this is a follow up on the glitch issue observed by Marcus on his HoTT PPM Setup
Marcus and I have spent some tough weeks investigation this nasty problem and found a solution now
.We did a lot of testing and now its time to share it with you.
We hope this contribution helps and your feedback/discussion is welcome.
Actually Marcus uses Graupner HoTT GR-16 and GR-24 receivers in PPM mode. He observed glitches on the throttle and the roll channel. But it was not clear why it happend sporadic, and why it happend just on one copter, where as another copter with same components did not show any problem.
This a is screenshot of a typical log, with some glitches on channels 0 and 1
We could reproduce the problem with the following special setup.
- Hott-GR-16-24GHz-Interference-Worst-Case-Setup-B.jpg (202.39 KiB) Viewed 16780 times
What's going on?
In fact the backchannel of the GR-16 Receiver is an RF transmitter. If the cabling of the PPM signal is not appropriate, the PPM wire acts as a receive antenna. The PPM signals gets interfered by an additional noise signal, which causes false timing measuremts at the AQ side.
Above picture shows a worst case setup, just for demonstration and testing. The PPM signal wire is appart from its ground return wire, which makes it sensitive for interference. And the 2.4GHz antennas between make it even worse.
Please not that the HoTT receiver works perfect all the time, just the output PPM signal is interfered.
Basically the described effect is a general one. But actually we could observe it with HoTT receivers GR-16/GR-24 only, with another product like Jeti SAT2 we could not reproduce something similiar. Probably the timing relationship between backchannel to ppm edges is different.
What's the solution?
(1st) apparently, care for good PWM signal cabling. Either use twisted pair cables or use standard servo cables. The important thing is, the distance between the signal wire and the ground wire must be close. The cabling should be as short as possible as well.
- Hott-GR-16-24GHz-Good-Cable.png (1.47 MiB) Viewed 16780 times
(2nd) Addtionally an update of the AQ ppm code is recommended. Due to a minor issue in the ppm.c code, occasional bad ppm signals can cause the observed glitches. We made a fix and a lot of testruns already, showing that even heavy ppm noise will not cause any spikes.
I will come up with further updates soon
regards
Alois
