Spikes in radio channels (stable release)

[Cookiemonsta]

Hi,

after updating my Frsky D8r-XP receiver to 27ms firmware I get some very large spikes in two channels (rudder and pitch).
Did anyone else observe such a behaviour or does someone know any cure for that?

In the attached pic are only spikes in the rudder channel visible and the spikes seem to be independent of the radio quality.
I use the stable release 6.6 btw.

Thanks in advance,

cookie

Spikes in Rudder channel

EDIT: I made some investigations and logged for some time and realized that the spikes are always for 10 datasets at -1024 . Does anybody have a cure for that? I already removed the gnd lines of the esc connections (vcc has never been connected)

[Cookiemonsta]

I also tested it with a non upgraded receiver and the problem persisted. I recently rebuild my hexa to a y6 on basis of the flyduino FCP HL XL and played a little with the placement of the receiver but that didn't help. Furthermore does the radio quality rise and fall between 90 and 100 when using the flashed receiver and solid 100 unflashed. Maybe I should try the beta if there's a solution to that. Or can somebody make a hotfix for that so that no values under -700 are accepted and the last received values are used?

Any ideas are welcome

EDIT:
I flashed the beta version but that didn't help. I also swapped the "problematic" channels with non problematic but the spikes persist. really odd...
Please help

[Cookiemonsta]

Nobody got an idea?

I even changed the cables but that didn't help

[chschmid]

Maybe something related to that:



Cheers
Christof

[Andrej]

Same thing with Graupner Hott receiver. All leads are good separated.
Seems its problem with all PPM radio connection.

[Max]

OK, sorry for the slow response, but I had to verify this. There are bugs in both 6.6.0 and 6.7r100 which can cause this problem when the PPM signal contains invalid (out of range) pulses. The code actually changed between those 2 versions, but the problem manifests itself in a different manner in the newer code (should be better, but still not great).

The problem is fixed as of r130. I recommend building that version or later, or making a change in ppm.c in the older versions. You can see the difference in line 114:

https://code.google.com/p/autoquad/sour ... c?r=95#114

https://code.google.com/p/autoquad/sour ... ?r=130#114

In tags/6.6.0/onboard version you could remove line 109:

https://code.google.com/p/autoquad/sour ... /ppm.c#109


But the problem ultimately stems from some kind of interference in the PPM signal. I have one AQ board which also has poor quality with PPM input, also with FrSky Rxs (several different ones with either Rx firmware version). Seems to be something related to the ground circuit on the AQ, but my troubleshooting so far has been inconclusive. My LA captures always come up clean, and the same Rxs work fine on another AQ, so it's a mystery at this point. It still flies fine and I don't notice any problems in the air, but the PPM code needs to be fixed as I describe above.

I would try to isolate the noise issue by disconnecting everything else from the AQ board, using a very short cable to the Rx, and making a quick log. Also make sure it's not grounding against a standoff or screw, or something like that. If that fixes the quality then you can add connections back one at a time and see if you find the problem. You could also try powering the Rx from a BEC (keep ground and signal to the AQ), though that didn't help in my case. If none of that fixes it, as seems to be my case, then I'm not sure what to do next

-Max

[Cookiemonsta]

I disconnected everything but the rx and the xbee and the problems persisted. Can you compile a version of the stable release 6.6.0 where every channel is checked for invalid (-1024) values and if so replace them with the last valid value?

[Max]

I'd especially disconnect the XBee for the test.

The problem is fixed in latest code and doesn't occur to begin with a good PPM signal.

-Max

[Andrej]

Is it possible to get compiled version? ... and with Hott code?

or link to "how to" install soft for compiling and compiling ?

[Max]

Andrej, do you actually read anything in the docs or the forums? Mav2Hott has nothing to do with the firmware, it's a standalone project. And you certainly must have overlooked the in the docs. Your posts make it seem like you can't be bothered with any of that reading/searching stuff.

-Max