Replies to This Discussion

Permalink Reply by Uwe Gartmann on November 24, 2010 at 1:26pm

I use a 8 channel optocoupler device to separate APM and receiver from the other gear. The ESC delivers power for Servos. The APM is powered by a 2S lipo and a standalone UBEC from Castle Creation. I would never connect Servos directly to the APM because servos can make much noise on the power line. Remember, there sensitive sensor on the IMU. Stable power means more stable values from this devices.

• ▶ Reply

Permalink Reply by Bill Noyce on November 24, 2010 at 4:51pm

Interesting, i have seen other people be concerned about noise down the ESC data cable. One suggestion which i am considering is to use ferrite rings to suppress the noise, but that is not really what i was refering to in the original post. My problem is random activity from OUT0 of the mux during code upload.

• ▶ Reply

3D RoboticsPermalink Reply by Chris Anderson on November 24, 2010 at 5:15pm

I've never seen that. I've uploaded code hundreds of times on dozens of APMs, and never had a prop spin up. If the servos are twitching during the upload, I've never noticed it. Are you using analog or digital servos?

• ▶ Reply

Permalink Reply by Bill Noyce on November 25, 2010 at 2:44pm

Interesting if it is just my APM which is doing this. With the fixed wing UAV it was just a plain analog servo on OUT0 but the rogue PWM signals would stress the servo beyond the normal range of movement. The Arducopter code obviously has an ESC on OUT0 so the outcome is more elaborate! I am now used to holding down the Quad when uploading code. Next time it happens i will video it.

• ▶ Reply

Permalink Reply by Darren on November 25, 2010 at 3:44pm

Bill, I too get the jumping around etc when doing that, but only when powering my apm from my older laptop, i dont get it when powering it from my new laptop. i think its a power issue. it could also be a symptom of the brown-out fuse setting issue reported here... but i havent tried the pwm fix for that.

• ▶ Reply

Permalink Reply by Mitch Robbins on November 26, 2010 at 7:20am

I haven't had it happen on APM but have had it on several other custom and commercial boards. I simply assume servo and esc movement will be unpredictable and plan accordingly. A quad or heli inadvertently powering up on the bench or in your hand will definitely get your attention and make you think about it for next time!

• ▶ Reply

DeveloperPermalink Reply by Doug Weibel on November 26, 2010 at 7:47pm

There is a way around this issue: If you have your transmitter turned on, are using channel 8 for mode selection, and have the transmitter configured to give you a PWM pulsewidth > 1750 (IE you are in "hardware" manual mode), then channels 1-4 will not be affected by the undefined state of the main processor while reprogramming. You could still get some glitching on channels 5-8 outputs if you are using those....

• ▶ Reply

DeveloperPermalink Reply by Mark Colwell on November 27, 2010 at 1:01pm

This is happening for my APM beta, I have both digital and analog servos powered through USB port, I just thought it was undefined states during upload as I have no interference. I disconnect all servos when loading programs. ( I have stripped gears when testing with old AP). I am using a servo tester to drive Ch-7 for Mode control. I will be testing Turnigy 9x system soon so I can select any of 6 modes vs 3. and slew pan & tilt too.

• ▶ Reply