Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Apr 18, 2012, 04:57 PM
Multi Rotor Maniac
Stunt Double's Avatar
Australia, VIC, Melbourne
Joined Oct 2007
2,734 Posts
Quote:
For some reason the ESCs woudl not allow me to test individualy. Kept getting a series of error beeps. Tried reversing throtle etc., no dice.
Thats doesnt sound right. Beeps on a tri usually indicate that the throttle range has not been set. You did calibrate each esc first then calibrate them to the kk right??
Stunt Double is offline Find More Posts by Stunt Double
RCG Plus Member
Latest blog entry: Cruisy Quad Flight
Reply With Quote
Sign up now
to remove ads between posts
Old Apr 18, 2012, 05:01 PM
Multi Rotor Maniac
Stunt Double's Avatar
Australia, VIC, Melbourne
Joined Oct 2007
2,734 Posts
Sparklet, thats a clever idea calibrating the acc in the gui by matching trimmed channel values.
Stunt Double is offline Find More Posts by Stunt Double
RCG Plus Member
Latest blog entry: Cruisy Quad Flight
Reply With Quote
Old Apr 18, 2012, 06:51 PM
Registered User
Joined Feb 2011
628 Posts
Ive noticed when I switch to stable mode it wants to move off in one direction pretty violently. I have dropped P value from 9 to 5 and its better, I will drop it again to 2.5 and hopefully that will be better again ?
The corrections should be very minimal because its pretty well trimmed now so I am assuming a much lower P will just put in small corrections which is all you need.
I have been doing the full throttle and moving sticks to trim the ACC but it always overshoots.

Thanks for you r screen shot.. Why are your D values so high ?
jbalat is offline Find More Posts by jbalat
Reply With Quote
Old Apr 18, 2012, 08:49 PM
Expat
Japan
Joined Apr 2010
3,010 Posts
Thanks. Ya, I followed the guide at rcexplorer for calibration/setup. It all seems to work fine on the bench in tri mode, just individual escs in rx throtle refuse to fire up.
Pulling all but 1 red esc wire did nothing. Most prevalent problem when all trimmed and in hover, is any movements of stick results in randomly weak right motor.
I'll try adding more padding to FC to eliminate more vibes? Bit endless I suppose. 8 wires are difficult to isolate from all vibes. Any other ideas? Recalibrate escs again?
TheNiceGuy is offline Find More Posts by TheNiceGuy
Reply With Quote
Old Apr 18, 2012, 11:43 PM
Registered User
Joined Feb 2011
628 Posts
Did you do a range check with motors idling ???

Also noticed on my quad with kk board the newer firmware did not like my cheap esc's so I went back to version 1.1 and all was good. I think the newer firmware runs at a higher frequency ?? so its good for turnigy plush or equivalent
jbalat is offline Find More Posts by jbalat
Last edited by jbalat; Apr 18, 2012 at 11:48 PM.
Reply With Quote
Old Apr 19, 2012, 12:40 AM
Expat
Japan
Joined Apr 2010
3,010 Posts
Ah, sorry, useing 1.6 firmware. And did a brief range check. While idling I collapsed antenna and went to other end of the house. No problem, but will try farther outdoors ASAP.
TheNiceGuy is offline Find More Posts by TheNiceGuy
Reply With Quote
Old Apr 19, 2012, 10:26 AM
Registered User
Cumbria in the U.K
Joined Jan 2008
1,920 Posts
Quote:
Originally Posted by Stunt Double View Post
Sparklet, thats a clever idea calibrating the acc in the gui by matching trimmed channel values.
Not my idea I came across it somewhere online.

To get the first check on how far out the accelerometer calibration is I do the following....

Start motors and increase the speed of the motors a bit. This won't work if the motors are left at the start up speed. You don't want enough speed for the multicopter to be threatening to lift off though.

Having got a nice speed then listen carefully and switch in the accelerometer.
Almost certainly you will hear some of the motors change speed. Switch back and forward from stable mode a few times and try to identify which motors speed up when the accelerometer is switched in. Be ready to switch back out of stable mode if one of the motors revs up enough to try and tip the multicopter over.

Now stop the motors and use the calibration technique to slow down the required motors. Start up and repeat the procedure again until there is virtually no difference in motor speed when you switch the accelerometer in and out.

This is a good procedure before the first test flight. Much better to find out the accelerometer calibration is way out while the thing is still on the ground. This prevents any unexpected veering off course when you switch to stable while flying.....

This won't achieve perfect calibration but it will put you pretty close. Flying and then trimming will complete the process.
sparklet is offline Find More Posts by sparklet
Reply With Quote
Old Apr 19, 2012, 10:49 AM
Registered User
Cumbria in the U.K
Joined Jan 2008
1,920 Posts
Quote:
Originally Posted by jbalat View Post
Ive noticed when I switch to stable mode it wants to move off in one direction pretty violently. I have dropped P value from 9 to 5 and its better, I will drop it again to 2.5 and hopefully that will be better again ?

Thanks for you r screen shot.. Why are your D values so high ?
Hi,
These things all tend to be different so one set of PID's may not work on another copter. They can be a starting point though.

If it is the P for level mode that you are dropping then I think you are masking the problem. Dropping the P value for level makes it less efficient and it will respond slower to changes in attitude. By backing it down to 2.5 you are, in effect, making the control so weak that it looks as though you have improved the problem.... you haven't, it just isn't responding as much as it was before and your level control will be very weak. Dropping it to 5 may make it look as though it has improved the situation whereas it is just a much weaker response now.

Both my quads run at the settings shown on my screen shot. Both quads are similar in build. The same settings also worked well on my tricopter.

I have seen people dropping P value to 2.5 before. If they need to go this low then I think there is a problem (I may be wrong !). The stock PID values should work for most multicopters with just some deviation to suit personal requirements.

I dropped the P values to 4.5 on my second quad when I built it. It was pretty wobbly and I was not impressed. It was nowhere as near stable as my first quad with the higher settings. Took P up from 4.5 to 5.1 and it was utterly transformed and was now as stable as my first quad and that was very stable even before the level control was switched in.

If the quad shoots off to the right in stable mode then the accelerometer needs calibrating as I explained in my post.

Stop motors, disarm the ESC's. Put throttle stick to maximum and then, in your case, hold the roll (aileron) stick over to the left. You will see the red LED flash quickly on the arduino board. Just guess about 10 flashes (it is very quick). Fly the copter and see what happens when you switch into stable mode. Keep repeating until you get it right....... also see my other post for checking it on the ground.

My D values are high because, along with the P and I settings they give me a very stable copter for photography. When auto level is working properly you will find that the copter is more reluctant to obey stick movements than when auto level is switched off... it wants to maintain it's position and doesn't respond as quickly to stick inputs.... it is now an absolutely superb camera platform
sparklet is offline Find More Posts by sparklet
Reply With Quote
Old Apr 19, 2012, 02:20 PM
Registered User
BigJimNZ's Avatar
Auckland New Zealand
Joined Sep 2010
1,294 Posts
Quote:
Originally Posted by jbalat View Post
Ive noticed when I switch to stable mode it wants to move off in one direction pretty violently. I have dropped P value from 9 to 5 and its better, I will drop it again to 2.5 and hopefully that will be better again ?
The corrections should be very minimal because its pretty well trimmed now so I am assuming a much lower P will just put in small corrections which is all you need.
I have been doing the full throttle and moving sticks to trim the ACC but it always overshoots.

Thanks for you r screen shot.. Why are your D values so high ?
Not taking the mick here John but are you sure your ACC is mounted parallel to your craft ?
I guess it makes a difference ?
BigJimNZ is offline Find More Posts by BigJimNZ
Reply With Quote
Old Apr 19, 2012, 02:23 PM
Registered User
BigJimNZ's Avatar
Auckland New Zealand
Joined Sep 2010
1,294 Posts
Quote:
Originally Posted by Stunt Double View Post
Sparklet, thats a clever idea calibrating the acc in the gui by matching trimmed channel values.
All that lead has not affected his brain. + 100 for that Sparkle lad
Darn that's a pretty slick cycle time you got there. How did you get it that low ?
BigJimNZ is offline Find More Posts by BigJimNZ
Last edited by BigJimNZ; Apr 19, 2012 at 03:09 PM.
Reply With Quote
Old Apr 19, 2012, 04:21 PM
You can't take the sky from me
scott page's Avatar
United States, WA, Richland
Joined May 2009
5,837 Posts
For Pacific North West Quad Folks.....

Slightly off topic -- but not entirely.

There is a charity Fun Fly in Richland Washington May 25 - 28. I convinced the organizer to dedicate two hours Sat and Sunday for multi-copters only on the second runway. I would love to see many multi-guys come together for food and fun -- but there are very few locally .. until they get so see some. ;-) For more information as it unfolds see http://www.rcgroups.com/forums/showthread.php?t=1618247 (also there will be a very rich raffle -- over $3000 in raffle items .. so far!)

Sorry if too far off topic -- don't know how else to reach multi guys in the North West.
scott page is offline Find More Posts by scott page
RCG Plus Member
Old Apr 19, 2012, 06:46 PM
Registered User
Joined Feb 2011
628 Posts
Well Sparklet P=2.5 didnt improve anything.. I dont know any more..

I regret going for this stable mode, it sends my heart rate up every time i switch it on.. And you are right its almost impossible to fly with it while its on.

When you say the red light flashes.. I actually see my Green light flashing on the arduino when I use full throttle to make my adjustments. I can never get it calibrated right, it always goes off one way or another. Sometimes if I try to fly it in stable mode it will lose its calibration altogether. It could be there is something wrong with the arduino or the wmp+ since it still doesnt stream properly to the pc gui.

Saying that it does fly ok in normal mode although it glitches a little now and then

I have received my KK board that was reserved for my new Quad.. All those in favour of dumping my MultiWill for the KK say I ???

Ian this is my setup here.. Looks pretty parallel
jbalat is offline Find More Posts by jbalat
Last edited by jbalat; Apr 19, 2012 at 06:52 PM.
Reply With Quote
Old Apr 20, 2012, 12:31 AM
Multi Rotor Maniac
Stunt Double's Avatar
Australia, VIC, Melbourne
Joined Oct 2007
2,734 Posts
Quote:
I have received my KK board that was reserved for my new Quad.. All those in favour of dumping my MultiWill for the KK say I ???
Nay, keep at it you will figure it out. Should try over at the multiwii forum those blokes will know whats going on. If you cant get the nunchuck to work just replace it with a sirius that way you will get baro and magnetometer too.
Stunt Double is offline Find More Posts by Stunt Double
RCG Plus Member
Latest blog entry: Cruisy Quad Flight
Reply With Quote
Old Apr 20, 2012, 02:02 AM
Registered User
Joined Feb 2011
628 Posts
Quote:
Originally Posted by Stunt Double View Post
Nay, keep at it you will figure it out. Should try over at the multiwii forum those blokes will know whats going on. If you cant get the nunchuck to work just replace it with a sirius that way you will get baro and magnetometer too.
Ok I will stick to it a bit longer.. Actually I did your trick and listened to the motor rpms to make sure they did not change when I flicked the switch and yeh it was much better.. I think it's more that I am a very nervous flyer especially when I have to trust a new gadget

Hey just want to bounce this off you. I find the tricopter is flying very nicely at low speed now but once I increase forward speed then it starts to twitch every now and then in the roll axis.. Is it possible that its not a problem with the board but that it is an aerodynamics issue caused by the landing rings, dare I say 'rolling vortex' ?
jbalat is offline Find More Posts by jbalat
Last edited by jbalat; Apr 20, 2012 at 07:04 AM.
Reply With Quote
Old Apr 20, 2012, 09:32 AM
Flying a Chipmunk in Portugal
jhsa's Avatar
NRW, Germany
Joined Sep 2010
3,270 Posts
are you flying with multiwii 2.0? I think there were a few people complaining about this twitching in the multiwii thread.. could be a software problem.. maybe you could also ask there??

Joćo
jhsa is online now Find More Posts by jhsa
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Yet another tricopter build. !!Updated parts list!! SebastianJ Multirotor Talk 2669 Dec 30, 2011 05:48 PM
Discussion What to do? Failed tricopter build! Razor_Racer Coaxial Helicopters 9 Mar 07, 2011 09:20 AM
Discussion Tricopter Build Razor_Racer Coaxial Helicopters 21 Feb 20, 2010 08:24 PM
Build Log Tricopter build SebastianJ VTOLs 1 Dec 22, 2008 04:34 PM
Discussion Trying to build a mechanically actuated tricopter, Help! stefalo Coaxial Helicopters 3 Jan 09, 2008 04:10 PM