Discussion forum for UAVDevBoard, Sparkfun's 3-axis IMU-based autopilot board

This is the place to discuss the new open source IMU-based autopilot board by Bill Premerlani.

Views: 14888

Reply to This

Replies to This Discussion

Hi Chris,

I use the Turnigy Plush ESCs (18A and 30A) with the UDB, so I know they will work OK. You might be on the very edge of the throttle range.

When the ESC doesn't arm, are you getting a constant beep-beep-beep-beep... from the motor? If so, you might try setting the range. For the Turnigy Plush, you set the throttle to full, turn on the power, and expect to hear two beeps from the motor. At that point, lower the throttle to minimum, and you should get the normal start up beeps.

Good luck,
Rusty
Rusty

thanks for that, seems the UDB must have a slight difference in pulse output, recal'd the ESC and it seems to work odd it worked fine direct from the Rx.


Bill thanks for the tip on the servo swap, I feel a bit stupid not trying that first

I'll have to re edit the options.h now as the throttle goes full on in stabilise, need to edit the throttle output reversing bit

regards and thanks

Chris
Hi Chris,

Glad that worked. The UDB outputs pulses at a slower rate than the receiver does, so maybe the ESC takes that into consideration somehow. You'd really have to put a scope on the throttle signal going to the ESC to see what's different about the two.

As for the full throttle in stabilized mode, I'm betting you have altitude hold enabled. The normal behavior for the UDB is to go to full throttle to climb to altitude, so that's probably what you're seeing. If the throttle is nearly closed, the motor should stay off, but if you open it slightly, it will go to full throttle until it can climb to the desired altitude.

Cheers,
Rusty (being soundly thrashed by Eagle CAD)
Hi Chris,
The behavior of the throttle in stablized mode will depend on what options you select. There is an option to apply altitude hold in stabilized mode. If you pick that one, the throttle will kick way on when you switch to stablized mode on the ground, because the controls are trying to reach altitude.
For your first few flights, I suggest that you turn altitude hold off, until you get the feel for how stabilized mode works, and you get the gains set the way that you want. In fact, if you are not using altitude hold, you can connect your ESC directly to the Rx.
Best regards,
Bill
Hi Rusty,

The UDB output pulses are at 40 Hz. That is a legacy from my initial work 6 years ago based on some testing that I did. At the time, I had fried a few servos by trying to run them at high pulse rates. I did some testing, and backed off from the point where the H bridge in the servoe shorts out, which actually is pretty close to 50 Hz. I did not realize that most Tx's generate pulses at 50 Hz. Most servos and ESCs will work fine at 40 Hz. I suppose we should raise the pulse rate to 50 Hz, it would not be that hard to do. I'll put it on my "list".

I hope that you get the upper hand with Eagle CAD. ;-)

Bill
Bill, Rusty

thanks for all the help guys, I re-read the pdf's and worked out its the altitude hold and have now set it to off

Just got to be patient and wait for a calm day

regards

Chris
Hi Bill

I am new to this forum and I am trying to read up on this. So please excuse me if this has been asked before.

For a "DevBoard3" in think it would be easy to replace the three LISY300AL gyros with a LPR530AL (for pitch and roll) and a LY530AL (for yaw). These devices have almost identical specifications to the LISY, so there is no need to make a special SW for that.
So my question is, why would you prefer the Invensense gyros? Do they have less drift or lower noise, or do we need the high angular rate?

Since the sensitivity of the Invensense gyros (500deg/sec) is 40% lower than the LISY, do you expect to have more noise(thermal- and quantization noise) on the pitch/roll/yaw angles?

Thanks,
Quist
Hi Karsten,

Your questions are very timely.

The team recently did some aggressive fixed wing and heli flight testing of prototype boards with the Invensense 500 degree/second gyros. They performed beautifully. No thermal or quantization noise issues. So, our plan is to use the Invensense gyros on the next generation of our board.

There were three reasons for wanting to change gyros. The LISY gyros are vulnerable to vibration, so they are really not suited for helicopter applications. They are also vulnerable to electromagnetic interferance, if they are located too close to a transmitter, as might be used for video or telemetry. Finally, there are pilots who want to do acrobatics with rotation rates that exceed 300 degrees/second.

So, we considered several types of gyros and did some tests on some of them, and we have been following the reports of tests that others have been doing. In my opinion, the Invensense 500 degree/second gyros are the best ones available for both aircraft and heli application. Our tests showed the following:

1. The Invensense gyros were not affected by heli vibration.
2. They are immune to RF interference.
3. They peformed well in flight.

By the way, from the point of view of the software, it is very easy to swap gyros. Besides the two different gyros used in production, there are 3 different prototypes sitting on my desk, for a total of 5 different gyros. The same software will run with any of the gyros, there is a short "configuration file" that configures the gain, channel assignments, and relative polarity.

Best regards,
Bill
Hi Bill

Thank you for the elaborate explanation! The Invensense gyro sound like a much better choice for this application. I cant wait for the new board to come.

Regarding the gyro sensitivity, I didn't realize it was that easy to change the polarity and sensitivity in SW. I like when people write flexible SW!

Du you have any numbers for the resulting angle noise, with the different gyros? By numbers I mean how many degrees RMS of noise. (I am just a little curious :-))

I do not know how much influence you have on the PCB layout on the v3 board, but here are some just a few ideas that came to my mind:
- Mayby we can put a footprint for a RF shield around the sensors. It does not cost anything and it gives the posibility to add a RF shield later if we see an issue. (that has saved my a** before :-))
- An other thought was to change the footprint for the Debug UART port to fit the Openlog module(from Sparkfun).

Best regards,
Quist
Billu,

Would you be so kind as to answer a few questions regarding the green UDB.
I would like to configure this board to fly a 3 cannel ( elevator, rudder, throttle ) polyheral glider

I have aquired a pickit2 programmer and intalled it on the laptop together with the MPLAB IDE 8.3

which program would you reccomend, matrixpilot?

If matrixpilot is used, what do I do with the aileron output as I can't seem to find any setup procedure for this particular application

Also can the ublox GPS be used with the old green board

Kind regards
Jonny
Hi Jonny,

Yes, I recommend MatrixPilot 2.0.3.

In order to set it up, you select settings, options, and gains in the file, options.h. Instructions are in the file. More detailed instructions are available under the Wiki tab of the project web page.

The default settings in options.h are for a plane with ailerons. To use rudder and elevator instead, turn off all of the aileron options and turn on the rudder options. For rudder, elevator, and throttle, you will use 4 inputs and 3 outputs, so set the number of input and output channels accordingly. For any control surface that you do not have, set the channel to CHANNEL_UNUSED. Be sure to set up the servo reversing options, either with the switches, or hardcoded in software.

All of the firmware, including MatrixPilot, will run on either the green board, or the red board. The uBlox GPS should work with the green board.

If you run into any problems, I suggest you post any guestions to the google group, uavdevboard.

Best regards,
Billu Bhaiya
Hi Bill

Many thanks for the quick reply, if I had been less lazy and just opened the folder! I found the optionns.h file and have started to configure for my airframe, everything seems to be working quite well so far but I expect I shall have a harder time when I progress to waypoints and trimming. I have made quite a trick isolation mount to install the board, this will then be permanently mounted inside the fuselage so I think I will need an extention cable to plug in the programmer externally for making alterations and uploading waypoints.

I was wondering how the IMU will handle bad weather? particularly wind!


Kind regards
Jonny

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service