Beta Details - dmgoedde

[dmgoedde]

Thanks for keeping us up-to-date.

Can't wait to see the Attopilot community grow worldwide.
Hope the export regs won't ruin this dream.I would be SHOCKED if the export classification is restrictive in the way it is for full blown 9 DOF IMU autopilots that can do dead-reckoning nav. I'll be frank... the v1.8 Atto for all its self-trimming and proven accuracy is still just a horizon sensing autopilot and is 100% dependent on having good GPS lock and relies completely on GPS for heading information for the navigation. By no means can it do any sort of dead-reckoning navigation and there are no gyros.

On another story... IMU prototype work is in full swing and I have refined my IMU code approach even more. Using very clean methods, code, and latest hardware the results are fabulous. The 1st IMU was about 10X noiser and less accurate and still worked fine in about 8 hours of "CoPilot" and autonomous flight on the Funtana by E-Flite and a flying wing. This new IMU is very very clean on the table top and I expect to get great results in flight testing this week.
http://www.vimeo.com/4994999

The link doesn't seem to work in this browser for me, so you might have to go to Vimeo to view this.

[CenTexFlyer]

Can't wait to see the Attopilot community grow worldwide.
Hope the export regs won't ruin this dream.

What? You mean like that other federal agency?

[Jack Crossfire]

Looks like a solid pine door & pendant chandelier. Must take a big motor to fly it.

[Gary Mortimer]

More power, are you saying Deans made his house fly now Jack?

Thats amazing, then perhaps the UAV will come under planning laws?

I wonder which is worse, town or airspace planners.

[saabguyspg]

Looks Good Dean, how do you prefer to be contacted? I PM'd you and realise you are likely quite busy but I am wondering if you have a delivery date in mind for the IMU atto? Also if you are looking for beta testers.

[dmgoedde]

Looks Good Dean, how do you prefer to be contacted? I PM'd you and realise you are likely quite busy but I am wondering if you have a delivery date in mind for the IMU atto? Also if you are looking for beta testers.My Gmail is really the best, but even more direct is the phone # on Attopilot.com, bottom of the page. Why make you browse over to it... the # is (480) 279-4557. Don't call right now I am going to bed after I finish this Guiness.

I already have 6 beta testers for the IMU Atto. 2 "civilians" and 4 UAV companies. The companies already have extensive experience with other high quality IMU autopilots including the TGE, MP2028, MP2128, and the Kestrel 2.x. One of the 2 individuals has some knowledge of the MP2028 but not a lot of direct experience, while the other individual has no IMU autopilot experience. Do you have IMU experience to compare the Atto IMU to? Call me if you do, as I would appreciate a bit more feedback compared to the other products already out there and 'proven'.

I was saying previously June for the IMU Atto. I'd really like to be the 1st to release a full-featured 6 DOF IMU autopilot for < $2k. It's looking like for highest perfomance in the real world, each IMU autopilot will require a fair amount of factory calibration and testing for quality assurance. I have come up with a clean method to do this as a batch process however. For example, some IMU 'snobs' I know of out there might say "A 'real' IMU is self correcting and tracks all the biases". Yeah, I know but tell me what your IMU's accel bias and scale factors do over the full mil-spec range ends like 80 or 85C or -40C, and how do you track and correct accel bias and scale factor? Not trivial! Also, I'm not sure there are enough independent variables to tease out that sort of data unless a magnetometer is included as well as tightly-coupled GPS into the IMU. I don't know of an accelerometer that has zero temp dependence for bias or scale factors, nor do I know of any rate gyros with all 3 axes perfectly orthogonal. Corrections are a must-do. Variation is a reality in any manufacturing process, statistically speaking. I can assure you that every Atto IMU I sell will be gauranteed to function and be right on target even if used at 85 C in the desert sun or at -40 in a baloon drop glider from 100k feet altitude. Also, I won't require any stupid calibrations by the end-user besides getting it approximately level. One time install and done. No day-to-day calibrations. Set it and forget it. Don't worry about the temperature either. I'm 100% focused on no-nonsense real world usage in crappy conditions. I fly a lot.

BTW - today is my 1 year anniversary of the crazy thing I call "quitting a well paying day job at Intel corp to do AttoPilot full time". I read Jack Crossfire's blog today for 1 hour, and in honor of him I will crack a joke: Thank you to all purchasers of Attopilot, as you have helped pay my mortgage so the Gov't doesn't have to bail me out.

[dmgoedde]

As I mentioned, I was reading Jack Crossfire's blog today. Ever since I annouced a serious "I'm in the club now" regarding IMUs, I have felt a special kindred spirit with Jack. I also read and saw that he has been doing a lot of PCB prototyping of SMT components lately. He respects me more now that I am in the IMU club, and because he is doing SMT of QFN parts, I respect him more. Ahhh, I recall the first QFN part I mounted in 2006 (that was my 1st ever SMT proto made 100% at home... no leaded parts, I went right for the jugular of the IDG-300 which is a 0.5mm pitch leadless package, and it worked). Solder paste stencil etched from 5mil copper sheet and the whole 9 yards.

In the last 2 months since I decided "hey, why not accelerate the IMU and get it working finally?" I have made 4 or 5 new prototypes, all of them 100% from scratch at home, err I mean at AttoPilot LLC. I've had to stretch my capabilities and figure out how to reliably use my laser cutter to make great mylar paste stencils for those nasty QFN parts like the IDG500. Vias are a lot of fun too, and I considered buying the equipment to plate through holes, and even to buy an LPFK from AltitudeAP. In the end I still just solder wires through Dremelized hole to make vias. All this effort so I can go from idea to working proto in 6 hours. No, waiting for PCB Express is too long. OK, so I'm on my 2nd Guiness now after a very kick-butt IMU day of killing 3 problems that appeared to be bad, but I defeated. The well deserved beer is making me ramble I guess.

[saabguyspg]

Thanks Dean for all the info, our company does not have any IMU auto pilot experiance sorry. We would only be able to beta test as a "potential user customer that is not a rocket scientist" We are in the market for a autopilot such as 1.8 for the price point immediately. For the IMU price point of sub 2000 we will be in that market of spring next year unless our biggest contract comes through this year then you will be getting a phone call from me ASAP. For our needs the 1.8 likely will be sufficient. I will try and give you a call next week. For now we are working with the DIY drones style autopilot because it gets the job done. However we like to look foreward in this industry as much as possible.

[Connexxion]

Dean,

incase you'll need to proof the system is foolproof...I'm all yours! ;)

[Robert Wagoner]

We are happy to report our initial trials with the AttoPilot worked beyond expectations. surprisingly when meeting with Dean and Bill the installation took less than 30 minutes and then we were off to the flying location. :)

One flight manual to double-check gains and the second flight we were flying autonomous. our third flight was aborted do to high winds for this airframe.

Dean then made a surprise test of his new IMU version in the 25+ MPH winds that was very impressive. we are very excited to proceed with our developments and integrations with our projects.

Here is a pictures and video of the EDF prototype platform in manual control that later flew autonomously. as we progress I'll share more of our Electric Jet UAV developments.

http://www.ejf.com/media/ejf-uav1.wmv

--RW

[airmcn_3]

We are happy to report our initial trials with the AttoPilot worked beyond expectations. surprisingly when meeting with Dean and Bill the installation took less than 30 minutes and then we were off to the flying location. :)

One flight manual to double-check gains and the second flight we were flying autonomous. our third flight was aborted do to high winds for this airframe.

Dean then made a surprise test of his new IMU version in the 25+ MPH winds that was very impressive. we are very excited to proceed with our developments and integrations with our projects.

Here is a pictures and video of the EDF prototype platform in manual control that later flew autonomously. as we progress I'll share more of our Electric Jet UAV developments.

http://www.ejf.com/media/ejf-uav1.wmv

--RW


Flys slow....... Nice!

[Tom Harper]

Very nice - I like the airplane!

[sim559]

We are happy to report our initial trials with the AttoPilot worked beyond expectations. surprisingly when meeting with Dean and Bill the installation took less than 30 minutes and then we were off to the flying location. :)

One flight manual to double-check gains and the second flight we were flying autonomous. our third flight was aborted do to high winds for this airframe.

Dean then made a surprise test of his new IMU version in the 25+ MPH winds that was very impressive. we are very excited to proceed with our developments and integrations with our projects.

Here is a pictures and video of the EDF prototype platform in manual control that later flew autonomously. as we progress I'll share more of our Electric Jet UAV developments.

http://www.ejf.com/media/ejf-uav1.wmv

--RW
Good start but what is the flying weight now?
How much more weight can be added?
Will you use this platform for gathering information from above?
Cheers Sim

[Mecha]

guys quick question.

To get Atto to read the WP.txt file it just needs to be in the SD card right? Meaning there is nothing to set in the SET.txt file for it to know there is a WP file.

I was testing mine and the plane loiter in autonomous without WP as planed, however, when the WP file is present, the plane does not follows the coordinates it just wants to flight in a straight line. I have put the $R in the first line of the WP file and also tried it without it using local coordinates. I have checked the coordinates using a hand held GPS and Google Maps.

Can anyone tell me what could be rong.

WP Latitude Longitude Alt Triggers Airspeed

$R00000,25.770444,-80.366194,0100,00000000,50
$00001,25.772222,-80.365861,0100,00000000,50
$00002,25.772083,-80.367861,0100,00000000,50
$00003,25.770917,-80.367583,0100,00000000,50
$00004,25.771417,-80.366944,0100,00000000,50
$00005,25.772222,-80.365861,0100,00000000,50
$h00006,25.771417,-80.366944,0100,00000000,50
$####### END OF FILE

[Tom Harper]

Here is a WP file for comparison:

WP Latitude Longitude Alt Triggers

$00000,34.05829,-106.92590,0140,00000000*29
$00001,34.05908,-106.92896,0140,00000000*21
$00002,34.06149,-106.92979,0140,00000000*2C
$00003,34.06332,-106.92774,0140,00000000*20
$00004,34.06503,-106.92561,0140,00004000*21
$00005,34.06320,-106.92406,0140,00000000*23
$00006,34.05579,-106.92395,0140,00000000*24
$00007,34.05383,-106.92573,0140,00000000*28
$00008,34.05599,-106.92873,0140,00000000*27
$00009,34.06318,-106.92865,0140,00000000*2D
$00010,34.06490,-106.92676,0140,00004000*2A
$00011,34.06313,-106.92465,0140,00000000*23
$00012,34.05569,-106.92454,0140,00000000*2A
$00013,34.05386,-106.92632,0140,00000000*2E
$00014,34.05599,-106.92916,0140,00000000*28
$00015,34.06315,-106.92908,0140,00000000*27
$00016,34.06483,-106.92758,0140,00004000*23
$00017,34.06310,-106.92521,0140,00000000*27
$00018,34.05579,-106.92506,0140,00000000*27
$00019,34.05379,-106.92699,0140,00000000*25
$00020,34.05602,-106.92971,0140,00000000*2F
$H00021,34.05795,-106.92585,0140,00000000*6E
$####### END OF FILE

The asterisk may act as a delimiter so I'd eliminate the comma and add an asterisk.

The last 2 digits are the checksum which is no longer used. You might leave your 2 digits to give the software something to chew on.

For initial flights drop the 'R' and use absolute coordinates.

Don't know what version you have, but airspeed is a function of the SET file, not WP.

A WP to fly around that lot should look something like this:

WP Latitude Longitude Alt Triggers

$00000,25.77002,-80.36526,0100,00000000*15
$00001,25.76956,-80.36656,0100,00000000*19
$00002,25.77106,-80.36729,0100,00000000*1F
$H00003,25.77148,-80.36618,0100,00000000*5F
$####### END OF FILE

Tom

[Mecha]

Thank you Tom,

The log file says V1.8 2_28_09, what version do you have?

I thought the air speed was only a function of the SET file if not set in the WP otherwise you could choose different values. I may be wrong.

How do you figure out the check sum value?

In the sample WP Dean has: trigger, speed*check sum

I will test again dropping the speed all together see what it does.

[Irv]

Thank you Tom,

The log file says V1.8 2_28_09, what version do you have?

I thought the air speed was only a function of the SET file if not set in the WP otherwise you could choose different values. I may be wrong.

How do you figure out the check sum value?

In the sample WP Dean has: trigger, speed*check sum

I will test again dropping the speed all together see what it does.

My WP.txt file that works is:

WP Latitude Longitude Alt Triggers airspeed

$00000,33.24568,-117.18011,0060,00000000,45*
$00001,33.24704,-117.17863,0060,00000000,45*
$00002,33.24660,-117.17641,0060,00000000,45*
$00003,33.24601,-117.17641,0060,00000000,45*
$00004,33.24533,-117.17746,0060,00000000,45*
$00005,33.24541,-117.17825,0060,00000000,45*
$00006,33.24567,-117.18006,0060,00000000,45*
$00007,33.24568,-117.18011,0040,00000000,45*
$00008,33.24704,-117.17863,0040,00000000,45*
$00009,33.24660,-117.17641,0040,00000000,45*
$00010,33.24601,-117.17641,0040,00000000,45*
$00011,33.24533,-117.17746,0040,00000000,45*
$00012,33.24541,-117.17825,0040,00000000,45*
$00013,33.24567,-117.18006,0040,00000000,45*
$00014,33.24568,-117.18011,0020,00000000,45*
$00015,33.24704,-117.17863,0020,00000000,45*
$00016,33.24660,-117.17641,0020,00000000,45*
$00017,33.24601,-117.17641,0020,00000000,45*
$00018,33.24533,-117.17746,0020,00000000,45*
$00019,33.24541,-117.17825,0020,00001000,45*
$00020,33.24567,-117.18006,0020,00001000,45*
$00021,33.24626,-117.17832,0050,00000062,45*
$####### END OF FILE

Fly 7 point rectangular path at 60m, 40m, and 20m altitude,
trigger at WP19 (Point 6), then loiter 100M North of the RW C/l
in a 100m loiter circle, Altitude 100M and 45Kph

The airspeed is after the 8 trigger digits
The ending * is leftover from the check sum value.

[Robert Wagoner]

Good start but what is the flying weight now?
How much more weight can be added?
Will you use this platform for gathering information from above?
Cheers Sim

Hello Sim,

This is the proof of concept vehicle for flight testing using a 2.5" EDF unit not intended for much of a payload about 1 lb. the larger version in completion uses a 5" EDF unit for flying a couple of different camera systems weighing about 10 lbs.

Regards,

--RW

[Tom Harper]

Irv,

Good stuff - does the GCS software accomodate this airspeed format?

I'll try it.

Thanks,

Tom

Mecha,

Don't worry about it. The check sum is no longer used - just put the * as a sentence delimiter.

Tom

[Mecha]

The problem with my WP files where the * for check sum and the 6th decimal place on the Lon and Lat values. Atto only accepts 5 decimal places. Now it works fine. Thanks.

BTW this 6th decimal will cause the GCS not to read the telemetry info.

[dmgoedde]

The problem with my WP files where the * for check sum and the 6th decimal place on the Lon and Lat values. Atto only accepts 5 decimal places. Now it works fine. Thanks.

BTW this 6th decimal will cause the GCS not to read the telemetry info.Firmware will be updated to ignore the 6th decimal. The * at end of each line will be dropped as well (as a requirement). This is legacy stuff.

If you don't manually add the , and 0-255 airspeed at the end of each line, then your Atto defaults to the MIN airspeed per SET file.

[Tom Harper]

Dean,

I just learned about the airspeed setting from Mecha and Irv's WPs. Mine has been defaulting to tne Min setting. Will try this out today.

Individual airspeed settings for each leg should allow automated landing approaches. Not good enough for touch down, but it should bring you right down the runway on final. Will play with that today. How low can you get with the FMA XY sensors and still be safe?

Tom

[airmcn_3]

Dean,

I just learned about the airspeed setting from Mecha and Irv's WPs. Mine has been defaulting to tne Min setting. Will try this out today.

Individual airspeed settings for each leg should allow automated landing approaches. Not good enough for touch down, but it should bring you right down the runway on final. Will play with that today. How low can you get with the FMA XY sensors and still be safe?

Tom


Tom,

That’s a good question, we have used assisted all the way to landing. So..... You should be good, just make sure you have a good clear view to the horizon.

[sim559]

Hello Sim,

This is the proof of concept vehicle for flight testing using a 2.5" EDF unit not intended for much of a payload about 1 lb. the larger version in completion uses a 5" EDF unit for flying a couple of different camera systems weighing about 10 lbs.

Regards,

--RW
so you are looking at 600 to 1000 watts?
Sim

[dmgoedde]

Dean,

I just learned about the airspeed setting from Mecha and Irv's WPs. Mine has been defaulting to tne Min setting. Will try this out today.

Individual airspeed settings for each leg should allow automated landing approaches. Not good enough for touch down, but it should bring you right down the runway on final. Will play with that today. How low can you get with the FMA XY sensors and still be safe?

TomWell, Chris (airmcn_3) was actually here and witnessed some assisted landings of a flying wing, using mode "2" (HARC) wherein the throttle is an RC passthrough and can be shut off but the plane is still in CoPilot mode. It worked fine, and the plane flying right past 4 of us standing there got a lot LESS disturned than we all thought it might. Keep in mind the thermopiles have a FOV of like 110 degrees... a small object won't corrupt their attitude estimation much.

BTW - IMU works a LOT better than thermopile for assisted landing, and in fact that is the only way I land my IMU test vehicles. I am progressing smoothly towards the standard of all other sub-$10k IMU autopilots worth their salt: 100% autonomous take off and landing. It is something that most of my commercial customer want: ability to detect release from a bungee and then firing up the motor some distance out in mid-air.

Getting back to thermopiles and assisted landing for a second... my tiny 8 ounce 18" span plank MAV was almost impossible to fly in RC. Takeoff wasn't so bad as the speed was high and increasing with full power, but at low speed and trying to flare for landing really sucked and gave me heart palpitations. I ended up HAVING to use "Co-Pilot" mode for all landings with that thing. I'll be trying the IMU in a 22" plank by this weekend, and the purpose of that is to polish 100% autonomous deep stall landing, AND autonomous takeoff insofar as I throw the "armed" plane into the breeze and let the autopilot detect launch shock and fire up the motor.

[Tom Harper]

Dean,

Worked pretty well, brought it right onto the end of the runway at 50 ft altitude.

Question: Is there a magic test mode that will cycle the trigger output at the parameters in the set file? For some reason my A650 does not take pictures in flight. Everything works on the bench but no photos in flight. Need to check out the electrical link.

Tom

[Robert Wagoner]

so you are looking at 600 to 1000 watts?
Sim

Or more. we have 3800W to work with on the larger platform.

--RW

[dmgoedde]

Dean,

Worked pretty well, brought it right onto the end of the runway at 50 ft altitude.

Question: Is there a magic test mode that will cycle the trigger output at the parameters in the set file? For some reason my A650 does not take pictures in flight. Everything works on the bench but no photos in flight. Need to check out the electrical link.

TomI'd need to also review your SET and WP.txt files.

If you have a modem link then you can manually fire the trigger on the ground from the GCS (you can do this on the bench in your shop), or do it via a wired connection and cut out the RF link.

[Tom Harper]

Here they are.

[dmgoedde]

So you want to trigger the photo for 1 second duration in the "on state" and trigger it every 2 seconds on the selected waypoints?

[Tom Harper]

Yes and that is what it shows on the log file, but the camera shutter never fires. I've checked out all of the connectors and wires and everything seems OK. That's why I wanted a scheme to force the signal from the Atto driver.

Using a cable into the telemetry input seems like an excellent test facility.

Tom

[carguy84]

How are you triggering the camera to fire? USB? Servo? Infrared?

[dmgoedde]

Yes and that is what it shows on the log file, but the camera shutter never fires. I've checked out all of the connectors and wires and everything seems OK. That's why I wanted a scheme to force the signal from the Atto driver.

Using a cable into the telemetry input seems like an excellent test facility.

TomThis might sound goofy, but you can also do a "driving" test with the plane in your car (motor disabled) and waypoints at intersections and on some paths your intended triggering scheme. Is the servo moving far enough to hit the shutter? Is 1 second trigger duration long enough to get auto focus and fire the shutter? Is the 2 second delay long enopugh to allow the camera to reset between photos (seems awfully short to me).

Your two servo positions are nearly at the end of servo travel also... is the servo fast enought to get to the triggered position and stay there to cause a photo?

[Tom Harper]

Carguy,

I use a servo on the camera that is connected to the Atto servo driver. It goes to focus at 1500Us and actuates at 1650Us. The driver output that is logged is 1100 to 1900. I think I have a connector problem because this has always worked with other cameras - I just can't find it!

Tom

[Tom Harper]

Dean,

Is it OK to use the #1 forward position on the XY sensor?

Tom

[airmcn_3]

Dean,

Is it OK to use the #1 forward position on the XY sensor?

Tom

All positions work as described...

[carguy84]

I think when using a servo to press the shutter, 1 second isn't going to be enough time for the P+S to focus and fire off a photo. Especially when the plane is at a decent altitude and vibration comes into play. Try changing it to 3 seconds and see what happens.

[Gary Mortimer]

What carguy said, try lowering the image resolution if you want to get the shot taken and processed in time.

I know you need to bang the shots quick off unless you are awfully high if you want to stich verticals.

[Tom Harper]

Not the problem. The camera works on the bench at this rate and full resolution. What I am after is a test set up that will use the whole system. Deans idea of a telemetry cable provides it. I just have to put it together.

Tom

[Gary Mortimer]

Ah but when its flying and shaking it might not

[brakar]

Know I must be missing something here, but what is the adwantage from using the autopilot to trigger the camera, why don't you jurt use canon and chdk? brakar

[Tom Harper]

brakar,

Attopilot advantages:

Set and forget - no start up process.

Attopilot will trigger the camera on selected legs only using either time or distance over the ground. The camera operation is synchronized with the desired coverage.

Distance over the ground is a major advantage since it gives you the same overlap on the upwind and downwind legs.

The Attopilot log provides the lon/lat of every image.

I suspect there may be some advantages to using the two in combination.

Tom

[brakar]

Thanks for your answer Tom,

I am hoping to buy an Atto when it becomes awailable to Europe. My thought was to use an Ixus 980 and a chdk-script I have downloaded to trigger the camera at fixed time intervals. No wiering, servos or programming, (must push the button to start the script though). With a 16GB memory card It would be possible to shoot one 25MB RAW-picture every 3 sec for about 30 minutes. If Atto rocord/provides lan/lot and time, I would have all datas required - since the camera also stores time when the pictures are shot.

I understand from your previous post that Atto has some sort of clock or timer onboard, but is the time also stored in the log file?

Different wind-condition on the legs might be a challenge.

brakar

[airmcn_3]

Thanks for your answer Tom,

I am hoping to buy an Atto when it becomes awailable to Europe. My thought was to use an Ixus 980 and a chdk-script I have downloaded to trigger the camera at fixed time intervals. No wiering, servos or programming, (must push the button to start the script though). With a 16GB memory card It would be possible to shoot one 25MB RAW-picture every 3 sec for about 30 minutes. If Atto rocord/provides lan/lot and time, I would have all datas required - since the camera also stores time when the pictures are shot.

I understand from your previous post that Atto has some sort of clock or timer onboard, but is the time also stored in the log file?

Different wind-condition on the legs might be a challenge.

brakar


Clock is in the LOG file at all times.

Different wind conditions on each leg will be managed by Atto..... You will have no problems.

Chris

[kbosak]

Clock is in the LOG file at all times.

Different wind conditions on each leg will be managed by Atto..... You will have no problems.

Chris
Of course he will have as he plans triggering the photos at fixed time intervals. So he will end with less photo overlap in one direction. On the other hand camera continuous fire freq will not allow him to make 60% overlap no matter the situation without going above say 200..300mAGL.
He will have a lot of fun interpolating the data to match the position to the photos.

[airmcn_3]

Of course he will have as he plans triggering the photos at fixed time intervals. So he will end with less photo overlap in one direction. On the other hand camera continuous fire freq will not allow him to make 60% overlap no matter the situation without going above say 200..300mAGL.
He will have a lot of fun interpolating the data to match the position to the photos.


There is this program that you load in your entire picture file to and it starts working them out. The finished product is a 3D picture image that you pan around on your computer. I am headed to a friends who has this program; I will find out what it is and post a link.

Chris

[phenomii]

hey im new to the idea to an autopilot

im looking to make it my self.

i have programing skills, and electronics skills

would some please provide some links

that have the steps involved and all the components that ill be needing

any help would be great :D

cheers

[sim559]

hey im new to the idea to an autopilot

im looking to make it my self.

i have programing skills, and electronics skills

would some please provide some links

that have the steps involved and all the components that ill be needing

any help would be great :D

cheers

Get a life
Sim

[phenomii]

all i was asking is a lil guidance

[airmcn_3]

all i was asking is a lil guidance


Don’t let Sim get to you he obviously does not feel like being nice. This particular thread is full of information on how to start building your own autopilot; there are also other threads to check out. Try the search engine and look for Build your own autopilot.

My suggestion and your best bet would be to start with something like the ArduPilot and find out what your about to get yourself into.

Good luck.

Chris

[phenomii]

Thanks for that,

ive been looking around on the net, everything pretty much leads here, ive been reading this thread and ur right it helpd alot

cheers

[Gary Mortimer]

Heres very good as well http://www.diydrones.com



Thanks for that,

ive been looking around on the net, everything pretty much leads here, ive been reading this thread and ur right it helpd alot

cheers

[sim559]

all i was asking is a lil guidance
If what you say about your skills are true, then I see no reason to ask for help on this thread.
It is in my nature to treat all questions like yours as one that comes from a tosser.
As this is the internet and any one can say or claim anything.
I will be waiting for proof before I reconsider my opinion of you or any one else.
As for Atto. I did buy one and it is an extremely good system.
Have log over 100 flights.
sim

[airmcn_3]

If what you say about your skills are true, then I see no reason to ask for help on this thread.
sim


Come on Man.... You have any idea who and what kind of individuals ask questions on this forum. You have everything from rocket scientist to everyday Joes.

Anyhow back to the Atto thread.

Atto V1.8 is a great unit, might be a bit steep learning curve for the beginner but with your background I am sure you will be able to figure it out. Besides there’s great help here on the forum.... Well.... Most the time.

Good luck,

Chris

[CenTexFlyer]

It is in my nature to treat all questions like yours as one that comes from a tosser.


The "JettPilot" of the UA forum! :cool:

[sim559]

The "JettPilot" of the UA forum! :cool:
You bet .
I bet he is not what he claims.
Sim

[dmgoedde]

all i was asking is a lil guidanceYou can go my route of starting without guidance and figuring it all out from scratch (I think some people don't want to believe me on that). Got no kids and lots of free time? Then it might work, and I think in the end is most fruitful, but hardest. If you have demanding day job and kids and can only spend a few hours/week on it, then I agree with Airmcn_3: buy an Ardupilot, or get into Paparazzi and you will learn about hardware choices, PCB layout, programming for this type of ver demanding embedded application, and flight control (stability, navigation, payload management). You can make it as complex as you want to.

My Avatar pic is from 2006 when I still swam and ran a lot. Nice flat stomach and muscular arms and chest. Now I am 20 Lbs heavier because I do AttoPilot 60+ hours a week. Just be careful.

Also, I recently started reading the JackCrossfire blog. He has tons of time to not only work on the nuts and bolts stuff and gets results on complex projects, but has apparently 2X the free time I have because he adds about 4 new pages of blog each day. Bewarned of the "4U"... everything is "PCB making at home 4U" and "My trip to NASA 4U" and subtle references to drugs like his "Vicoden Copter" and something about "Heroine" worship refering to a robot, then later to his girlfriend that I think left him a couple weeks back.

So there you have it: if you get too deep into your own autopilots you can expect to lose yourslef to insanity.

[dmgoedde]

Why not time for some pics from me? Sure, I can prototype 2-sided PCBs on-site with line widths down to 0.15mm (4 mil), laser cut mylar stencils for solder pasting, got a real reflow oven (not a toaster oven). What I can't do (yet) is drill 12 mil holes and through plate them for good vias, so in the end my protos have to be enlarged, unless the design is verified good then I submit it for professional PCB manufacture and draw it at THEIR manuf limits, not mine. The results are shown below.

Watchout Sparkfun... I have created what is probably the smallest micro breakout for the ADXRS gyros. PCB is barely larger than the gyro itself, all passives are on the backside (6 caps, 0603 size) and 5 pads to physically connect it to a main PCB at right angle, and appropriate solder mask onthe edges to cross brace 2 of these.

[Tom Harper]

You're having too much fun!

Tom

[airmcn_3]

You're having too much fun!

Tom


Fun might be pushing it! ;) Lots of sleepless nights......

[Gary Mortimer]

too out there so I removed it, need pub filter

[sim559]

too out there so I removed it, need pub filter
I may need one too. Do you have a good link to one?

[2jers]

Dean, any update for release date (international)???.

[brakar]

Gary, you may need to do some fine-tuning on that filer of yours. Currently, it appears to be 100% effective, making this place a lot less interesting.

Best regards, brakar ;)

[biloxthecat]

No news from state department yet regarding the export classification of the v1.8. The window for a result is between mid May and mid-June. Because the technical side of work is alone far more that I can chew, but I can't afford to drop the ball on export expertise, I have procured an experienced to handle it.

Bill Nailor, owner of Royal Hobbies in Chandler, AZ is the guy. Bill brings lots of experience in shipping, including export from US. He is also a whiz at documentation and training. Bill will be the export "czar" for AttoPilot v1.8 (thermopile) and IMU products. He is also doing lots of flying with the Atto v1.8 and making the documentation more complete and user friendly.

Although I can write user-friendly manuals, lately it has worked out that I just don't have time to write documentation past a rough cut level. Bill is filling this gap.

I have been sold out of the initial batch of Atto v1.8 for about 2 weeks now. More will be made in anticipation of the sales to Europe and Australia. There will also be announcements regarding distributorships and availability of the IMU AttoPilot

Hi Dean,

Any updates on an international release date?

Thanks!

[dmgoedde]

Well, last detail for the Atto v1.8 is my export compliance lawyers are waiting on me to fax a couple forms, and I have Bill Nailor working the Attopilot e-mails to help people out.

[2jers]

Come on........PRESS THE BUTTON.... :-) :-) :-)

[CenTexFlyer]

Well, it won't be going to Mongolia....... :(

[Gary Mortimer]

Oh for the Gengis thing

[saabguyspg]

well I am on board now.. after spending hours messing with ardupilot I am putting it on the back burner. I need something that works and works well.

I have sent am email to Dean and called him... now I can't wait to hear back and get things rolling.

I was originally planning to put it in my Multiplex Mentor that has full pan and tilt for the camera but now I am thinking of just putting it in the multiplex xeno with the camera pointing straight down... so many options I may need two of them if it works as well as it sounds like it does.

Is there a popular thread dedicated to atto anywhere? or is this it?

[2jers]

Is there a popular thread dedicated to atto anywhere? or is this it?


http://attopilot.ning.com/

[sim559]

well I am on board now.. after spending hours messing with ardupilot I am putting it on the back burner. I need something that works and works well.

I have sent am email to Dean and called him... now I can't wait to hear back and get things rolling.

I was originally planning to put it in my Multiplex Mentor that has full pan and tilt for the camera but now I am thinking of just putting it in the multiplex xeno with the camera pointing straight down... so many options I may need two of them if it works as well as it sounds like it does.

Is there a popular thread dedicated to atto anywhere? or is this it?
Atto works very will. Not hard to learn or to set up. May take as few as 3 flights or as many as 10. It just depends on how good you can read and see what was logged. From the log you can make adjustments.
so what were some of the pit falls in the ardupilot? you can send link or let me know here. As it would be nice to have a comparison on thread
Thanks
Sim

[saabguyspg]

Ardu is good if everything goes well and you have lots of time to get it going... I don't have the time but I am still going to keep trying. Here is my list of problems and comments but remember that I still think it's a great project but lets call a spade a spade. Also I realise it is DIY and for you to learn about things so that is why I am still going to persue it.

1) ordered the board from spark fun with recommended cables to find out that the 5.0v cable does not work
2) ordered the recommended cable from DIYdrones store to find I spent 30$ US for a usb cable that just has wires split to a connector (something I am pretty sure any diy person can do)
3) the adru code loading program is very fincky (it won't run on one of my computers and the other is it hit and miss) this has nothing to do with the ardupilot but it's still a pain
4) my board came with the solder jumper not soldered for power from the servo lead (easy fix) but it's not easy to find when all you are told to do is look at the trouble shooting tips
5) my sheild board does not turn the 406 gps off so that I can load code (no biggy but that is the whole point of the board)
6) the 406 gps is not ublox so apprently it is not as smooth and accurate as ublox 4hz
7) support is really non existant any questions are answered with "read the trouble shooting guide" or "we have sold lots and there have been no complaints" which may be true but there are also almost no testimonials of working ones that I can find either.
8) I ordered fma sensors to work with this, then fma ran out of z sensors so now I have to order a new set of sensors from DIY when they are ready or modify my sensors.
9) cost: I have spend a couple hundered dollars in parts and shiping and really all i have left that I can keep using is the main adru board and the shield which are only like 100$ in themselves so the cost is quickly adding up to half that of the atto which is obviously more powerfull.

I have found that most of my problems are had by everyone else too so I am not a complete idiot ;-)

Again though it is a diy project. I will keep on with it but right now I need something that works and I can tinker this winter with this other one.

so to re-cap, ardu is cool... very cool... I want to keep at it and dont get me wrong but this is just what I have experianced so far.

My only fear is that as soon as I buy atto he will drop his prices (he mentioned in another post that he would be dropping them soon) I only hope he remembers us that bought his first set and gives us a deal on the next one or supports us.

that is the other thing I want something I can get answers on within a day. I would imagine if paying 800$ for something you will get some support.... but I have sent a few emails to Dean and even sighned the waiver etc but have not heard back from him as of yet to place my order ;-)



Atto works very will. Not hard to learn or to set up. May take as few as 3 flights or as many as 10. It just depends on how good you can read and see what was logged. From the log you can make adjustments.
so what were some of the pit falls in the ardupilot? you can send link or let me know here. As it would be nice to have a comparison on thread
Thanks
Sim

[airmcn_3]

Ardu is good if everything goes well and you have lots of time to get it going... I don't have the time but I am still going to keep trying. Here is my list of problems and comments but remember that I still think it's a great project but lets call a spade a spade. Also I realise it is DIY and for you to learn about things so that is why I am still going to persue it.

1) ordered the board from spark fun with recommended cables to find out that the 5.0v cable does not work
2) ordered the recommended cable from DIYdrones store to find I spent 30$ US for a usb cable that just has wires split to a connector (something I am pretty sure any diy person can do)
3) the adru code loading program is very fincky (it won't run on one of my computers and the other is it hit and miss) this has nothing to do with the ardupilot but it's still a pain
4) my board came with the solder jumper not soldered for power from the servo lead (easy fix) but it's not easy to find when all you are told to do is look at the trouble shooting tips
5) my sheild board does not turn the 406 gps off so that I can load code (no biggy but that is the whole point of the board)
6) the 406 gps is not ublox so apprently it is not as smooth and accurate as ublox 4hz
7) support is really non existant any questions are answered with "read the trouble shooting guide" or "we have sold lots and there have been no complaints" which may be true but there are also almost no testimonials of working ones that I can find either.
8) I ordered fma sensors to work with this, then fma ran out of z sensors so now I have to order a new set of sensors from DIY when they are ready or modify my sensors.
9) cost: I have spend a couple hundered dollars in parts and shiping and really all i have left that I can keep using is the main adru board and the shield which are only like 100$ in themselves so the cost is quickly adding up to half that of the atto which is obviously more powerfull.

I have found that most of my problems are had by everyone else too so I am not a complete idiot ;-)

Again though it is a diy project. I will keep on with it but right now I need something that works and I can tinker this winter with this other one.

so to re-cap, ardu is cool... very cool... I want to keep at it and dont get me wrong but this is just what I have experianced so far.

My only fear is that as soon as I buy atto he will drop his prices (he mentioned in another post that he would be dropping them soon) I only hope he remembers us that bought his first set and gives us a deal on the next one or supports us.

that is the other thing I want something I can get answers on within a day. I would imagine if paying 800$ for something you will get some support.... but I have sent a few emails to Dean and even sighned the waiver etc but have not heard back from him as of yet to place my order ;-)


Nice post!

Just a little info.

Dean is crazy busy right now working on the IMU, we have been in talks about getting customer service and technical support together so that he can focus on engineering. Talks have completed and there will be full time support very soon.

I just spoke to Dean specifically about your order and he will be sending you an e-mail this evening. There is a 1 week backorder so your unit will not ship out until after the holidays.

Can you please send your e-mail address to my PM box so I may include it in an e-mail to Dean?

Thank you,
Chris

[saabguyspg]

No problem Chris I undestand and don't expect the impossible!!!

As for the post above I want to re-state that I am not unhappy abou the ardu project but maybe some of those things will help others avoid some pitfalls.

I will pm you my email. Hey have a good 4th of July! (July 1st is Canada Day for us as I am sure you know ;-) so we are off work tomorrow and I am taking the rest of the week as well to enjoy my newborn son!!!!!

[airmcn_3]

No problem Chris I undestand and don't expect the impossible!!!

As for the post above I want to re-state that I am not unhappy abou the ardu project but maybe some of those things will help others avoid some pitfalls.

I will pm you my email. Hey have a good 4th of July! (July 1st is Canada Day for us as I am sure you know ;-) so we are off work tomorrow and I am taking the rest of the week as well to enjoy my newborn son!!!!!


Thanks man! Congratulations on your new boy! Yes it’s a double for me as my birthday is on the 4th as well!

I will be looking for your PM.

Thanks,

Chris

[saabguyspg]

pm sent... excitement building... this system i beilve is just what my business needs to reduce our costs and increase our reliablity and safety.

[dmgoedde]

4 servos / 4 Rx channels, for the simplest UAVs requiring only 3 flight control and 1 trigger.Dedicated expander port (serial com at 230.4 kbaud) to drive up to 30 additional servos11g accels and 300 deg/sec gyros. Fully 3D capable IMU Improved airspeed filtering Dedicated 2-way telemtry with GPS data in the $ATTO string. User defined baud rate from 4800 to 230.4 kbaudSD data logging is now an on-board capture of the serial stream, as a backup in case of COMS loss9 parameters are temperature compensated over the full mil spec -40 to +85C range, using 14 point linear interpolation.Uncorrected gyro drift is < 1 degree / minute

New GCS (non-hobby) writing is underway for commercial/military users. Will start to utilize onboard non-volatile memory for storage of tuning parameters, GUI-driven tuning, and grid-based integrated moving map in the GCS (goodbye Google Earth) and extensive real-time retasking of the flight plan. This autopilot is not merely an Atto v1.8 with an IMU, it is a somewhat expanded capability version of the previous Attopilot, with an all-weather IMU.

Rather than use 2g or 3g accels, I chose 11g and with proper RC filtering and IMU filtering, this IMU is quite tolerant of airframe vibration, or at least I should state that regarding vibration it is not overly fussy.

Oh, BTW, this shown unit is a quickie 2-layer PCB design. The final version on 4 or 6 layer PCB will be somewhat more compact, and lighter of course. Standar size header will remain in the design.

[The Touch]

looks great Dean. Keep us updated with that export info ;)

[rbeall]

bad ass bro. I made some money off of this move. I'll most likely be looking to buy one of the new systems from you!

-Ryan

[dmgoedde]

bad ass bro. I made some money off of this move. I'll most likely be looking to buy one of the new systems from you!

-RyanWhat move are you talking about?

[borneobear]

[QUOTE=Keep us updated with that export info ;)[/QUOTE]

Still no news for international buyers? :(

[Jack Crossfire]

He tacitly abandonned the IDG500 & the flat IMU concept. Well, no need to replace our 3 year old IDG300's. Wonder if Procerus patented the super compact ADXRS arrangement & that's why everyone else is building these sort of but not quite Procerus arrangements.

[airmcn_3]

Still no news for international buyers? :(


A pre-order list is being drawn up as we speak; we will have this available to add your name to by next week. Prices and Export information will follow.

The cost is not set in stone just yet so there will be no price quote until talks are completed.

We will not crud up this thread with pre-order, prices and Export information so expect a new thread.

Thank you,

Chris

[2jers]

A pre-order list is being drawn up as we speak; we will have this available to add your name to by next week. Prices and Export information will follow.

The cost is not set in stone just yet so there will be no price quote until talks are completed.

We will not crud up this thread with pre-order, prices and Export information so expect a new thread.

Thank you,

Chris

We thank you i advance... :) :D :)

[rbeall]

Analog devices doesn't make kestrels accelerometer anymore....The gyro choice is a no brainer.....they are bad a$$.

Military moved me to TX

[rbeall]

Dean here is a matlab plot of that track smoothing we had talked about. The circles should all three be touching eachother but due to a roundoff on my GPS and angle calculations I could only get close....plus despite what people say....the earth isn't flat. Ha The x inbetween the two tracklines is the aircraft position, the line coming out of it is the calculated heading. The aircraft is setup (in this example) to fly towards the vertex from the bottom trackline to the top trackline making a right hand turn. Each of the circles represents the orbit radius for a orbit hold control law.

Hope this helps. When you get the 3.0 hardware all settled out and venture towards some of the fancies of nav etc. Hit me up I'll walk you through it

-Flight school starts monday

[sim559]

I have found that most of my problems are had by everyone else too so I am not a complete idiot ;-)

;-)
Thanks very much for that post.
I think the big problem is in working with different vendors and the vendors are not working together. As a result all that try to make this work end up spending too much time reinventing the wheel.
From what you have posted, it appears that no one has a good working unit.
Sim

[mboland]

Thanks very much for that post.
I think the big problem is in working with different vendors and the vendors are not working together. As a result all that try to make this work end up spending too much time reinventing the wheel.
From what you have posted, it appears that no one has a good working unit.
Sim

I too thank you for the post.
I have been beating my head against the wall with Ardu for some time.

Same answers to queries that you got "well it works for me"

I have purchased 2 boards, the old one and the new one with the shield board, and although I can get it to stabalise, it won't navigate.

My only option was to rewrite the software, which is why I have been sweating on getting the Atto.
I have thrown my hat in the ring for the Australian Outback UAV Challenge, and frankly, without Atto, it would not be a goer for this year.

Keep up the great work Dean, you have quite a fan club out here.

[sim559]

I too thank you for the post.
I have been beating my head against the wall with Ardu for some time.

Same answers to queries that you got "well it works for me"

I have purchased 2 boards, the old one and the new one with the shield board, and although I can get it to stabalise, it won't navigate.

My only option was to rewrite the software, which is why I have been sweating on getting the Atto.
I have thrown my hat in the ring for the Australian Outback UAV Challenge, and frankly, without Atto, it would not be a goer for this year.

Keep up the great work Dean, you have quite a fan club out here.
hope you can get an Atto in time. As you will not have a problem winning the event with Atto.
When does the happen?
Sim

[dmgoedde]

He tacitly abandonned the IDG500 & the flat IMU concept. Well, no need to replace our 3 year old IDG300's. Wonder if Procerus patented the super compact ADXRS arrangement & that's why everyone else is building these sort of but not quite Procerus arrangements.Jack... you are the only person I know of that uses the word "tacit"... did you play an instrument? Ever been to band camp and had a "and one time, at band camp, I put my IMU in my..." experience?? "Tacit" as I'm sure you know means to be silent and is used on sheet music scores.

I was 100% on track with the flat IMU, but really like the ADXRS610 after exploring it in a 3-axis system, so I decided to adopt it. I do like how the Kestrel has them tightly clustered on a corner. Maybe that influenced me.

[mboland]

hope you can get an Atto in time. As you will not have a problem winning the event with Atto.
When does the happen?
Sim

I too am hoping the Atto gets here in time, everything is crossed and I pray to the gods daily.

The AOC is being held on the 28th September, but we do have to lodge a technical overview of our system by the 28th of August.

Times a runnin out real fast :-O

[biloxthecat]

I too am hoping the Atto gets here in time, everything is crossed and I pray to the gods daily.

The AOC is being held on the 28th September, but we do have to lodge a technical overview of our system by the 28th of August.

Times a runnin out real fast :-O

Seems like a few of us are in exactly the same boat! :eek:

Luck for us we have a plan B!

[Gary Mortimer]

What a lovely picture Biloxthecat, my boys would be asking why they could'nt fly in the plane at that age. Is that Balsa overcast?? The big Telemaster.

[kbosak]

Jack... you are the only person I know of that uses the word "tacit"...
It reminds me 'Tacit Blue, a nice object for uavisation.
BTW What is the correct etymology of Tacit Blue, Swift and Sad?
"Northrop engineer Norm Cashen was quoted in 1996 as having said,
"You're talking about an aircraft that at the time was arguably the most unstable aircraft man had ever flown."[1]"

[biloxthecat]

What a lovely picture Biloxthecat, my boys would be asking why they could'nt fly in the plane at that age. Is that Balsa overcast?? The big Telemaster.

If you have been talking to Tom Hunt it could be! :cool:

[saabguyspg]

I have been reading the manual and some posts in other groups... man this thing is truely powerful.

lol... I have no other comments than that.... Dean is working away on mine and I am building the plane for it as I wait.

Are there any Canadian atto users on this forum? I went to the Atto ning page and read a lot of Gary's posts.

[brakar]

saabyguyspg, btw, know if Gary eaten (lion) lately? brakar (pub) :cool:

[jtprouty]

Has anyone used the XBee Pro with Atto yet? I'm having some self induced issues that I need help with.....

Jimmy

[airmcn_3]

Has anyone used the XBee Pro with Atto yet? I'm having some self induced issues that I need help with.....

Jimmy


Jimmy,

Dean and I are working on one right now. I am using this on our UAV. We will be working on it at the beginning of next week. If you have not figured out your issue by then I may be of more help.

What kind of issues you having?

Chris