SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
This thread is privately moderated by batfire, who may elect to delete unwanted replies.
Old Jun 12, 2012, 03:48 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Data
Tutorial for v1&v2 fc systems but useful to all quadcopters

TUTORIAL 1
Digital Compass and GPS for V1&V2

TUTORIAL 2
Compass / Yaw Issues and Troubleshooting

TUTORIAL 3
Multicopter Frame Tuning Adjustments /Electronics and PID Tuning Part 1 and Part 2

TUTORIAL 4
How to conduct test flights and find hidden gremlins

TUTORIAL 5
Flight Logs and Maintenance Log Examples

TUTORIAL 6
Questions and Answers
This tutorial will constantly be updated as more questions are asked

TUTORIAL 7
Kv, Motor Designations, Kv vs. Torque and Kv vs. Propeller Choice

TUTORIAL 8
How to save weight and have a nice 2 axis camera mount

TUTORIAL 9
Common issues and solutions for 2.4 GHz onboard Receivers and GPS interference

TUTORIAL 10
Multicopter PID / Gain Flight Control Theory Part 1, 2 and 3

TUTORIAL 11
After the Crash[/B]

TUTORIAL 12
Which battery is best for your applications



THIS SECTION IS FOR TUTORIALS ON THE XAIRCRAFT FC-1212-S AND P-SERIES ASSOCIATED ISSUES AND FIXES

NOTE: A LARGE PORTION OF THESE TUTORIALS APPLY TO ALL MULTICOPTERS



X-830-S series QuadCopter


TUTORIAL 1


DIGITAL COMPASS & GPS INSTALLATION AND TESTING for FC-1212-S & P SERIES X-830 QUADCOPTER


X-830-S Owners - This can apply to the X650 and X450 owners also but PID settings will vary and are not covered in this tutorial.

Introduction:

After spending considerable time testing both the FC-1212-S and P-series I think you will find the tutorials will cover most aspects of both series and there operations. The real difference in the 2 series is mostly firmware. The P-series has a nicer looking GPS head unit and the autopilot allows faster acquisition of satellites and better pacc counts. The AHRSV2 unit is slightly different from the AHRS-S unit in firmware and the capability to Auto-Gyro calibration on power up not requiring a level surface.

Another feature you may not realize is that with the S or P-series complete systems you will have 9DOF ability to achieve full functionality of the heading and attitude algorithm (quaternion, Euler angles, etc.).

You will find in the P-series XA Center an extra gain control for RTH feature which will allow you to slow the speed of return. You will also find a few differences in the RC settings section too with RTH standing alone and Show Lamp for locating downed copter.

The below section will be a short tutorial of the capabilities and how to install the FC1212-S, AHRS-S flight control system with Compass and GPS-S. As you know these mods are not available yet on the X-830-S Multicopter per Turbo Ace and WowHobbies the distributor. This also applies to the P-series unit. See below:

X-830 LISTED SPECIFICATIONS:

* Four separate 35A ESCs with 2mm spring-loaded motor connectors for ease of maintenance.
* BEC MODE: Linear
* BEC OUTPUT: 5V/2.0A
* LIPO BATTERY: 2200mAh, 25C, 11.1V (3-Cell) FLIGHT CONTROLLER:
* FC1212-P and AHRS-V2 auto-stabilization system, AP/AV based.
* FC1212-P flight controller with Camera Mount Gyro Compensation outputs :
* No additional purchase of gyros is required to operate the 1, 2 or 3-Axis camera mount. * 12 input ports with support for up to 12-channel RX.
* Controller includes 8 motor outputs which can be configured to quad, X6/Hexa, X8/Octo.
* 4 servo outputs for camera mounts, roll, pan & tilt, or 3-axial camera mount with shutter control servo.
* Modular designed with future expandability. Port for AHRS module, extra 4 expansion ports,
* Firmware is upgradable online with configurable parameters using the powerful XAircraft Configuration software.
* GPS module and compass will be available soon.
* 4 GB USB Flash Drive for setup, upgrade and electronic manual

WARNING! Installing these modifications is at your own risk and may void any warranty that may be on multicopter. I also have no affiliation with any of these companies. I am just an X-830-S owner that couldn’t believe the mods were not available or that the FC was not capable of implementing their usage.

NOTE: Before you say this has all been done before please read the tuturial. This mod has never been done to the X-830-S nor offered for sale by TA or WowHobbies. Some techniques used in this tutorial have NOT been used by XA in any publications nor video’s I have been able to locate online or in XA Center. Consider the installation a new fad if you really want this FC to work correctly with Compass and GPS.

These fixes apply to all working modes of X-830-S including NORM, ATT, GPS ATT (RTH), HH and CF.

WARNING! After each change you must (BENCH TEST) your copter. Once your bench tests have completed satisfactory you can then move on to the flight test.

( No amount of software (firmware) fixes will correct hardware errors.)

After many hours of research and testing of the X-830-S now accomplishes all modes of flight using the compass and gps. Looking at the pictures below you will see that the electronic compass and gps are mounted outside the top cover. The interior of the top cover and lower cover both are copper foiled to cut down on the EM from the electronic boards, wiring and esc’s along with any other equipment you may be carrying on board. If you are going to carry a OSD you will also need to foil between the OSD and esc’s but make sure you have plenty of airspace for cooling both esc’s and OSD.

Before attempting installation of the compass and gps your multicopter must first be correctly aligned , COG balanced and props and bells balanced with NO drift when hovering. Don’t forget to verify all motors are aligned in the horizontal plane to each other . One other thing to check if you have had a minor crash is that the arms are still straight. If not replace the bent arm, forget trying to straighten it. Your quad should be able to lift off climb to 10 ft. and set up a nice level hover in NORM and AUTO-LEVEL mode in a no wind condition. If you are having drifting issues and not able to hold a level hover do not install compass or gps until drift condition is corrected. A slight yaw left or right with no compass plugged in is normal.


Compass

Why mount the compass on top of the cover? The electronic compass uses the earth's natural magnetic field to find direction the same as an aircraft compass. Being a pilot the first thing I noticed was XA did not isolate the compass from the EM field. This is very important because even with a small aircraft you must guard your compass from EM from within and outside the aircraft. Aircraft compasses must have inspections done periodically to make sure nothing is influencing the compass. In aircraft something as small as a screw replacement on a instrument panel can influence the compass if the screw becomes magnetized. I don’t have the room here now for war stories. Just understand that the compass and gps work together so if the compass is giving a bad output to FC along with gps data what is the FC going to do??? The errors will keep growing and growing

That said, the compass should be mounted on top of the cover on a mount with vibration isolators as shown facing forward. The mount should also have copper foil installed under the mount facing toward the top cover. You must align, calibrate and test compass first and correct any deficiencies before installing a gps unit.

NOTE: Anytime you move from your old flying field to a new location ALWAY recalibrate your compass before flight. I also recalibrate every third outing when on home field and any time you do equipment changes.

UPDATE:There is one exception to the alignment procedure. If you are having drifting on Yaw axis to left or right note this from manual V1.03 for P-series and I believe by testing this would apply to S-series also.
Quote:“Note: Copter heading keeps turning to left or right drift slowly is normal when there’s no compass plugged-in.”


GPS

Correct installation of the GPS-S unit is very important if you wish to have a working unit, not a pile of debris on the ground!

The GPS-S unit is located under and attached to the top cover on foam isolators. A small hole is drilled in the cover to allow the gps cable access up the tower to the gps head. The distance from the top cover is extremely important.THE GPS HEAD MUST BE A MININUM OF 8” ABOVE THE TOP COVER. Period! If you place it any closer than that you will cause deterioration in signal and will start inducing Toilet bowl affect at minimum. I also use a copper foil isolator and support between gps and compass. I have found from much experimenting the GPS needs to be dead center over quad to help prevent TBE. All quads are different in where they install there esc’s and other assorted EM makers but with the X-830-S they are located in center under bottom cover. This is why the GPS must be at the tower height specified above. I have tried installing out on an arm but this still causes TBE and slower locking. This also happens if the tower is shorter than specified height above.

With the P-series GPS you may mount this unit with the kit provided with gps unit. Due to the increase sensitivity of the gps head and better shielding this unit will work well only 11/2 to 2" above the foiled cover.

This brings up another point. You must allow the GPS to LOCK before you move the quad. This is very important for 2 reasons. 1st If you move the quad before lock and engage the gps it will not know how to respond correctly and decide to go whichever way it wishes. This random flight movement is reported often by XA multicopter fliers. 2nd The quad will not know where home is and go its merry way when you select RTH, while you are dumb founded trying to figure out where and why. This can also occur if you loose gps lock intermittently due to incorrect placement and EM interference. This very lack of waiting until gps lock and loosing lock cause more preventable accidents with multicopters. I highly recommend that you add these features to safe guard your quad. Buy yourself a PILOT LAMP so you can monitor all phases of copter operation visually and audibly. I also highly recommend you buy a low voltage alarm with buzzer for your Lipo pack which can be set to produce a loud audible cue that its time to land. In the photos you can see both mounted on quad.

Now let us look at the settings needed on the X-830-S. Since I do not want new people or veterans with no prior PID tweaking experience jumping into the PID screens we will bypass setting up a personal profile and use the profile and gain adjustment fields supplied. First before we start Version Firmware 1.33 will be used throughout this tutorial. What does that mean to you in this firmware version? See below:

1. Height hold is automatically on. You can tune the gain or disable it in Flight Control section under HEIGHT CONTROL GAIN.

2. ATT mode has the position hold feature if GPS-S is plugged in.

3. GPS ATT mode will return to home. But this will change in feature firmware to reflect RTH and GPS ATT will be discontinued on script.

That means ALTITUDE HOLD and POSITION HOLD are automatic in version 1.33 if you have the Compass and GPS plugged in. This is great since it frees up some switches. Just to reiterate you no longer have a GPS ATT position on switch, this position is now the RTH mode.

So to wrap this up below explains each mode in the 3 position switch. I also explain which hardware the Flight Controller gets the information from. Your 2 position GEAR switch which prior to our changes was used for NORM/ HOVER mode will convert to Heading lock or Carefree mode if you only have a 7 channel radio. If you have 10 channel have fun putting them anywhere you like.

FC-1212-S & DEVO 7

I have these three modes on my DEVO 7 three position MIX switch.

Normal (NOR) Mode: Like most of RC helicopters, in this mode, the copter is not auto-leveled. It's the same as Cruise mode for previous firmware versions.

Attitude (ATT) Mode: The copter will auto-level by itself. The [U]GPS is plugged in not working but is sensed by the FC. Quad will hold the position when pitch and roll stick is released as long as GPS plugged in. (Position Hold)

(RTH) Mode: The GPS IS WORKING IN THIS MODE to help in hover and to return you to home. IMPORTANT:There's a delay before Position Hold entered in GPS ATTor RTH Mode, the delay depends on forward speed of copter.

 If speed > 0.2m/s, there's a delay. (0.447mph)
 The delay stops if the delay-time more than 2 second.
 The delay stops when speed is lower than 0.2m/s.
 If speed < 0.2m/s, FC enter position hold directly.


Heading lock is activated by ON/OFF Heading Lock Sw. I have this assigned to the 2 position GEAR switch. This feature requires compass to be connected and calibrated to work. See below in Devo 10 setting for CF mode info.

FC-1212-P AND DEVO 10

I have these modes on my DEVO 10 switches.

Norm, Attitude, GPS Atti is assigned to my FMOD 3 position switch.

Normal (NOR) Mode: Like most of RC helicopters, in this mode, the copter is not auto-leveled. It's the same as Cruise mode for previous firmware versions.

Attitude (ATT) Mode: The copter will auto-level by itself. The [U] GPS is plugged in not working but is sensed by the FC. Quad will hold the position when pitch and roll stick is released as long as GPS plugged in. (Position Hold)

GPS Attitude (GPS ATT) Mode: The GPS IS WORKING IN THIS MODE to help hold position in hover and and used in (RTH)mode. IMPORTANT: There’s a delay before Position Hold entered in GPS ATTor RTH Mode, the delay depends on forward speed of copter.

RTH Mode: The copter will return to home, speed of return is adjusted under RTH Gain setting in RC settings. This feature is assigned to the GEAR switch.

HEADING LOCK Mode: is activated by Heading Lock Sw. I have this assigned to the 3 position MIX switch. This feature requires compass to be connected and calibrated to work.

CAREFREE Mode: is activated by Carefree Sw. I have this assigned to the 3 position MIX switch. This feature requires compass to be connected and calibrated to work.

With limited switches on 7 channel you may want to have CF instead of HH so please understand the difference of CF vs HH. I will give you a quick explanation of the differences.

The only two differences between HH and CF, are that in HH: 1) yaw orientation will not hold on rudder stick release and 2) spatial steering is affected by yaw position ie, the quad will move in the direction its pointed with forward ele stick when the rudder stick is held off center and the quad is yawed. When you release the rudder stick, the quad will return to its ORIGINAL orientation and spatial movement will follow the "square" pattern you started with. In other words, you can fly the quad around with the rudder and elevator, but once the rudder is back to neutral - you fly spatially with the ele/ail stick. In practice, in HH, if using rudder, you'll use all three axes controls. If the rudder stick remains center neutral, the quad will fly exactly the same as it does in straight ahead orientation in CF mode.

In Carefree the unit remembers which direction the front of the copter was facing when you power it up. When you enter Carefree mode the copter will fly in the direction it was facing when powered up, no matter what direction the front is actually facing. For example, when powered up the copter was facing north. Take off, enter Carefree mode, and turn the copter so the front is now facing west. Give forward stick and instead of the copter moving in the west facing direction it will fly north while facing west. Or you can fly north while giving continuous yaw without having to worry about losing orientation. 'Forward' is always going to be to the north, no matter which way the copter is facing. This is a nice feature to use when taking video say of a moving object or fixed position as you fly over without having to worry about losing copter orientation.

Now let’s look at the settings I am using on the S-series X-830-S.

Stick Gains Pitch – 36, Roll – 36, Yaw – 36
Attitude Control Gain Pitch – 30, Roll – 30, Yaw – 25
Auto leveling gains Pitch – 40, Roll – 40
GPS gain – 30
Height gain – 35

Current firmware used for all systems onboard: FC1212-S V1.33, AHRS-S V1.23 and GPS-S V 1.01


These are the settings used on the P-series X-830-H.

Stick Gains Pitch – 40, Roll – 40, Yaw – 40
Attitude Control Gain Pitch – 35, Roll – 35, Yaw – 25
Auto leveling gains Pitch – 50, Roll – 50
RTH Gain - 35
GPS Position gain – 50
Height gain – 40

Current firmware used for all systems onboard: FC1212-P V1.36, AHRSV2 V1.26 and GPS Autopilot no updates

Due to issues with Firmware 1.34a for S-series DO NOT UPGRADE TO THIS FIRMWARE! It was replaced with Firmware Version 1.34b to correct issues with 1.34a but this firmware upgrade is not necessary since it really didn’t accomplish any major change.

Please read your flash drive manual for the Remote Control settings depending on the radio and Rx you have. The only thing I will bring up here is if you just received your X830 no matter if it was RTF or ARF go to this section to check and adjust your radio to the system and ALWAYS BENCH TEST UNIT FIRST after any changes. You will also be able to see which switch does what function. For the ARF crowd the rc stick reversal test screen will be a good tool to see what your servos are doing and see the response of the quad in the simulator.

Hope you have enjoyed the tutorial. Oh and one last thing, people new to quads please take note.

WARNING! These quads are no toy and will cut you to itsy bitsy pieces if you do not respect them and keep your hands and other body parts clear. The most common culprit in injuries with multicopters is the THROTTLE which is inadvertently caught on neck strap when putting on or taking off transmitter. The next common culprit is laying transmitter down on ground and strap lands on throttle. There are many stories and pictures of injuries sustained just in this manner.

Well that is it for now, if anyone should want more info on installation etc… you can drop me a PM and I will get back to you.

Oh, and the disclaimer: Not responsible for loss due to crash, freak of nature, earthquakes and whatever else you can imagine, your results may vary and can be significantly affected by incorrect procedures, working on multicopter in the dark, bad karma, the wife left me and the dog died syndrome or your medication is off.
batfire is offline Find More Posts by batfire
Last edited by batfire; Apr 24, 2014 at 11:37 AM.
Reply With Quote
Sign up now
to remove ads between posts
Old Jun 19, 2012, 12:04 AM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Proof of concept flight test footage X-830-S&H: FC-1212-S & FC-1212-P complete system

PROOF OF CONCEPT," FLIGHT VIDEO OF FC-1212-S & P SERIES FLIGHT CONTROLLERS ON X830 COPTERS WITH COMPLETE COMPASS AND GPS ADDITIONS"

Hi All,

Well I had many request for "show me the video proof" . Well here it is and I purposely picked the worst spot I could find to fly out of to prove once and for all that the FC1212-S can do what it was originally suppose to do if you do my mods. The area is 20ft from tree line to center field on each side and only 10 ft behind me and 100ft to the telephone poles and wires. It is tight with no room for software and hardware errors.

My research for this phase of flight testing is complete and the next phase is underway.

S-series video test flight w/ Compass & GPS system:

http://www.youtube.com/watch?edit=vd&v=kYZ4f9B73Sc

BTW this was my first use of the gopro and I rambled a little bit and repeated myself also please forgive I was a little nervous to be narrating and flying at same time.

P-series video test flight w/ Compass & GPS system: Edited just to cut the time down on video.

My Edited Video (11 min 26 sec)


UPDATE: Just completed testing of P-series 8/15/12 . The P-series was tested and filmed out of the same location where the S-series was tested. This makes for easy reference of flight stabilization in all modes on both series.

After many hours of test flight time I would NOT recommend flying the P-series with compass and gps while carrying expensive camera gear onboard. The V2 firmware is not stable over time and can react unpredictable. I believe this is due to introduction of the new AHRSV2 with auto gyro on startup. The AHRSV2 seem to change attitude from flight to flight and as flight time progresses from last calibration reset which seems to show instability in the firmware interaction with the AHRSV2 and compass. This error condition is compounded with the GPS also connected to FC. Many people have noticed this issue but not known what was happening or causing issue. Things such as roll drifting in hover left or right but usually left and forward after checking and rechecking copter balance and using bubble level or AHRSV2.

Many who have notice this behaviour report that the drift gets more and more prominent and steeper as time goes by. I have found with both bubble level and electronic leveling you still have this occur as flight time goes by. The issue begins occurring during the 1st flight after recalibration and progressively gets worse flight after flight. For normal flights with just a gopro onboard (close in low flights) or with fpv gear on board (open sky flights) you would be ok but for shoots from altitude I WOULD NOT RECOMMEND carrying expensive camera gear onboard the P-series without at least an FPV setup on board to back you up.


UPDATE 2: As I have reported on the X830 site post #306 I was doing a high flight around 400 ft and just turned in gps hover mode to the west filming when the copter decided to start going backward and it started climbing until it went out of sight and flew away to never-never land. I had failsafe and RTH activated and it never responded. I have done this kind of flight in same location over and over again so it wasnt caused by Xmitter or reciever failure.

I never saw a news story so it must have crashed in woods somewhere thankfully. I had over 3grand in that copter including a full flight recording system which if I am able to ever retrieve the copter will know what was going on with FC and flight controls through the whole flight including sd video footage. I still had the testing rig on board due to the anomaly I have been recording in the AHRSV2. I have come to the conclusion I will not be able to find it due to the area we live is hilly and very wooded. I never installed the gps tracker equipment due to working close in. My bad.

I have to admit that the V2 P-series for me has also been a very unpredictable FC and even with all the testing and flight time it never was a really rock solid FC as is the V1 S-series. The S-series flight control has been rock solid and does exactly what it is told. Period!
batfire is offline Find More Posts by batfire
Last edited by batfire; Oct 31, 2012 at 02:25 PM.
Reply With Quote
Old Jun 19, 2012, 04:03 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Parts, parts, parts, For the V1 S-series Compass & GPS addition

MODS ON THE X830-S FOR S-SERIES COMPASS & GPS

I've had a few questions about where to get the parts to build the mods I have.

I will give a list of required parts, and where you might find them with prices listed.

First here is list of what you need. Caution: Only buy V1 parts! Version 2 (P) series are not compatible with V1 except the compass.

(1) 1 Turbo Ace Lipo Battery Voltage Meter With Programmable Buzzer Alarm
(2) 2" ground plane kit for gps head which includes copper foil and tower parts.
(3) 1 electronic compass with connector to AHRS-S
(4) 1 GPS-S with gps head etc... as seen below or you can buy parts individually
(5) 1 Pilot lamp

These are possible locations to buy parts and their listed prices.

WowHobbies:

Turbo Ace Lipo Battery Voltage Meter With Programmable Buzzer Alarm : Measures 1-6 Cell Lipo Batteries, 2.8V-25.2V, For All Helicopters
Price: $9.95

CNChelicopters

Xaircraft GPS compass first version special upgrade package
Price: $10.00

XAircraft X650 upgrade parts Electronic compass New
New style 2012
Price: $53.99

xaircraft x650 x450 GPS ACTIVE ANTENNA
Price: $15.00 This is the 15" replacement for old antenna if you get GPS-S off ebay etc...

New XAircraft Pilot Lamp E1030 New style with plastic cover
Price: $10.60


-----------------


Cnchobbies is out of the gps-s for now so I did a quick look and here is one source

Free shipping!!!X650 - XAircraft GPS-S module accessories
Price:US $221.05
http://www.aliexpress.com/fm-store/7...cessories.html

or maybe complete set Gps-s compass and 2" ground plane.
http://www.aliexpress.com/fm-store/7...mpass-set.html
Price: $312.00

even better deal Hobby-wing XAircraft GPS and Compass Combo (Latest Active GPS Antenna)
Price: $249.98
http://www.hobby-wing.com/xaircraft-gps-compass.html

I'm sure you could go to eBay etc... for some of these parts but I would only buy new off eBay. I would not buy the gps or compass off ebay due to you dont know if it has been damage in a crash etc....unless from a storefront shown as new on listing in box.

-----------------

http://quadframe.myshopify.com/
Quadframes handles an excellent inexpensive skid system ($85.00) very well constructed -- Materials: Fiberglass G10 2.0mm thick, aluminium tubes 6-8x1mm -- along with an gopro 2 axis gimbal ($69.00) -- Materials: Fiberglass G10/G11 1.5mm thick SATIN BLACK COLOR -- that is super smooth. All these parts come from Poland and I highly recommend these kits.

--------------------------------------------------------------------------------------------------------------------------------------
MORE MODS AND IDEAS FOR EXPANDING THE X830 QUADCOPTER
--------------------------------------------------------------------------------------------------------------------------------------
Hey, how about adding dual batteries to this copter for longer flights but keeping weight to minimum.

This was brought up a little while back but after doing a little thinking on the subject I have decided this is an easy way to add a dual battery system to your copter with the least hassle and have a nice clean setup.

The TA-X830-35: Super Structures for X830 kit now available at WowHobbies is a usable platform which would work great for adding batteries. Let me show you the kit and then my diagram for mounting batteries. You will need these items shown:

Update:

(since this tutorial WowHobbies has made the dual battery holder available please see their site.)

Total outlay without s/h would be $63.00 or so +s/h on each. The structure would be obviously twofold. First to strengthen the arms of quad, and second to hold battery trays. Use the collars to lock structure in place on legs. Use Velcro and wow straps to lock batteries in place. See pics below:


Hope this helps.
batfire is offline Find More Posts by batfire
Last edited by batfire; Apr 25, 2014 at 02:13 PM.
Reply With Quote
Old Jun 28, 2012, 03:02 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Compass / yaw issues and troubleshooting

TUTORIAL 2


COMPASS / YAW ISSUES & TROUBLESHOOTING


This is a short tutorial on how to calibrate the AHRS and Compass correctly. This info can also be found in the Xaircraft Center. I will try to give step per step instructions and follow up with others videos on this subject. At the end of this tutorial I will refer you to my blog above for compass placement and trouble shooting techniques. Do understand, "any changes that are different from XA recommendations are at your own risk". What else is new?

A few Warnings and Precautions before we begin:

1st Any time you are working on your copter in this manor always remove the propellers.

2nd Completely read through the instructions before you do an operation.

3rd Always, Always “BENCH TEST BEFORE TEST FLIGHT”.

In calibrating the AHRS-S OR V2 you will need to have access to the internet and be able to log into the XaCenter. Go to the AHRS tab which will bring you to the Calibration screen. The first Calibration shown is for the AHRS itself. Normally there is no need to recalibrate the AHRS unless you fit the following circumstances:

1st After a firmware upgrade

2nd AHRS is not sitting level and needs to be re-leveled and shimmed on FC.

3rd Flight environment temperature has changed greatly i.e.… winter to summer, summer to winter seasonal change.

This procedure assumes that you have done the following:

Before attempting installation of the compass and gps your multicopters must first be correctly aligned, COG balanced and props and bells balanced with NO drift when hovering. Don’t forget to verify all motors are aligned in the horizontal plane to each other.

Your quad should be able to lift off climb to 10 ft. and set up a nice level hover in a no wind condition. If you are having drifting issues and not able to hold a level hover do not install compass or gps until drift condition is corrected.

UPDATE:There is one exception to the alignment procedure. If you are having drifting on Yaw axis to left or right note this from manual V1.03 for P-series and I believe by testing this would apply to S-series also. Quote:
“Note: Copter heading keeps turning to left or right drift slowly is normal when there’s no compass plugged-in.”

Once your online in XA Center go to the tab AHRS to start this procedure.

Step 1 When calibrating the AHRS in copter, level the copter and AHRS with bubble level on top of AHRS. Make sure AHRS shows a centered bubble then without touching copter click the calibrate AHRS button. It will take a few seconds to calibrate. Once this is completed check your results on the 3D Diagram. You should be very close to 0 in pitch, roll and yaw. There will be minor fluctuations but this is normal.

Step 2 Make sure compass is plugged in correctly which will allow you to click the option calibration at next power on. You can then go to the bottom right of screen and click disconnect.

Step 3 Now take the copter outside and as far away from any electrical wires, metal etc.

Step 4 Power up the Tx as normal and make sure throttle is closed. You only have a second or so after copter power up to start calibration procedure so make sure you are ready to go. Plug in copter and you will see the compass light begin to fast flash. You will do Horizontal calibration as soon as light starts flashing slowly. You will then turn copter flatly 360 degrees continually until the light goes back to a fast flash.

Step 5 Second fast flash is to get ready to turn the copter 360 degrees on vertical plane continuously. Once light starts flashing slowly do the vertical calibration with the compass arrow pointed either up or down in vertical plane. Hint: I normally put quad to chest and do 360 rotations in place. Seems to work just fine.

Step 6 Once light on compass goes to solid you have completed calibration procedure. For clarity see diagram below.

Step 7 There is one last step they don’t mention in the XaCenter on this procedure that is VERY IMPORTANT. You must after calibration power down the copter and Tx. Then power back up. For some reason if you don’t you will sometimes get wild reading and the copter will try to flip over.

Here's the video tutorial for Compass calibration outside the copter. This is a good technique to use if having AHRS or Compass issues such as 0 degree glitch. It will alleviate the quad from the equation.

:
XAircraft AHRS & Compass Calibration (7 min 1 sec)


Here's a video calibrating Compass outdoor: http://v.youku.com/v_show/id_XMjU3ODc0MDIw.html from time 2:35, the video showing how to calibrate Compass according to the light flashing shown in the video tutorial.

I have one last thing to mention here. I do not like the way XA wants you to install the Compass on top of the AHRS. Why? Please read my above blog entry: Digital Compass and GPS for V1.

I hope this helps answer some of the questions on how to calibrate and troubleshoot common problems with the compass unit.

Don’t forget all the nice testing screens in the XA Center area for all the electronics on your copter. Just remember to KISS and you will do fine'

Some helpful hints, first check the compass wiring to make sure plug is plugged in tightly and not loose. Check to see that nothing is coming into contact with compass such as other wires etc... If all else fails remove units and test on bench then mount compass outside of cover on top as I have and this will help reduce a lot of EMI related issues.
batfire is offline Find More Posts by batfire
Last edited by batfire; Nov 02, 2012 at 01:29 PM.
Reply With Quote
Old Jun 30, 2012, 10:06 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Multicopter frame tuning adjustments

TUTORIAL 3

Part 1


MULTICOPTER FRAME TUNING ADJUSTMENTS / ELECTRONICS AND PID (PROPORTIONAL-INTERGRAL-DERIVATIVE) TUNING

This is a two part tutorial on multicopter adjusting and tuning. In this first section we will talk multicopter tuning and adjusting from a mechanical and balancing point of view. In the second section we will talk about how to tune your equipment for maximum stability in flight.


Multicopter Frame Tuning Adjustments (COG)

Center Of Gravity:

COG is the first thing we need to determine before proceeding on to the next part of tuning the airframe. This procedure is a quick and easy way to determine your X-Y-Z balance points.

1st Determine where the X and Y axis are by hang testing the multicopter from the center top or bottom. You may be required to move components around if a DIY design or additional equipment was added from the factory kit to cause an imbalance condition. This center balance should be easy to accomplish.

2nd Use a table top edge and tilt the copter on the skids until the copter finds a balance point. While holding copter have someone take a picture. Now repeat and balance on side skid. Again take a picture.

3rd Print photos and draw a line depicting where you balanced the multicopter down the center during hang test.

4th Now draw a line vertically up from the skid tip where you balanced the quad. Now draw a vertical line from the side skid up. Where these lines intersect will be your Multicopter COG.

This information will be useful for future alteration of the multicopter.  This info also helps us eliminate balance from our checklist in the event we should have drift issues with copter.

Frame Checks

Next we need to go completely through the frame and check all bolts and nuts and make sure all are tightened and Loctite all required nuts.  I use Loctite blue on frame work nuts but if the nut will directly affect a plastic or CF piece it is imperative that the Loctite not contaminate the piece. For these nuts I use a small drop of CA to hold them in place.
 
While we are investigating and tightening nuts it is a good time to check all electrical fitting such as bullet connectors for tightness. Bullet connectors have been a frequent cause of motors stopping and crashes resulting in multicopters.

First look at the spring on the tip of the bullet connector. If you are able to spin it on the shaft, then these are known to eventually cause glitches and failures. The best thing I have found to do with these for the ultimate multicopter reliability is to solder the wires directly and do not use bullet connectors at all.

If you do not wish to solder wire to wire then the next best thing is to take a small knife blade and open the springs up and before connecting slide on a shrink tube. Once I am sure they are as tight as possible I then heat shrink them in place. Some people also slide a little solder on and slam them together before heat shrinking but I have never had one fail after just heat shrinking them.

Trim and Tx Checks

It is now time to check that all trim is centered in the Tx and make sure sticks are also centered in the Remote Control section of XAircraft Center.  To do this we are going to USB into XA Center and go to the Remote Control tab.  Within this tab we will see General Input section.  This is where we can see what each stick movement is doing real-time.  We want to make sure when we stop moving sticks that each centers in middle of it throw range with exception of the Throttle.  

If for some reason your stick throws are off then all you need to do is go to the RC Calibration section click the Calibration button and follow the instructions step by step.  Save your calibration once you are finished.  There is a nice video that XAircraft puts out showing how to do this which I will include below. 

Video of calibration: http://v.youku.com/v_show/id_XMjIyNzc1NDg0.html

Motor Maintenance

Now this might be a controversial topic but motor maintenance is very important to the electric motor life span.  The issue comes up with do I or do I not oil and grease the bearings on the motor and around the circlip on bottom of motor?  The answer seems to be how often do you wish to oil the bearings once you start?   If you decide to oil then once every 5 hours of flight time is about right.  When you oil or grease this bearing clean off any excess so you won’t have a dust and grit collector.  That also goes for the bottom of the motor greasing the circlip, otherwise known as the Jesus clip.  

I feel one of the best ways to get longevity out of your motor is NOT to tighten the crown nut more than 1/2lbs of torque when putting the prop on.  When you tighten any further you start pulling the Jesus clip up against the housing of the motor.  This causes at minimum an increase draw on the battery and at most an inflight motor or esc failure.

  A good system I use is after I take prop off for inspection I screw the crown nut back down hand tight and take a small rubber mallet and tap a couple of times.  If you are drawing the shaft up when tightening unknowingly this will free up the motor shaft and you will see the difference when you put the prop back on and flip it around.   I truly believe some of the videos I have seen with people asking why did I crash was caused by just this issue.

Another thing that must be checked is the motor shaft for straightness.  If there is any type of wobble in the shaft replace the motor and order a new shaft if motor is fairly new.  Once motor gets a lot of hours on it it’s just as well to replace the motor.  

Just a quick list of motor killers to watch out for: 

 Dirt ingestion, stones, debris including wet grass, heat stress, crash impact tension, wire failures at bullet plugs, thermal runaway due to OVERLOADING THE COPTER, bad solder joints, poor bearing maintenance just to name a few.

 If you see a slower running motor (stop/start) dragging, hesitation or stuttering land NOW and call it a day until you have completely corrected issue.  If this occurs as you are taking off ABORT your takeoff and call it a day.

 Real Pilots at the slightest hint of a problem discontinue the flight and land ASAP. 

UPDATE: This is why you always run new motors on a bench test for at least 2 battery packs before flight.

I had a motor on my quad that started momentarily seizing. In-flight this momentary seize would cause a flip and crash. The reason for the seize turned out to be a small washer in the armature from the factory.

I have seen more than one occasion this has happen to motors and takes a little break-in-time to be noticed. First symptom is very slight drag when starting or stopping motor. Unattended this will within an hour of flight time or so become an in-flight motor momentary seize.

If you are not attuned to what is happening you would write it off as wiring, bad solder joins etc after the death flip crash. You will read in later post what I did to find this issue.

This brings up another point; these are miniature versions of aircraft no matter how you look at it.  Keep an aircraft log from day one.  That makes it a lot easier to record wear and tear and know the true hours of operation.  Always keep a log it will pay for itself a hundred times over.

Examples of logbook entry:

1. Description of work performed ie. replaced props, greased bearings....
2. Date you completion of the work
3. Crash of hard landing and damage.
4. Parts ordered, prices
5. Last calibration and last updates added
6. Total flight time on multicopter
7. Total time on propellers before replacement and why replaced
8. Parts flight time until replacement due to wear or breakage

The list can be as long or short as you wish but you should keep a log so that you can see time on unit and when things had to be replaced. This allows you to set up a maintenance schedule.

Here is a great site with explanations and videos on maintaining your Brushless Motors and extending there life.

UPDATE: After working with Boca bearings we have found the replacement bearings for the TA motors. This is the bearing size etc......4x9x4mm and just email Michael at Bocabearings for replacement upgrade bearings when you need them. Prices vary depending on what quality level you desire. Bearings do make the motor when it comes to longevity and quality. If you look at your high end motors they all have the top quality bearings and get easy 600 flight hours with no issues as long as you dont crash the motor. Even small crashes that dont show damage can actually damage your bearings.

RC BRUSHLESS MOTOR BEARING MAINTENANCE & INSTALLATION
TIPS FOR REMOVING SHIELDS & SEALS

http://www.bocabearings.com/dropin.a...stallation.txt

RC Brushless Motor Bearing Replacement

http://www.bocabearings.com/brushles...t-article.aspx

RC BRUSHLESS MOTOR BEARINGS LISTED BY MODEL

http://www.bocabearings.com/bearing-...rings-by-model


Propellers


1.Blade Tracking - Most pilots overlook this one but it's, by far, the biggest contributor to vibration. Hold your copter up to an incandescent light source and look (not too closely) at the blade tips horizontally while it's spinning at low RPM. If you see a blur, rather than a fine point on them then your blade tracking is off. This is usually caused by mild crashes or blade strikes that you thought didn't do any damage but in fact slightly bent a blade. The solution to this is to toss them for FPV flight but they can still be used for an emergency prop.

2. Motor bells - I found that prop balancers (even the precision magnetic ones) are often useless because they don't seat perfectly with the motor bell and you end up with worse (not better) balanced motor bells in the end. Solution is using a multi balancer because it has several different shafts from 2mm thru 5mm. You'll have to remove your motor bell and shaft to balance it. A common practice that I use to balance motor bells via trial & error is by using a tie lock tighten down to one side and run the motor up listening for vibration and tone of motor. Move the tie end around until you have the least tone variance then add weight to that side of bell. That is better than nothing but the balancer is much more accurate. It'll tell you exactly where the heavy spot is (even when there are multiple spots) and how much weight to add/remove to perfectly balance it.

There is also one more variance where the prop is left on the motor on the copter and one motor is electrically connected at a time. They use the throttle for spin up and access the tone balance adding tape to the prop to balance both the motor bell and the propeller at the same time. I like the other ways better due to I just can't hear the tone change with it on the copter but if you can it sure saves time.

3. Mount and Arms - Do not forget to verify your motors and arms are aligned in the horizontal plane to each other. This will alleviate a drift issue due to miss alignment of arms or motor mounts.

4. Props – Props are easier than Motor Bell balancing but can be just as time consuming. For this choir just use any decent floating magnetic prop balancer. I use the magnetic mount and balancer from WowHobbies. I balance my props and hubs using scotch tape for the blades and a small glue gun for the hubs.

NOTE: This comment I feel is very important. How do you correctly install the propellers on the shaft and tighten? First do NOT use Loctite OF ANY TYPE on your propeller shaft or crown. The reason, it is the number one cause of prop hub deterioration which leads to hub failure in flight! Instead find some thin nylon washers and use those under the prop. Insert a small screwdriver through the top hole of the crown caps and tighten it in a clockwise direction. When the crown cap lock touches the propeller, use no more than half pound of torque to lock the propeller in place. This is very important. Over tightening the propeller is the next major reason for prop failure and also damage to shaft and crown caps. Below I have added a few videos of tuning motor vibration and prop and hub balancing.

One last thing on props NEVER EVER use a PETROLEUM BASED product on any part of the copter that will interact with plastic or Carbon Fiber. When you need a lubricant use Silicone instead such as the prop shaft. Only use a tiny amount wiped off to create a film on shaft which will keep your props from becoming a permanent fixture to the shaft.

Source for replacement wood props http://www.rcgroups.com/forums/showthread.php?t=1779091

Video on tuning motor vibration

http://www.youtube.com/watch?feature...&v=OgvayWfey78

PROP AND HUB BALANCING

prop balancing video tutorial (5 min 5 sec)


HorizonHobby.com How To - Prop Balancing (13 min 41 sec)


This is the end of part 1. In the next segment we will talk about the electronic side of Tuning.
batfire is offline Find More Posts by batfire
Last edited by batfire; Dec 04, 2012 at 11:22 PM.
Reply With Quote
Old Jul 05, 2012, 10:32 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Electronics and pid (proportional-intergral-derivative) tuning

TUTORIAL 3

Part 2


ELECTRONICS, ELECTRONIC COG, SHIELDING & PID (PROPORTIONAL-INTERGRAL-DERIVATIVE) TUNING


NOTE: Make sure before you read this tutorial that you have completed reading the first 3 tutorials so that by the time you get to this tutorial your copter is ready for these adjustments.


ELECTRONIC AHRS (IMU) Center of Control:

You will notice on my V1 multicopter in the below photos that I have the gps antenna directly over center of quad and 8" above the top cover and below explanations will be based on this location. There is a reason beyond the one I explained in the first Tutorial. I have found through bench and flight testing that there is a dual pendulum effect which can be incorrectly assumes as TBE. From testing and doing the math I have found an answer to solving this effect. I will skip the entire math lesson and just give a quick understandable explanation and example.

All multicopters have a center of vertical balance which is easy to find by doing a hang test of multicopter and adjusting the equipment around to obtain a good center balance. Next you need to find the center of gravity of multicopter. You can do a quick and dirty COG by tipping the quad on its skids until a balance is achieved. Take a picture of the multicopter and then draw a line from the vertical center of copter down and a line vertically from the skid balance point up. You can go one step further if you wish and balance quad on side of skid and draw a vertical line off bottom of skid upwards. Where those 3 lines meet is roughly the COG. In my example below in picture 1, I am using 2 lines to locate COG.

This is where it starts to get fun. The AHRS is the center of the universe for the multicopter and where all IMU leveling is derived from. On your picture draw a line horizontal through the middle of your AHRS. This is your center of control. Now take the distance from your COG to your center of AHRS. This number will always be a negative number. Let's use the example of my copter shown in the first picture. From my COG to center of AHRS is -9 cm. The distance from center of AHRS to GPS head is 19 cm. This number will always be a positive number. Take the difference of these two numbers and you get 10 cm. Now if we transpose the number we come up with a -1 difference. This fits the guidelines I have come up with for total quad balance electronically and physically. As long as the value of the difference in height of gps head vs. distance from COG are between 0 and -2 you will never see the TBE or pendulum effect causing the wondering and circling during position and altitude hold times.

Think of it as 2 pendulums working against each other. See picture 2 below. One pendulum is the gps height to center AHRS and the other is the COG to AHRS center. By keeping the final number between (0 to -2) you will cancel the effect. Picture 4 shows a gps mounted below the center of control which will cause the wondering effect.

So to sum this up the software was really not as much the issue, the dual pendulum effects were causing the wondering and circling (tbe) effects. Note this assumes you are mounting gps over center of multicopter. There is another factor called offset antenna location that must be taken in effect if gps is to be mounted away from center of multicopter as seen below in the 4 picture. That’s a whole different variable not covered in this tutorial.

SHIELDING EMI / RFI

When first installing your equipment please make sure you line the bottom of compartment with copper foil tape. This will require unscrewing and lifting FC to foil underneath. Then copper foil the esc wires leading up from bottom. Next copper foil the top cover with multi-layers of copper foil but leave all vent holds clear of foil. I use over 6ft of foil to insure maximum protection from EMI above this cover. This is why you mount the compass outside of cover and not on the AHRS. Make sure to Copper foil under the mounting plate for the compass. As it stands now you would then mount the gps head 8” above the top of the AHRS. This allows you to be on a groundplane 2”x2” patch out of the EMI field. IN the future I will come up with a better, prettier alternative. I promise. See diagram below for reference of EMI fields vs equipment and copper foil locations. This diagram is valid for V1, V2 and Naza installations.



ELECTRONICS WORKING TOGETHER – GPS, COMPASS AND AHRS

Before we start on this section if you look to the bottom of this tutorial you will see 3 pictures of ckt boards. Only a few people actually take the time to open up their units and see the circuitry involved in this Flight Control system. The main FLIGHT CONTROL board comprises of three, ( STM32F103C8T6 advanced arm-based 32-bit MCU with FLASH, USB, CAN, 7 TIMERS, 2 ADCs, 9 COMMUNICATION INTERFACES ), chips. There are 2 of these on front and one on back of this card. This card also has the 176333 IS 3.3V Low Noise Regulator on front of the card. This FC has a lot of potential for future add ons.

The next card below is the AHRS-S card which provides real-time flight attitude and heading for the flight controller. The AHRS is sensitive to vibration; it can stop working properly under serious vibration situations. It is extremely important to install the AHRS on an anti-vibration pad when installing. It's also just as important to install the AHRS in the center of your multicopter.

The third picture is the GPS-S ckt board. This card allows you the capability of Altitude and Height Hold. The gps head for this unit is EMI sensitive and must be mounted high above the other components and away from EMI field.

The AHRS, Compass and GPS all work together to level copter, give you directional data along with position and altitude data. One thing to remember when working with only the FC and AHRS, you have no directional data. This means you will see yaw drift whether it be left or right. Once you plug your compass in and calibrate the drift will vanish.

Always make sure when you plug in and power up your multicopter do not touch the controls until the AHRS on board gyros have stabilized along with card temperature. (about 1minute) This also allows time for your gps to lock on to required satellites. Do not take off until you have a lock on your gps system if you are going to use ATTI OR GPS ATTI. Both require gps input for values.

When you power up your multicopter make sure it is placed facing direction you wish to do most of your traveling. This helps your orientation when you activate Heading Lock or CareFree mode. This also helps in verifying your gps has good tracking during flight.

To verify your gps tracking line up the quad with a distant object. When you lift off fly toward this object and watch track of copter. If you find the copter tracking off to the left or right of the straight line to object just re-align gps head at a slight offset angle toward the straight line to object. This may take just a few attempts to get an exact correction angle established and complete the adjustment. See diagram below.

Why does the gps need to be corrected with offset angle? This has to do with your magnetic declination and the exact offset will depend on where you live. Remember the gps and compass both work together with the AHRS to create a 3d environment in which they operate in. BTW without doing this little test and adjustment you may find after a long run out that when you hit RTH the multicopter may wind up not coming back to your location. Depending on range out you could be off as much as 30-100ft. This could be the difference of a tree line and not being able to see your multicopter if you hit RTH say after LOS during FPV flying. Where do I find what my Magnetic Declination is for my location? http://magnetic-declination.com/ Once you locate your home flight location click on location and it will give you your MD. Now if you are negative turn gps to left, if you are positve turn gps to right the amount of offset needed.

A word about (TBE)

Toilet Bowl Effect can be caused by (1) your FC not having the function of setting up an x,y,z axis for location of your gps relative to your AHRS, (2) AHRS not located in center of copter on COG, (3) copter not balanced correctly, (4) motors and arms not adjusted to horizontal to each other.

Example of TBE would be lifting off in NORM mode and all is well, switching to Atti mode and also all is well, switching to Atti GPS mode and copter locks on location and starts a slowly increasing circle to CW. If you do not switch back to Atti mode or NORM mode the copter will go to such a roll angle that it will spiral down and crash. This is true TBE.

So how do we correct this issue if balance and centering are correct and you have set the gps to the correct MD? This is where you have to experiment a little to get the exact angle that the gps head will need to be turned into the turn. If your TBE turn is to CW then you will need to turn the gps head to the left until that CW motion stops. This will be the same procedure if turn is CCW. You will turn gps head to right until the TBE stops. That is all there is to it.

PID (PROPORTIONAL-INTERGRAL-DERIVATIVE) TUNING

WARNING! : To tune the Flight control settings, you need a complete understand of PID auto-controlling theory, please don't tune the parameters until you have learned how to correctly. Tuning PID parameter may cause DANGEROUS RESULTS.

What is the function of a PID controller?

The PID controller has three principal control effects.

The proportional (P) action gives a change in the input (manipulated variable) directly proportional to the control error.

The integral (I) action gives a change in the input proportional to the integrated error, and its main purpose is to eliminate offset.

The less commonly used derivative (D) action is used in some cases to speed up the response or to stabilize the system, and it gives a change in the input proportional to the derivative of the controlled variable.

The overall controller output is the sum of the contributions from these three terms. Due to the lengthy amount of space it would require to teach PID I will direct you to an excellent video on PID tuning which also shows the affects of tuning on a multicopter and other very helpful sites with sims.

http://wiki.openpilot.org/display/Do...ng++Multirotor

http://www.openpilot.org/pid-tuning/

Here is a nice simulator to make PID adjustments and fly them or not

http://www.jasonshort.com/

Here is a nice site with explanations and simulator, very well done.

http://code.google.com/p/arducopter/wiki/AC2_Tweaks

There are tons of information on the net explaining PID tuning with multicopters and many papers written also. Take the time to study and experiment in small scale with PID tuning and you will find your copter experience will be more enjoyable. Start out small working with the gains you have in your copters flight control section of XAircraft center. Learn the affect of tuning these gains first working with only 5% changes. Always write down your original settings before changing settings. Now go have some fun learning.

UPDATE: Please see Tutorial 10 for additional information and a non-technical approach to PID / Gain Tuning Theory.
batfire is offline Find More Posts by batfire
Last edited by batfire; Feb 02, 2013 at 03:30 PM.
Reply With Quote
Old Jul 17, 2012, 12:24 AM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
How to conduct test flights and find hidden gremlins


TUTORIAL 4


CONDUCTING TEST FLIGHT INVESTIGATIONS FOR HIDDEN ISSUES


I was out on a test flight recently testing the FC / AHRS combination when I had a “flip of death” episode. Fortunately it only cost me emotionally. I was only 2 ft off the ground in a heavy grassy area so it was a flip strait down and flops upside down type of landing. This test flight was brought on from an episode a few days before that cost and arm on the quad. Since that time I had not been able to replicate the “flip of death” even after 30 minutes of bench test time.

This time I was ready if it reared its ugly head. About 10 minutes into the hover in ground effect I was feeling a little better when all of a sudden the left rear motor momentarily quit and the death flip followed. It happened almost instantly but I was ready for it watching closely. If you were just flying around you would not have known what happen just a death flip with total destruction. In my case I was hovering 2 ft above a trampoline. This is why I test in or near ground effect, no distance to fall. I was able to narrow it down to either the #3 motor, wiring, esc or FC slot 3, only because I was close and could see the split second seize that lasted momentarily. After the crash I tested the motor and it started up along with the others. There was no delay or hesitation, all motors started and stop together.

What I learned from this flight was where to look for this gremlin. The first thing I tried was to recreate the problem with all motors removed from FC and only #3 connected. Within about 3 minutes at 50% power it did the dive. I then moved it from slot 3 of FC and attached another motor to that slot and ran the motor. No problem at all with another motor in slot 3. Hmm… ok I then went looking at the wiring. All wiring looked good so to prove it back to motor and not esc I replaced #3 motor with a spare motor I had. It ran fine so I put the old motor back on and again after a few minutes it took the dive. I replaced #3 motor with spare motor and decided that it was doubtful that WOW hobbies would take the defected motor back with a total logged time of 34 minutes.

To Wow Hobbies correct it was not an issue that they would have picked up in the brief test flight they do before shipping. This took 25 minutes of flight time before it developed. I do expect better though from these motors due to their cost.

I decided I had nothing to lose by taking the motor apart and low and behold guess what I found. Look at pictures below. Seems the factory left me a little gift. This small washer had been in there since day one notice how worn and shiny it is? This is another craftsmanship issue along with the motor wiring and esc to battery thin wiring found with this quad. This washer is the same kind of washer used under the jesus clip when putting motor together. I guess they gave me a spare.

Look at the picture of my replacement motor wires. Notice again cheap thin gauge wire on this motor, same as I have found throughout this quad. The pictures below show the April quad motor and chassis wiring is twice what this quad is and has had no issues at all. This wiring is ok as long as there is no real load on the wiring system but under an AOW load it will not be so. I have replaced most of the wiring and may have to replace all wiring onboard with heavier gauge to my other quads specs soon.

After reinstalling the motor I felt I had fixed the issue and the damaged motor might be usable. I then did bench testing and again it failed after a few minutes so the wiring in motor must have also been damaged. It looks as though I will have to write this $60.00 motor off.

In this hobby it doesn’t matter how much the parts cost it’s becoming a hit or miss whether you will get a good motor or quad or not. This one has had gremlin after gremlin, but this is part of ownership and troubleshooting your quad is worth the effort gained in knowledge.

I believe this is the motor C2809-30T Kv 600 http://www.lotusrc.com/xxzl.asp?ID=11 This motor has been almost impossible to find info on, it is not C2806 as shown in parts list unless there is something I am missing here.

---------------------------------------------------------------------------------------------------------------------------------------------------------------------

This is an example of diligent troubleshooting by another quad owner having issues with his AHRS-S. I find it very informative and explains why we must look at the firmware side as well as the hardware side of Multicopters when troubleshooting drifting, and inconsistent flight issues.


Originally Posted by vadmcse View Post
FINALLY!!!
I get the AHRS re-flashed.
The last time i tried, installed XAC v1.1, but ONLY to get the firmware files, then uninstalled it, installed the v1.8 and tried the flash.

Yesterday, i did the same as you, uninstalled v1.8 and installed v1.1, with this version, get the AHRS re-flashed.

Made some tests and it is very stable.

THANKS for your help.

Víctor


Author: jetjock51

Let me lay the foundation for a long forum post.

I own a XA 650 V4 with FC-1212-S V1 FW 1.33, AHRS-S FW 1.23, and GPS-S FW 1.01. According to Batfire, these versions work best. My quad would never fly consistently. It seemed to have a mind of it’s own and was a real challenge to fly from flight to flight. I never knew what it would do next. After doing everything Batfire and others suggested, I still had a problem with severe AHRS drift. Batfire suggested reflashing the AHRS even though I had the correct version. I could not find a download location for AHRS v1.23. Finally, Victor found it. I reflashed with it, and the machine flies better than it ever has. I believe my AHRS firmware was corrupt from the factory. I never had a crash until three hours flying time and that one was due to a failed esc. The following is the process for getting and reflashing AHRS 1.23.

Victor nor I could reflash AHRS v1.23 from XA Center 1.8. It is there, but we could not get it to work. We both got an error message and “no joy”.
I think Victor and I together may have stumbled on the right combination to get AHRS v1.23 reflash to work. I deleted XA Center v1.8 and installed XAC v1.1 at Victors suggestion and did find AHRS v1.23 there. I downloaded v 1.23 and did a AHRS reflash from XAC v1.1 and it worked. Victor deleted XAC v1.8, installed XAC v1.1, downloaded AHRS v1.23 for offline update and then deleted XAC v1.1 and reinstalled XA Center v1.8 but could not get the reflash to take. He kept getting the error message as I did when trying to reflash from v1.8. He then reinstalled XAC v1.1 and was able to reflash his AHRS.

As Batfire has stated many times in this forum, the system works when set up properly. XA wiki could be much more transparent about the process for reflashing to the best firmware versions for FC-1212-S v1, AHRS-S v1, and GPS-S v1. It does appear they have abandoned their users by failing to communicate and keep the latest information on the wiki. You can communicate with Jingchen by email and he did respond and offer his suggestions. His email is support @ xaircraft.com.

To summarize, this is the process we found to work for reflashing the FC-1212-S v1 AHRS to v 1.23.
1) Go to http://wiki.xaircraft.com/en-us/Down...ter_v1.1.0.908 and download this version of XA Center
2) Go to your computer and delete XA Center if you have any version higher than this one.
3) Install XA Center v1.1.0.908 from your download location
4) Disconnect AHRS from the Flight Controller
5) Open XA Center
6) Connect USB dongle directly to the AHRS and to your computer with the USB cord
7) Click on Firmware Update button
8) Reflash your AHRS to v1.23 no matter which version you have including v1.23. Remember, this only applies to AHRS-S V1.
9) According to Victor’s findings, you can’t download firmware from XAC v1.01 and install it from XAC v1.8. It didn’t work for him. It worked when I downloaded and installed from XAC v1.1. It worked when Victor again deleted XAC v1.8 and reinstalled XAC v1.1

Victor, does this process agree with your findings?

If you need to reflash the FC or GPS, FC v1.33 and GPS v1.01 can be found there as well. According to Batfire, these are the best firmware versions to work with XA FC-1212-S v1 system.

I have been very frustrated trying to get my quad to fly like Batfire and others have demonstrated it is capable of flying. I have run down every rabbit trail I could sniff out. Nothing completely cured my problem. The first milestone event where I saw a major improvement was when I placed the GPS and compass on a tower outside the canopy and directly above the FC as Batfire’s tutorial shows. Just for good measure, I removed the red lead from the male ESC connectors that plug into the FC except for one (you will notice in the pic that I had a UBEC connected. I have since removed it and reconnected one ESC power lead for simplicity). That one lead is all you need to supply power to the FC and receiver. I am told by guys with high time on MRs that this can prevent possible system problems such as burning an ESC. I lost an ESC and it caused the quad to crash. I am not sure this was the cause, but why does an ESC fail after only three hours flight time?

I added additional shock mounting to the FC mount plate. This was in the form of orange vibration isolating bobbins from http://www.getfpv.com/featured/orange-rc-bobbins.html. These are softer and I think will work better for our light weight equipment than the black rubber ones. These were installed between the frame and the FC mount plate.

The last significant change I made was to twist every wire on the machine. I already had them organized and separated (FC , rec, and FPV equipment). Twisting helps eliminate possible EMI, so that was one more preventative step I could take. Photo attached. (Photo 5) In this photo I have the FPV equipment removed for trouble shooting purposes.

I want to thank Batfire again for all the work and time he put into both publishing his tutorial and for his continued forum support for all the XA owners who are still trying to get their machines flying. I doubt I would have been as persistent and determined to make this thing work without his tutorial. You can find his tutorial here: http://www.rcgroups.com/forums/member.php?u=293281

There have been others who were encouraging without the bashing such as Canadian Geek and Chris Danger Pants. Sorry I can’t recall all the names. I am brain dead from reading all 1361 pages of this thread. You guys all convinced me these machines could be made to fly properly. Thank you all for your encouragement.

I only have four flights since I re-flashed and it was flying far better than it ever had…right up until it flipped and went in. I should have listened to what the quad was telling me. During the first two minutes of the flight I could hear a clicking type noise and it glitched mildly a few times. I landed twice to check it out and all seemed to be okay so I took off again. By this time the noise had stopped. About six minutes into the flight it flipped. I have it on video and can see which engine shut down. It was number four. It flipped and landed on the the opposite side on engine two. After the crash, all engines run fine. More stuff to drive me crazy. I am going to disassemble the motor to search for problems and hope I find something obvious.

I have figured this much out. Flying multirotors requires determination and attention to detail for all systems to work properly. There is a lot going on with these aircraft and small things turn into big problems fast. Everything has to work perfectly or you are going to crash. I fly turbine powered radio control jets for years now. They are complex aircraft and have to be built and setup properly to fly successfully. They are child's play compared to my experiences with this XA 650 quad.
batfire is offline Find More Posts by batfire
Last edited by batfire; Nov 09, 2012 at 02:43 PM.
Reply With Quote
Old Aug 07, 2012, 03:01 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Flight logs and Maintenance log examples

Tutorial 5

HOW TO KEEP FLIGHT & MAINTENANCE LOGS


With the complexities of multicopter operation today it is very important to treat these copters as real test copters. In many ways we are experimenting on the cutting edge of rotor flight capabilities. This is one of the newest areas in RC that only came about just a few short years ago. One of the best ways we can get a grip on our machines and increase the dependability, and longevity of them is to keep flight and maintenance logs. You should treat these copters like real unmanned aircraft (UAV), which the FAA, AMA and Government feel they truly are. Congress along with the FAA is writing legislation and passing new rules on their operation which whether we like it or not will control how and where we fly our multicopters in the future.

If we start today keeping these logs should a mandate come down that if you wish to use these for Civilian, commercial work or any worthwhile endeavor then the first thing you will have to prove is how your copter was maintained. If we start now getting use to logging our flights and maintenance then in the future we are going to be way ahead of the curve and have a better flight track record not to mention saving money in the process.

Below are example snapshots of my multicopter Personal Log (1) on my X-830-H quad copter. This log shows an in-depth explanation of cost and each test, change, adds or deletion I make to that copter. I also keep an X-830 Flight Maintenance Log (2) which at a glance tells me time on props, motors, what modes I used during flight, total time on airframe etc... This information is vital if you are to set up a good Maintenance Schedule for your copter.

Without this type of information you would not know when it was time to change out your props. As an example, if you begin keeping a history of flight maintenance you will find over time when you would have to replace the propellers to avoid hub separations or overall prop failure. You would know when maintenance to the motor bearings would be needed to extend the motor life. There are many things you can add to your log heading to track applicable to your particular copter.

What other things should we consider placing in our Flight Maintenance Logs? How about things such as plug, wire or battery total times/ failures? Anything dealing with Flight operations of the electronics is important. With today’s electronics you are able to keep complete telemetry flight logs on SD for review after the flight. Telemetry logs can be very important to determine if you may have an issue starting to develop that hasn’t become obvious yet. You can also track trends and anything that deviates from the norm is an issue that may be developing. This allows you to investigate before you have an in-flight failure.

Telemetry logs on SD also are good for determining whether you are tracking over ground as you should or do you need further adjustment to your multicopter gps to correct ground track.

I hope this quick tutorial has given you ideas to help with determining ways to maintain and keep your copter in top shape while eliminating common issues that could cause unnecessary crashes and monetary loss later.

These are the fields in my Flight Maintenance Logbook from left to right.

1. DATE
2. TOTAL AIR FRAME TIME
3. PER FLIGHT TIME
4. BATTERY PACK# AND PACK TIME PER FLIGHT
5. PACKS NUMBER OF CYCLES ( MY PACKS ARE LABELED A,B,C, ETC
6. REPLACED PARTS (due to crash, maintenance replacement due to time on part)
7. PROPELLER TIMES (total until replacement of set)
8. INDIVIDUAL PROP TIMES #1,#2,#3,#4 ETC...
9. IN-FLIGHT MODES USED DURING FLIGHT I.E. NORM, ATT, ATT GPS ETC...
10. COMMENTS / NOTES

This is the layout of my spreadsheet log which not only is good for testing but later you can look back and see when to expect to replace parts or props. From this log you can began to make your own maintenance schedule for your multicopters.
batfire is offline Find More Posts by batfire
Last edited by batfire; Nov 06, 2012 at 09:28 AM.
Reply With Quote
Old Aug 18, 2012, 04:45 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Commonly asked questions

TUTORIAL 6


COMMONLY ASKED QUESTIONS & ANSWERS


This tutorial will constantly be updated as more questions are asked.

Gains, Gain Adjusts


What is Gain?

Quite simply gain is a measure of the ability of a circuit to increase the power or amplification of a signal from the input to the output. When dealing with multicopter gain it is usually defined as the mean ratio (0-100) of the signal output of the FC to the signal input of the FC.

What is Stick Gain?

When you go into the XA Center Flight Control screen and adjust the Stick gain for Pitch, you are actually adjusting the amount of sensitivity of the copter to react to your stick control (Tx) input. If you push the stick (pitch) on your Tx forward the copter will react say with 20% gain very sluggish and slowly. Now if you change the stick gain lever to 80% and move your stick forward you will get a very violent quick response pitching forward.

With Stick Gain you should always change the amount of gain for pitch, roll and yaw so that they are the same percentage amount.

What is Attitude Control Gain?

Attitude control gain is set amount of gain necessary to hold the multicopter in a set attitude that you have chosen. Say you wanted to copter to hold a 10’ bank and the copter kept falling out of this bank angle. By increasing the gain you will cause the copter to seek a 10’equlibrium faster and have better control of holding that angle. By decreasing the gain you will cause the copter to fall off the angle and or not be able to hold or maintain the bank angle.

What is Auto-Leveling Gain?

Auto-leveling gain adjusts the amount of force needed to maintain a steady level hover. If the gain is set to low the copter will wonder and fall off in roll or pitch unable to hold a hover. If the level is set to high the copter will snap to level and the motors will fighting change which decrease battery power more quickly.
These are the common gains associated with multicopters and I hope this explains gain usage a little better.


[B]Prop questions[/B]


This one from (Originally Posted by joemi)

As somewhat of a newbie to Quads and putting a fair amount of $$ on mine, my fear of loosing it due to a prop coming loose or breakage grows.

My seemingly dumb question is - should one do anything different when tightening a prop to a quad?

Does anyone use something like blue/medium loctite? Although I've read that loctite near plastic is a bad idea.


Response:

I use 12x3.8 props on my X-830 which are keyed. These are known to break easily and I have had 4 prop breaks over the time I have been flying these copters. I have learned a few lessons on how to extend prop life along with motor life. All of my prop breaks have been at the hub and fortunately I have had no failures in the air.

This is what I have learned. 1st never ever use loctite on crown or motor threads. Why? It will slowly seep down the threads and contaminate the plastic, cf of your prop hubs causing failure in a short amount of time. 2nd Never tighten your crown nuts beyond ½ lb of torque. This is 2 fold. If you torque your crown nuts down any further they will cause the prop hub to bite into the bottom of prop holder causing fractures that will cause hub to fail eventually. Also if you torque down more than 1/2lb then you will start pulling the Jesus clip up against the housing of the bearing. This will cause wear and less flight time and will cause eventual motor or esc failure.

UPDATE: never use loctite, blue or plastic on your copter anywhere near plastic. You can use on screws to lock them on airframe areas as long as they are not coming in contact with a plastic area.

If you wish to feel more secure with your prop nut use a thin nylon washer under to lock into place.

On my X-830’s I use nothing to hold down props. I screw down crown nuts until they are tight and prop is seated then I apply 1/2lb of torque to nut and that is it. Never had one come off or loosen up. I also remove nuts and check props every 3rd flight with magnifier to verify no cracks are starting to develop on prop hub or blades.

Since incorporating these steps into my flight maintenance I have not had a failure of a prop. In the early stages of this maintenance schedule I found one hub cracked before a flight which would have caused an in-flight failure had I not done the checks. When will it fail, who knows but probably within 2 or 3 flights?

Only a couple of things more to mention
Always balance your motors and props which will cut down on harmonic vibration which can cause cavitations in the air and prop blade failure.

Keep a log of flight time on your props this will help determine under normal use when it time to swap the blades out.

If you have prop strikes take them seriously even if they are minor. Remember your blades are doing 6000 rpm and even minor strikes cause blades to loose alignment which will eventually cause a blade failure. I swap them off and replace, keeping the old ones as emergency spares. I mark them so that I can watch them closely if I do use them down the road.

Last thing that breaks blades is weight. Keep your copter within weight specs for the blades and motors. This will help prolong blade and motor life. The heavier the weight the more bending motion you have of blade to hub and this is why you see many of these vids explaining that the blade snapped. Even being within max AOW is no guarantee if you are doing fast descents and max climb outs explained in above statements.


Question:

Is there something special about doing a calibration with an X650 v8 with AHRS and compass (v2)?

I sold a RTF unit to a guy and now that he has received he is getting weird results when he switches it to ATT and ATT + GPS. The manual mode seems fine. He says it is doing a strange yaw and also rising up on it's own. Any ideas would be great.

Thanks,
Dave

Answer:

Hmmm.... I would disconnect gps and compass go back to the start and check the AHRS completely making sure not loose on sticky shock pad, tilted or something like a wire bumping it in flight. Next check for vibration which the AHRS cannot handle by running props up with copter strapped down in bench test. Check all bolts and parts that none are loose or vibrating (usually from crash damage) which may be minor crash but still can cause issues.

After all checks indicate good for AHRS go into XA Center and correctly calibrated AHRS with bubble level. Once satisfied and AHRS looks good in calibration data testing screen Pitch and Roll, do a test flight to confirm no drift and copter hovers correctly. If drift roll or pitch then recheck again including motors to make sure horizontal and aligned with each other’s props no arms slightly bent etc.

Once satisfied then connect compass and do calibration on compass. Go test fly again and there should be no change except no longer a slight yaw right or left.

If all is correct then you can add gps and check in XA Center that it is acquiring satellites correctly.
If all checks out correctly to this point when you add gps and it starts a side tilt the only things it can be are compass calibration is inaccurate or not working correctly and or getting interference from current sources, example wire from compass laying on current wire from esc ….etc..

If after all this when you go to atti gps and it drifts out more than 10 meters then turn off that mode and don’t use until you can isolate issue. See above all over again.

The only thing I haven’t covered here is where is gps located in reference to all other electronics? Is it low and behind the electronics in rear of copter? IF so move it to a high location such as you see with my copters. You may have to spend 10.00 on a gps upgrade kit to get pieces you may need to do this. CnChelicpters.com has the parts kits in stock.

This has covered anything and everything that could be causing this problem that I may have not covered completely in my tutorials except for one thing.

If the AHRS has been in a bad enough crash you may have loosen the ground trace welded shielding inside the unit. That you can’t fix without electronics background and micro miniature soldering kit.

I hope this helps, let me know what happens just remember you can’t skip any steps.

Response:

Thank you so much for your detailed response. After he recalibrated it is working fine.

-Dave

EMI issues

Originally Posted by aeroparts  

Question:

I have a lot of 4" to 6" aluminium foil tape in my work shop, what about using this for EMI shield? anyone used alu, instead of copper foil.
/Lennart

Answer:

Hi aeroparts,

Personally I find copper foil tape much easier to work with and easy to solder if needed for grounding etc... and the adhesive is 100% conductive! Either can be used for shielding. Aluminum foil tape is normally used for EMI applications where Copper foil tape excels in RFI applications. Both can be used or one or the other for multicopters BUT Aluminum duct tape does not have conductive adhesive backing, nor can the pieces of tape be tack soldered together as solder does not adhere to aluminum,(in air) therefore there will not be the required electrical continuity between the individual pieces of tape needed for effective shielding in some EMI applications. My best recommendation is to stick to copper foil tape for multicopter applications since most of the issues with EMI are really in reality RFI issues i.e. Video Tx, Rx, ESC's, and so on......




IMPORTANT PAGES FOR V1 AND V2 OWNERS:

http://wiki.xaircraft.com/en-us/Products

http://wiki.xaircraft.com/en-us/Down...ircraft_Center

http://wiki.xaircraft.com/en-us/Special:AllPages
batfire is offline Find More Posts by batfire
Last edited by batfire; Nov 24, 2012 at 04:50 PM.
Reply With Quote
Old Aug 25, 2012, 10:31 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Understanding Brushless Motor Lingo

TUTORIAL 7

Kv, Motor Designations, Kv vs. Torque and Kv vs. Propeller Choice & "C" RATINGS



Quite a few people have asked which motor is best for them. That is a loaded question. It is much easier to explain what you are looking at and let you make the choice. I will endeavor to explain Kv, motor designations, KV vs. torque and KV vs. propeller choice. With this knowledge you will never have to ask again what type of motor and Kv should I get.

How does a brushless motor work?


On the most basic level, an electric motor’s only job in life is to convert electrical energy (battery power) into a mechanical energy (revolution of propeller). There are two basic facts which allow an electric motor to work:

1. Electric and Magnetic Fields of Flux and they are related – moving protons (charge) produces a magnetic field, and magnetic fields can produce electric charge.

2. Magnets Interact – Magnets align when placed near each other. Brushless motors consist of two magnets. One is a permanent magnet, the other is a wrapped coil of wire which when charged becomes a magnet.


So how does a brushless motor turn and how does the ESC fit in?


The brushless motor is designed so that each magnetic field is out of alignment with the other. So when battery power is applied this miss alignment allows the motor axle to rotate trying to catch the other magnetic field. This is done by turning on different coils of wire that surround the motor axle in succession. Now can you guess what controls this never ending chase? If you said the ESC (electronic speed control) you would be correct. The ESC switches the coils on and off rapidly in a synchronized sequence which is based on the motor axle position. The ESC should always have a 20% greater amp Capacity compared the maximum continuous amps the motor generates.

Common terms and their meanings

There are a number of terms that are thrown around in the multicopter world associated with brushless motors. Here are some of the most common terms and explanations:

RPM – the measurement of angular speed, how fast something is rotating. Rpm is how fast the brushless motor will rotate.

Kv Rating - The relationship between Vt, Kt, and rpm are linear for a brushless motor. That makes the math easy. It turns out that the number of RPM’s provided by each Vt is the same and the Kv number. The Kv number is useful mathematically so you can calculate how many Vts you need to reach a certain RPM, or vice versa.

Continuous Rating / Burst Current Rating – Continuous current rating is how much current the motor can handle continuously for extended periods of time. The Burst current rating is how much current a motor can handle in a short burst of a few seconds.

Outrunner Brushless Motor – this is a motor has a permanent magnet that is located on the inside shell of the motor placed outside the stationary coils of wire on the motor shaft.

Torque – is a measurement of angular force, or the amount of push a rotating shaft has.

Watt – is the measure of power, or how fast the energy is used.

Volt (Vt) – Measurement of electric potential or how much push the electrons from a battery has. That said, the greater the voltage is the more energy is being applied to a given amount of charge.


Watts = Volts x Amps
This is the most important formula you will ever encounter in multicopter power.



Kv ratings a further explanation

When you look at a motor the first thing that seems to pop out is the KV rating. What this means is how many revolutions (RPM) the motor will do per volt in a no load (no prop) condition. Let’s say you had a motor that had an 850Kv rating and like many of us you use a 3 cell 11.1vt battery to power you copter. If we take that motor with no prop installed on a test platform and you powered it up and went to max power you would see 850(Kv) x 11.1(Vt) = 9436 (RPM) from the shaft. So using that logic if the motor was rated as 3-4S then if we connect 14.8vt battery we would see theoretically around 12580 rpm. That’s a 25% increase in rpm for one extra cell. Using a 2S would yield 6290 rpm. Which is best for your application? It all depends on what you want to do with your copter.

For say an 11.1vt battery, a lower Kv allows you to use a larger propeller and will give you more overall thrust but at a loss of speed. That speed may not be important if you are just videoing or photographing real-estate. If you feel the need for speed then you would go with a higher Kv rating giving up thrust (carrying ability) and flying a smaller prop with higher rpms gain aerobatic agility and higher top end speed.

To summarize with our 11.1vt example battery, if you want speed and aerobatic performance go high Kv’s and smaller props. If you want to climb to the moon and hover effortless with larger load capacity go low Kv’s and larger props.

Ok, here comes that question I know you want to zing me with. Why does a 2s pack swing a bigger prop than a 3s pack? My 2s pack is swinging an 11x4.7 but my 3s is swinging a 10x4.5 that seems to contradict the above statements.

Well not really, that 2s swinging that large 11” prop will have a hard time reaching the motor amp limit, but if we tried to swing the same prop on the 3s pack you will exceed the motor limit and poof. So that means normally we run smaller props on 3s and they spin faster generating a higher rpm. (Kv x V)

Motor Designations explained

Ok, this really seems to trip up people when they are scouting for a new copter build. Let’s walk through and example:

We have two motors which look almost identical; one is a MT2215-25 with a Kv of 1000 and the other a MT2215-20 with a Kv of 1200.

Taking a closer look we see the 2215-25 Kv 1000 uses 25 turns of 22 AWG wire. The other motor 2215-20 Kv1200 only take 20 turns of wire on it coil. Since there are fewer turns it is wound with 21 AWG wire.

Since we see the higher Kv motor is wound with a slightly larger size wire it can handle more current, and if both motors take the same 11.1 vt battery the motor that can take the higher current will have the higher max power output. This is why the 2215-20 Kv1200 can take 230 watts of power, while the 2215-25 Kv 1000 can only handle 180 watts of power.

I need more Torque

Torque is a motor constant (Kt) and inversely proportional to the voltage constant (Kv). That means that as Kv increases, Kt decreases. Huh? It means the 2215-20 Kv1200 motor has a lower Kt and draws MORE current on fewer volts to make its power. The 2215-25 Kv 1000 motor will get you to the same point with MORE cells but less current.

So which motor do we go with? The 2215-20 Kv1200 will be great for running small props (6, 7, 8 inch)more agility and very fast for speed. The 2215-25 Kv 1000 will give you a little more torque, loading capability and ability to run larger props generating extra thrust at the expense of speed.

In simplified terms, prop diameter determines thrust, pitch is airspeed, so think of pitch like a car’s gears. Low pitch, low gear, high pitch, high gear.

Remember, bigger the prop, more thrust, the higher the efficiency.

Here is a link to one of the most helpful calculators for determining your multicopter motor/prop combinations:

http://www.ecalc.ch/motorcalc_e.htm?ecalc



Additional information on motor,Kv and prop selection may be found below:

APC Props Efficiency - Official Data
________________________________________
Thrust and power needed according to RPM.
Values given by APC being not easily usable, I have made them more readable for common props sizes.
Efficiency shown is about propellers, global efficiency is slightly lower.

What could be expected from 150W, 250W or 350W motors ? Which KV to choose ?

Source : APC Propeller Performance Database

http://www.apcprop.com/v/downloads/P...B/datalist.asp


This information courtesy of “Mortimer”

This summary table is made on purpose with maker's values.
It's easily understandable and may help choosing motors and props.

I've also done some real tests:
Nov 28, 2011 - Five motors compared: http://www.rcgroups.com/forums/showp...0&postcount=50
Dec 17, 2011 - Five propellers tested: http://www.rcgroups.com/forums/showthread.php?t=1557506
Feb 15, 2012 - Efficiency test: DT700 and 24 propellers: http://www.rcgroups.com/forums/showthread.php?t=1594904
Jun 30, 2012 - 2 motors and 10 propellers: http://www.rcgroups.com/forums/showp...&postcount=159
But they're less understandable and now forgotten.




“C” RATINGS AND DISCHARGE RATES

So what is “C” Rating and what does it mean to me? C rating is simply how fast a battery can be discharged safely. If we have a lipo battery with a 20C label then this means you can safely discharge the battery at 20 times more than the capacity of the battery, a 25C battery is 25 times, and a 40C is 40 times the capacity and so on.

Let’s use the 1000mah battery rated at 10C as an example. So with a 10C battery we can pull a maximum sustained load of upto 10,000 milliamps or 10 amps. From a time in air standpoint this means the battery will be exhausted in about 6 minutes or so. We determine this by first calculating the mah per minute for the battery. 1000 / 60 = 16.6mah per minute. Now we take the C rating of 10 x 16.6mah of drain per minute and / that into the battery packs capacity of 1000mah which = 6.02 minutes.

So if we take a cheap battery pick of 10000mah with C rating of 10C and work out the math we see the following. 10,000mah or 100amps. So now we take the mah per minute of the battery, 10,000mah / 60 minutes = 166.6 mah per minute. This means at 100amps of load it will be exhausted in 16 minutes. But at what cost?

My personal guide line is never buy a pack with less than twice the C rating. With a larger C rating the battery will run cooler than maxing it out each flight pushing the battery hard and close to its limits which will burn out a battery and reduce its useful capacity in short order. As a general rule your C rating should always be at least double the maximum rate you intend to pull out of it.

It comes down to do I pay now or pay a lot more later? Those great battery deals on the internet are not always the best deals. I would rather pay now for a better battery than to have my copter fall out of the air because of a cheap battery failure IMHO.

Here is a link to one of the most helpful calculators for determining your multicopters motor/battery combination:

http://multicopter.forestblue.nl/lip...alculator.html


Well I hope that this little motor tutorial has given you a better understanding when shopping motors on your next build.
batfire is offline Find More Posts by batfire
Last edited by batfire; Feb 20, 2013 at 06:54 PM.
Reply With Quote
Old Sep 03, 2012, 11:30 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
How to save weight and have a nice 2 axis tilt & roll camera mount

TUTORIAL 8 KINDA.........

HOW TO SAVE WEIGHT WHEN EVERY GR. COUNTS & STILL HAVE A GREAT 2 AXIS CAMERA MOUNT


I found a nice 2 axis camera mount that you can install on your X830 with the standard stock skid with very little effort. The fact that it is a 2 axis mount giving you tilt and roll controlled from your FC is not the best part. Nope, it’s the quality and price of this fine mount. At $69.00 this mount blows away the competition. You will need to grab a couple of servos which are sold separately (14.95ea.) but Quadframe has the servos designed for the mount. The nice thing is you don’t need a fancy landing gear to use it on the X-830. It so happens the X-830 landing skid bolt pattern is the same as the mounts tubing rail pattern. This makes it an easy install with a total weight addition of ONLY 95 grams. Below are some pictures of the install and I will go through steps needed to do this fast easy install.

Picture 1 shows the 2 axis camera mount with the tube holder locations at top of mount.
We will need to remove the landing skids from the bolts holding them.

Picture 2 shows replacement screw which are ¾” long with locking nuts.(one in picture is longer than ¾”in test mount)
Now unbolt and remove the CF gear plates shown which also hold the esc’s in place. The esc’s are also Velcro in place so they won’t fall out. Remove the front 2 skid bolts which we will be replacing with same diameter and thread type bolt but ¾” long. When you reinstall the new bolts make sure you use the old flat nuts which lock it to the CF plate.

Picture 3 shows the tubing that camera mount rides on.
Next we need to cut 2 pieces of 6mm tubing to make the tube rails. Cut each tube rail 1¾” long. At the center of the tubes (0.875") drill a hole through slightly larger than the skid bolt diameter. You may wish to use a wooden dowel inside the tubing for extra strength not allowing tube to deform as you tighten the skid locknut down. Note: If you have access to 6mm aluminum rod stock you may also use this instead of the tubing.

Update: I installed replacement set of 6mm tubes that are 3" long with hole drilled at 11/2". This alteration is for those running larger batteries and FPV systems. You can now slide the assembly forward to rebalance quad. You can use tiny tie wraps to lock in place.

Picture 4 shows the assembly diagram.
This diagram shows you the assemblies of all parts for this tube rail build. Once you have completed all parts and slid the tubes into place on the rails, install the landing skid then the tube rails on camera mount onto the front 2 bolts tightening down with the lock-nuts.

Picture 5 shows camera mount ready to shoot with fpv camera also installed.
Now all you have to do is install the servo connectors on the 9(tilt) slot output side and 10 (roll) slot output side of FC.

On the V2 XA Flight Controller you only need to install the tilt control from receiver to input side of FC. Once you calibrate these in XA center you are done. The controller will control the gimbal movements and you can manually override control of the tilt as needed.

This is a video showing the installed 2 axis camera mount with GoPro Hero2 in action. I made sure this was a calm no wind day so that I could also demonstrate what it looks like to be caught in your quads vortex ring. You will see me a few times drop into the vortex ring during descent and some descents were violently kicking me around so I could see how the mount and FC gimbal control reacted. Overall since gains were low it did well. I also did a couple of takeoffs with 45 degree tips so you could see not only the FC right the ship but how that camera mount and gimbal reacted to the lunge.

9 11 2012 FC-1212-P WITH 2 AXIS QUADFRAME.US CM (11 min 28 sec)


I hope this has given an extra avenue you can use when considering camera mounts for your X-830. Camera Mount found here: http://quadframe.myshopify.com/colle...-black-version Make sure you get the one for the landing legs version mount.
batfire is offline Find More Posts by batfire
Last edited by batfire; Sep 24, 2012 at 07:00 AM.
Reply With Quote
Old Oct 19, 2012, 06:17 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
COMMON ISSUES AND SOLUTIONS FOR 2.4GHZ Rx AND GPS INTERFERENCE

TUTORIAL 9



COMMON ISSUES AND SOLUTIONS FOR 2.4GHZ Rx AND GPS INTERFERENCE


One of my favorite subjects of lately is EMI/EMF/RFI issues that interfere with on board Rx and GPS performance in UAV’s. Some of the most popular A/V transmitters (900 MHz, 1.2 GHz, and 5.8 GHz) can cause significant interference to onboard electronic systems. The common symptoms associated with use of this A/V are poor GPS reception or lose of GPS lock entirely, and loss of RC receiver range.

Many of you have read my tutorials on the XA Flight Control systems and the issues with interference I had with them along with solutions to correct the issues. This tutorial goes one step beyond those solutions.

Correcting interference problems can be a very frustrating affair. If we use a thoughtful logical approach to this issue we can substantially reduce and in some cases eliminate the interference completely. The success of removing the interference will be judged in small changes to our equipment which will bring us one step closer to ridding the EMI /RFI problem. Please follow each step in the process, failure to or skipping over a step will most likely leave you without a workable solution.

As mentioned above most common A/V Transmitters cause interference from radiated (airborne) or conducted (wire path) signal radiation. This means you are not just trying to correct one issue but two issues when you attack this problem.

Radiated Noise and effects

First let’s look at (Radiated Noise) the distance from your FPV Tx Antenna to your GPS antenna. Simply, the further you can place these antennas apart the better off you will be. Forget looks and compact neat installation, you need to get the A/V antenna as far away from the GPS antenna as possible. Also as with most GPS antennas they are X distance above your COG so place you’re A/V antenna as low below your COG as possible. If we sit and go over the formulas (boring) we will find on most 900 MHz- 1.3 GHz systems we need to be at least 16”-18” apart and this is for low power A/V Transmitters. The higher the power the further the spacing needs to be. Let us not forget harmonic action also. Just because a GPS is tuned to 1.6 MHz or so and you’re A/V Transmitter is emitting 900 MHz does not mean harmonics multiples do not occur that will cross your GPS frequency range. You will be looking at multiples at 1.8 MHz, 2.7 MHz and so on, looking at a 1.3 GHz system you will have harmonics even closer to the GPS frequency. What makes this matter of harmonics worse is that most RF synthesizing hardware inside the Transmitter begin at low frequencies and intensify as harmonics increase. We should not forget that the A/V frequency is not stagnant. When audio and video signals are sent they vary which causes the frequency to vary which also causes the behavior of the RF to change. Another factor which can affect harmonics intensity is the type of A/V antenna you are using which can make matters even worse.

So what can we do to reduce this harmonic interference? We can use low pass antenna filters specially designed for the frequency range we are operating. This type of filter can help suppress the noise in the GPS RF band. But this solution alone will not stop the issue of GPS interference.

Conducted Noise and its effects

It would be great if all the interference, noise, was generated out the antenna but unfortunately this is not the case. We have to look at, (Conducted Noise), removing the EMI/RFI noise too. This type of noise is conducted down the copper wire cable from you’re A/V Transmitter to your camera, power supply battery, onboard microphone then out to your electronics such as OSD boards, common power, and FC system. This noise would also travel into the GPS receiver either directly through the host cable or indirectly from other cables.

So how do we correct this? The most common cure is mounting common mode filters known as Toroid Ferrite cores onto the wires and cables. These little cores are designed to attenuate the high frequency noise that travels along the cables by wrapping your cable through them a number of times.

Not all Toriods are created equally. Do not use the split type toroids which just snap over the wires. They are all but useless for EMI/RFI frequencies that need to be attenuated. Always use the one piece Toroids. When you wrap a core on the transmitter module do it within an inch or so from the module. Remember that ALL the transmitter cable wires must be wrapped through the Toroid, this includes the microphone wire. Next rule always use a minimum of 10 wraps around the Video cable to achieve good noise reduction. If you have to achieve this number of wraps splice in wire but make sure you do the minimum wraps. There is nothing wrong with making premade wraps and cutting and splicing on to the premade ends and don’t forget the shrink wrap. After you complete the video wrap move on to the microphone and camera wraps. Don’t forget to wrap your GPS receiver cable.

Last but not least is to consider the (Dirty Power) noise issues. Let me explain dirty power. Lets take your copters motors which are high current devices and while operating they cause large current spikes (changes) on the battery circuit. These currents take the form of transient or ripple voltages and you have seen this before if you have video onboard your copter. It takes the form of waves, jello which move in concert with your motor throttle changes.

The simplest solution to correct this issue is a dedicated battery for you’re A/V gear, but if you have to share your power with the main battery then you will need to filter the power supply. To correct this issue it requires a LC power filter. These you can find by doing a Google search for L-C power supply filters. The easiest solution is just to get a dedicated A/V battery.

I hope this tutorial has helped you gain a better understanding of how EMI/EMF/RFI issues can be corrected with a little work.
batfire is offline Find More Posts by batfire
Last edited by batfire; Oct 25, 2012 at 04:41 PM.
Reply With Quote
Old Nov 28, 2012, 09:53 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Multicopter PID / Gain Flight Control Theory

TUTORIAL 10


Multicopter PID / Gain Flight Control Theory
by BATFIRE

Part 1

Feedback Loops


Introduction:


I’ve been questioned numerous times on the subject “how to adjust multicopter PIDs and Gains”. After considering these questions I have come to the conclusion I should concentrate more on explaining generically how these adjustments affect the flight control of a multicopter instead of trying to explain heavy theory. I believe this set of tutorials will give you a better basic non-technical understanding of multicopter flight dynamics allowing you to actually comprehend adjustments you make and how they will affect your overall multicopters stability. I will not be covering complex “Kalman theory” because for basic tuning I do not feel this is necessary. I wish to make this series of tutorials simple enough not to lose your attention and easy enough that anyone should be able to comfortably do modest PID/ Gain adjustments with confidence.


Note: In this tutorial series to keep things simplified we will not be using complex math to explain how to do adjustments, but will show how easy it is to break down these complex formulas. (NO Complex Engineering Math is Necessary to do Simple PID Tuning and Gain adjustments) I will only speak of 3 dimensional spaces in Part 1 and 4 dimensional spaces in Part 2. In Part 3 of this tutorial series I will cover actual set point adjustments and how to calculate and graph them. You will have a basic understanding of how your copter will fly before you even leave the ground. I will also show you how to break down those long engineering formulas into simple math for adjusting set points. So let’s get started.

Flight Controller and Thrust

Multicopters completely depend on (torque) thrust to sustain flight. Though mechanically much less complex than helicopters to hover and move in controlled flight, the on-board system to control these motors is much more complex. The flight controller is the heart of the multicopter in which resides a complex firmware to control and balance power to each esc and motor/ propeller combination.

The average flight controller has multi-sources of input (sensors) so it can determine forces needed to maintain its location in 3 & 4 dimensional spaces. Sources that can feed information (fusion data) to the flight controller are - IMU (which give acceleration and gyroscopic information in the x, y, z axis), magnetometer (direction readouts), Barometer ( altitude reading), optical flow sensor (which sends info for horizontal altitude hold or can send information for collision or obstacle avoidance), GPS (global positioning system), ultrasonic range finder (which send information on the sonic range to an object in its detection zone) and feedback (we will speak of this later).

With this fusion data you can control hovering flight or movement throughout the flight controllers 4 dimensional cubic world. This data allows the flight controller to determine which motors to speed up and which to slow down to achieve the desired direction of flight along with height and whether to yaw left or right from a specified course.

Say What?

Once again, thrust is (torque force) and with multicopters you have CW torque forces and CCW torque forces. To yaw left or right requires changing the rotational torque of all motors so there is a total rotational torque in desired direction you wish to turn. Once you have yawed to the desired direction and heading you counter the rotational torque to stop the copter. To move forward or backwards, left or right you are adjusting the amount of torque force in the desired direction of travel. In other words in forward flight the rear motors is producing more torque and RPM than the front motors. The reverse is true if you are backing up. When traveling side to side (roll) the motors opposite of the roll produce the most torque. I know this sound easy but sometimes these concepts are forgotten when determining what PID or Gain setting you wish to input. Hopefully these explanations will help clear up some of the confusion.

Why looping?

Ok, now that you have a basic handle on how a copter flies dynamically let’s look at why a copter flies using an on-board flight control system. Remember that fusion data we spoke of earlier? Let’s looks at how that data affects decisions made by the flight controller. I will briefly explain what is known as a simplified basis feedback loop. This, in effect, is what you are changing when you adjust the PID or GAIN settings.

Standard Simplified Block Loop

The following simplified block loop diagram is found in almost any appliance, navigational aid, central air and heat controller etc…anything that requires a feedback loop. We will use this diagram to introduce you to a simple PID loop from a block prospective and where it is located in the scheme of things along with signal flow. Fig.1

If we overlay our flight control system on this diagram then an input (any Tx stick movement) is sent to your Rx. This data is then sent to the input side of your flight control unit. The stick movement information goes to the PID controller (PIDC) embedded inside the FC unit. After completing its move through the P, I, D (PIDC) section the corrected signal (Tx) now enters the PROCESSOR (P) which takes the signal along with any DISTURBANCES (D) input such as a cross wind component and outputs a process variable (PV). This PV continues to the Controller Output (CO) and on to your esc’s and controls the output of your motors.

Wait, now we have to complete the loop. A real-time sample of the PV now (CO) is not only sent to the esc’s but back to the sensors which then send the combined info plus new real-time Sensor data (S) back to the PIDC. This information plus the new incoming stick information continues back around the loop again and out the PV to the esc.

Did you notice anything about this feedback loop system that might explain how the FC knows where it is at all times? That’s right; this type of setup only uses one sampling of real-time information to continue the looping action each time around the loop. Who cares about information that is 5 minutes old or an hour old? This is how you are able to speed the process up and allow the FC to create a more accurate faster response to its environment.

A More Accurate Simplified PID Loop Made for Multicopters


Ok, so we made it around an everyday standard simple loop system to see what each step accomplishes. Now let’s look at a basic PID loop that we use in multicopters flight & tuning. See fig.2 This loop is an example of a loop you see located in your Attitude Control Tab in the Profile Editor section of XA Center. This loop is for the Normal Mode Pitch Axis. You will see how simple this is to understand.

The human input (Tx) enters from right. The feedback loop sends back the combined information real-time from last loop run (PV) which we subtract from our new human input. This new error difference (ED) is multiplied to each of our current P, I and D settings (set points). The new corrected error (CE) value of each is added (Tx + SPV) and continues to the Processor section of the FC card as we see in figure 1. When completed we have again a real-time value to use on our next input cycle.

This type of loop computations allow us a very fast responsive FC system able to resolve effects of wind, gust, IMU attitude changes and human input faster than a closed loop system that saves the memory of many previous loops. This is why companies strive for a faster loop rate with military types passing 1k+ speeds easily. This is also why the faster the FC computations the more you will pay. Basic to average FC systems work in the 50-500Hz range with some college experiments now reaching 1 KHz speeds.

Gain Adjustments are Simple Now


Gain adjustments are really a simple way of adjusting the multicopter without having to have knowledge of PID tuning. With the V1 and V2 XA flight controllers you would just go to XA Center and click on the Flight Control tab. Here as with many other Companies flight controller system centers are tabs for each area of flight control adjustments.

The Stick Gain Adjustment is designed so you can either have more aggressive control of your copter for aerobatic maneuvering or to tone down the stick gains for relaxed easy AP work.

The Auto-Leveling Gain function allows you to adjust pitch and roll gains to determine the amount of power needed to right itself and auto-level.

With the Attitude Control Gains you are setting the symmetrical gains of the copter for the % of gain needed to hold the attitude of the multicopter. If not enough % gain the copter will drift from auto-level attitude. If the gain % is set to high the copter will actual overshoot the auto-level center attitude and fight back and forth continuing to overshoot center.

Each of these conditions determines not only how stable a hover, gps lock, altitude or attitude control, your copter will maintain but also but how much valuable power will be needed to maintain this stability. The better your gain, PID tuning adjustments the less power you will consume and the longer your flight times will be.

I hope this tutorial has brought forth a better generic understanding of how basic stick movements travel through the PID loop of your FC to affect your esc's and motors operations. Part 2 of this tutorial will concentrate on closer look at P, PI, and PID loops.
batfire is offline Find More Posts by batfire
Last edited by batfire; Dec 05, 2012 at 05:32 PM.
Reply With Quote
Old Nov 29, 2012, 06:32 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Multicopter PID / Gain Flight Control Theory Part 2

TUTORIAL 10


Part 2

A Closer Look at P, PI and PID Looping



Well, now that we have walked through the overall operation of the FC (fig 1.) and a closer look at the PID Controller (PIDC) let’s take a deeper look at the PIDC and flow diagrams. Refer back to our block diagram as an overall reference (fig 2.) while we take a more in depth look at each component and term in this figure starting with the (P) Proportional block.

HINT: Before you change any values from the factory installed values always write the value down. You should keep a log on all your value changes in-case you have to go back and change them in the future. You will also be able to see each value you installed and what the flight results were for that value.

Proportional Block


Refer to (fig 3) below also as we explain P’s operation. We could operate with proportional control only taking the incoming (Tx) and simply multiplying this by the error (PV) processor variable and get the controller output corrected error (Tx x PV=CE). But how well would the copter fly?

Our current established value (P), (in this case 1.300) counters external effects as we discussed earlier, i.e. wind, gust along with smoothing human (Tx) inputs. To be able to do this it must receive the feedback gyro value input of (PV). When we make adjustments to (P) we are striving for a flyable stabilization of the copter. If we find the copter is oscillating back and forth at a high frequency rate then the (P) value has been set to high. If the copter feels very sluggish and slow to respond to leveling commands then the P value is set to low. Once we have tuned this value in just below the oscillating or wobble effect then we can incorporate the (I) block into our controller.

Integral (I) Block

The simplest explanation for (Integral) is "the signal sum of the instantaneous values plotted of the signal from the time you started counting until the time you stopped". So, each time we send an input through our PID circuit a portion goes back through our feedback loop and is measured against the last input sent.

If we go from a static value to time step values we can see a graphic trend starting to form. See (fig 3b). This trend starts out from 0 before we received our first feedback loop signal and climbs to maximum on the second or third sampling. From this point on if we have put in a small correction error (I value 0.030) then we will see the trend oscillation line over the next few feedbacks stabilize at a constant level. I like to call this the smoothing value. Next we will combine (P) and (I) for finer tuning.

Combined Proportional and Integral Blocks


Refer to (fig 4) below as we explain I’s operation and then how P and I work together to form a better solution.

This is beginning to look like a real Flight Controller now. When we combine Proportional and Integral actions together we get the PI controller and much better real-time calculations to stabilize our copter. In flight Controllers adjusting the proportional action is known as “seconds per repeat”. This means the value we have from the Integral box (0.030) is how long it takes for the Integral to match the Proportional action box (1.300). Please note that the larger the number you use for the Integral box the quicker it will match P action. So with that said, the smaller the P value the bigger the integral action!

Predicting the Future, the Complete PID Tuner


Now we will look at the (D) Derivative Action of the PID controller which will allow performance tweaking. If you noticed in (fig 2) we had the following settings (P- 1.300), (I-0.030) and (D-1.300) set into our PID controller. In this case you will notice our (D) is the same size action as our (P) setting. Derivative Action settings allow you to have larger P and I gains and still maintain a stable loop which gives you a faster response time and better loop performance. How? (D) acts as the crystal ball and looks into the future to predict future changes and projects the current rate of change into the future to compensate. This number is not the current measured value (rate of change) but a future measured value. So how far into the future does the crystal ball see? That depends on what you set in as the (D) action. If I have (P) action working in a 1.3 second time frame then I want to be able to see 1.3 seconds into the future. This system works great to smooth and tweak in your copter with the exception of one issue.

Noise Again?

Remember my pass tutorials dealing with eliminating noise as much as we can in our copters i.e. RFI and EMI? Well, depending on the amount of noise you have not eliminated will depend on how much it effects the future predictions. Those noise spikes can confuse the PID controller and specifically the derivative action future algorithms. This is why we strive to eliminate noise from our copter systems not only for video and video range along with copter Rx range but also for the FC to operate to maximum expectations. The clearer the crystal ball the better the future looks.

I hope you have enjoyed Part 2 of the tutorial for non-technical PID tuning and Gain setting. You should now have enough basic knowledge to understand Gain settings and how a PID Tuning loop works.

In Part 3 we will discuss Practical Application of P, PI and PID Tuning and how to perform small adjustments to tweak in your copter. We will be using some math but believe me we will keep the math simple. We will decipher those huge scary looking formulas and show you how to simplify them.

When you get through with Part 3 you will be able to manually calculate set points and how to create an Amp vs. Time graph to prove your set points before you ever lift off. You will conquer those mind boggling engineering formulas by substituting a simplified math in their place. You might even be ready after this tutorial to tackle the dreaded Kalman Filter.
batfire is offline Find More Posts by batfire
Last edited by batfire; Dec 06, 2012 at 10:20 PM.
Reply With Quote
Old Dec 06, 2012, 05:58 PM
Connecticut Quad Hacker
batfire's Avatar
Joined Dec 2009
956 Posts
Multicopters PID / Gain Flight Control Theory


TUTORIAL 10


Part 3


Practical Application of P, PI, PID Algorithms on Multicopters



Acknowledgement:


I will be using graphic representations from Xaircraft Wiki for this tutorial to assist Xaircraft multicopters owners in using the Xaircraft Center for PID tuning. I thank Xaircraft for use of these displays.

Even though I am using Xaircraft Center PID tuning examples, all FC systems on the market use a similar setup to PID tune their respective flight control systems. I will stay as generic as possible so that all may benefit from this tutorial.


Terms and Definitions;



Before we begin the PID Tuning there are a few keywords we will be using so I will note them here along with their definitions.


Quadcopter (QC) – the unit that will reflect the outcome of your PID settings.

Pitch, Roll or Yaw Output – what axis we will use for the quadcopter output adjustments.

Output Rate (OR) – time it takes for input to output change. Example; Time it takes from pitch down command: (Input), until quadcopter pitches down (Output).

Error Correction (CE) – This is the combination of your (set point) and the multiplied error difference.

Controller Output (CO) – this is the final corrected Output to the quadcopter esc/motors and to the processor feedback loop.

Control Action (Tx) – This is the manual human input you give the quadcopter (stick action) to create a change, such as copter in pitched forward flight and you give an input command to pitch back to level flight.

Processor Variable (PV) – This is your combined sensor and one real-time processed data feedback input to PID controller.

Error Difference (ED) – The difference between the Control Action and Processor Variable inputted to the PID Controller.



Proportional Controller Tuning Only



As we discussed previously in Part 2 the P tuning loop must be thought of as 3 individual loops. When we control a quadcopter we must control not only the Pitch but the Roll and Yaw of the copter. So that requires 3 P tuning loops with three setpoint values, one P setpoint value for each axis of movement.

The P tuning loop has disadvantages and is very seldom used alone for multicopters operation so I will only briefly mention the use of (P) alone. Just as a quick refresher following Fig 3 the input (control action Tx) enters from left and your sensor data (PV) is subtracted from your stick input. That difference (ED) is sent to the Proportional unit where it is multiplied with the setpoint you have installed. The completed signal (CE) continues to your controller output (CO) and on to your esc’s. Remember that the (P) tuner in our example is an open loop so the (PV) incoming is just real-time sensor data only.

Disadvantages: Since this is Proportional only mode, the controller simply multiplies the error (ED) by the proportional gain (setpoint) to get (CE) which leaves the Controller Output and (CO). Later we will break down the engineering formula and the proportional gain (setpoint) will be known as (Kp).

So, small changes (Kp) is the safest way to get to a setpoint, but your controller performance will be slow. If the Kp is increase to fast, overshoot in the signal will be present and faster oscillations will occur in flight.

To have any hope of holding a steady hover and stable forward flight you need a flight controller that has the ability to measure and feedback a portion of the (CO), determine the amount of (ED) and apply a (CE). This is known as a feedback loop. In XA Center Configuration Control Tab Fig. 5 shows where the P Tuning for Hover Modes three axes are located.


Boundary Setpoint


One last function in Fig 5 I will mention because I am sure you are wondering. What does the boundary setpoint mean? A simple definition would be: This is an electronic clipping of the signal to limit the amount of strength in the movement of your input vs. final output. Example: I push the stick forward and as the signal goes through the FC system it will look at the boundary setpoint to determine where to cutoff the amount of signal to keep the output response pitch from being a violent pitch over. This would be the same for Roll or Yaw. Boundary is used in all P, I and D functions for the V1 and V2 FC systems and is actually a (TIME) function.

I have not noticed this function available in some of the other FC systems on the market. I feel this is a great tool for closely tuning the copter for perfect stability. It works almost identical to PID tuning but limits the amount of strength your inputs will have on the output in time, i.e. fine tuning stability.



(PI) Proportional-Integral Controller Tuning





Refer to Fig.7 to follow along as we traverse the PI Tuning loop. I will endeavor to explain the operation of the loop but we will also use simple math to understand how this loop returns the copter to level semi-stabilized flight. When we complete this loop you should understand how to use simple calculations to determine set points to be used in the PI loop.

The PI tuning loop differs from the P tuning because we’re no longer looking at an open system and have added a couple of components to the mix. The PI Tuner comprises of a Proportional (outside loop setpoint) tuner and an Integral (inside loop setpoint) tuner loop. See Fig 7B. The combined sums are then split. One signal leaves on the controller output and applied to the esc’s and motors. The second signal is feedback to the processor to be combined with new sensor data for the next cycle. This combined signal (PV) is applied to your next (Tx) stick command.


Would you like to see the engineering mathematics of that? See refer to text (RTX) 1. (below in picture section).

Ok, before we get too excited here let us refer to Fig 8. If you take a close look at figure 8 you will see that I have installed our simple designations and math onto the formula. Now doesn’t that look a little easier to understand? Nothing has changed from our previous explanations in Part 1 and 2. So let’s break down the above formula.

U – Controller Output (in our case (CO))

t – Time - instantaneous

MV – manipulated variable (CO) returning to the Processor going out as (PV)

e– Error SP- PV this in our diagram is (Tx-PV) which = (ED)

Kp – Proportional gain, tuning parameter (setpoint) (P)

Ki – Integral gain, tuning parameter (setpoint) (I)

Kd – Derivative gain, tuning parameter (setpoint) (D)

T – Variable of integration; (simple it takes on the time from time 0 to the present (t))

0 – start of time in a sequence

So by looking at Fig 8 we can see that the only addition to our static walk through was the time required for the Integral to complete its looping and smooth the future Control Output Processed into PV back to subtract off our new Transmitter input.

Let’s recap what we have learned so far. The PI controller is really a special case scenario for our quadcopter because no derivative error function is used. In the PI we use the (Ki) integral gain as a trade off between decreasing overshoot and increasing settling time. With the lack of using (Kd) in the mix the PI controller system alone is less responsive to changes in Tx movement, slower to reach the programmed setpoint and slower to respond to perturbations compared to a well tuned PID system. This means you could overshoot the setpoint and oscillate also. See Fig 5b to see where these setpoints are located in the Xaircraft Configuration Hover Mode Control tab. Let's move on to the complete PID tuner.



(PID) Proportional-Integral –Derivative Controller Tuning



See (RTX2) below:

You will notice our formula has changed slightly. The Derivative function is now added to the formula. I bet you had no problem figuring that out since we have now added the (D) loop. But from reading my tutorials you can see this is no longer a scary heart stopping formula. But then again, this is just a static formula what if we add velocity to this equation?

(RTX 3) Below:

Now you can see why so many non-engineering types decide maybe this is a little more time consuming than I wish to put time into. You can see how quickly things start to become blurry. That is why I wrote this non-technical tutorial to cut through the entire PhD wise math and simplify the procedure so anyone can calculate PID tuning adjustments. Look also at Fig 5c which shows each loop scaled and what type of response to expect when tuning.


Ok, enough of that, lets get back to simplified tuning. Refer back to Fig 2 in part 1. By now I am sure you know how this diagram works so I will skip the boring walk through. Without further adieu we are going to do what is called manual tuning. This is used when you wish to start from scratch, don’t have the original setpoints or have a copter that does not fit factory setpoints due to weight, size etc…


Manual Tuning



WARNING -- First and foremost you must put safety first when doing PID manual tuning. I can’t stress this enough.

We need a perfectly level flat surface for the copter so we can calibrate our IMU (AHRS) to level. If you don’t do this step you will find your copter drifting which will not only make it harder to fly but also harder to see reactions and tune the copter.

We need to take some precautions with the copter. Once the copter IMU is leveled and calibrated in the XA Center we will need to decide how you will test the copter. You can either put the copter on a test stand allowing freedom of movement in the pitch and or roll axis using a string to tie the copter between two poles or you could just do free flight in air testing. Free flight air testing is very risky when we are first only tuning the (P) axis of the copter. Using this method you must make small adjustments and retest in steps being very patient. I would still tie the copter down only allowing a 6-18” lift off while doing this phase of testing. The height depends on how steady you can hold the hover.

We will need to set the setpoints for (Ki) and (Kd) values to zero in all three axes. We only want to tune one loop at a time. Please remember each loop has 3 axes to adjust. One word of advice, if you are test tuning for the first time you need to understand what subtleties to look for from the copter as you make your adjustments. Generally Proportional (Kp) and Integral (Ki) oscillations look the same when their values are much too high. As you get closer to a steady level hover, the oscillations change to the following:

Kp Oscillation – The frequency of the Oscillation is very fast. A good example would be seeing the copter vibrating (oscillating) quickly up and down and as time goes by the oscillations will continue to get worse until total stability lose and crash occurs unless you land.

Ki Oscillation – this is a slow circling oscillation close to a stable hover similar to TBE effect where you have the copter slowly circling a point instead of maintaining a steady no drift hover.

Kd Oscillation – this is not really an oscillation it’s more of and inability to correct from overshoot. Once you have reach a level of excessive overshoot then back the setpoint down until just a hint of overshoot is seen.

We will start by increasing the (Kp) slowly until the Output causes the copter to start oscillating. We will then back off the (Kp) setpoint until the oscillation stops. Leave the setpoint at this setting and move on to the (Ki) settings.

Slowly adjust the (Ki) setpoint from zero until the copter flies smoothly with no oscillations or wondering. How do you know when you have added to much (Ki)? Once you pass the point of stability any further increase in the setpoint will cause copter instability.

Once the (Kp) and (Ki) are giving you a steady no drift hover we will add a little (Kd) until the copter is acceptably quick on reestablishing hover after being disturbed by stick input or wind gust. Again, how do you know when you have added to much (Kd)? Once you have reached the limit any additional increase will cause excessive response and overshoot in-flight. Remember that the faster the PID loop tuning setpoints the more overshoot and instability that develops.


Hints and last words


When tuning your copter run the motors during testing at your hover speed. As your tuning gets closer and closer to stable flight you can raise the copter up out of ground affect.

(Kp) hints

If you are holding a hover and the copter all of a sudden breaks into a pitch or roll rapid oscillation, land ASAP and reduce the Kp value by 10-20% and try to hover again. If it breaks into a yaw oscillation use the same advice. When you have a point of no oscillations, increase the gain only until it oscillates a little when you kick it over in pitch or roll sharply then back off by 10% and retest doing so until oscillation is removed.

(Ki) hints

I call (Ki) my self trimming cruise control for high speed and windy flying. The quickest way to set (Ki) is to set it to at least the same value you have installed for (Kp) and once you are up and oscillating which you will then reduce the setting until the oscillation stops. Wait, you are not done yet. Now lift off and hover, kick the pitch over then bring back to hover. Did you see overshoot? If so continue reducing the setting until the overshoot is minimal. (There will always be a tiny bit of overshoot) Another thing, (Ki) is most prevalent on Yaw so make sure you adjust for that.

(Kd) hints

I see (Kd) as a way of enhancing the copter to a faster response and better handling performance level. Derivative action improves the controller’s action to predict what is yet to happen by projecting the (current rate of change) into the future. When you adjust the setpoint you are describing how far into the future you want to look.

I have computed a PI loop manually and projected that on a graph to show that you do not need complicated formulas to see what a setpoint value is going to do to the copters performance. I have included Fig 6A and Fig 6B in the picture sections so you can see 10 1/10th second samples and how the (Ki) and feedback loop work to bring the stick input back to a natural non oscillating state.



I hope you have enjoyed this Tutorial series and with your continued reader support and questions I will strive to bring more tutorials to bear with the information you wish to know.
batfire is offline Find More Posts by batfire
Last edited by batfire; Dec 06, 2012 at 10:42 PM.
Reply With Quote
Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
For Sale XAircraft Flight Control AHRS-S + FC1212-S V1 kevinv033 Aircraft - Electric - Multirotor (FS/W) 1 Jun 06, 2012 10:42 AM
Sold Here's a Cheap Radio System for an Entry Level Boat ALL American Boats (FS/W) 2 Apr 13, 2012 03:23 PM
Discussion Outfitting x-830-s help........ batfire Multirotor Talk 1 Mar 29, 2012 02:10 PM
For Sale XAircraft flight controller FC1212-S and AHRS-S barchiola Aircraft - Electric - Multirotor (FS/W) 4 Feb 18, 2012 06:45 AM
Qualcom 830 flight report (Lite Stik) Drewmotion Parkflyers 48 Apr 07, 2002 12:09 AM