Micro Motor Community

NotFastEnuf E011 / Bwhoop Silverware Fork


#1610

#1611

Hmm so I flashed the latest git version with your experimental angle changes however I noticed something (properly unrelated to those pid changes) for the first time.

My RSSI doesn’t go over 20 on my betafpv lite FC whoop.
I flashed the same hex on my e011c fc and it works fine there.
Swapped antennas too to make sure thats not the problem and same thing.

Did I damage the betafpv board somehow or is it scaling the RSSI funky and I just need to change the scaling in the telemetry? Any ideas?

Hmm I flashed back to an older hex and its 20 there too. I am not sure if it was always like that or I just started to notice.


#1612

Maybe swap out the crystal?


#1613

Test your angle pid stuff now and ran through 16 lipos.
Super great - I like it a lot :smiley:

I did reduce the small error values a bit to my liking tho:

float apidkp2[APIDNUMBER] = { 6.00 }; // Kp ROLL + PITCH
float apidkd2[APIDNUMBER] = { 3.0 }; // Kd ROLL + PITCH


#1614

Cool man! It’s gonna be hard to pick defaults. I’m gathering as much data as I can on it from testers. Your info is very helpful.

Preferences are ranging from 13/5 over 7/2 to 10/3 over 5/0 … Down to yours. This will be tricky


#1615

Ah cool to know, i’ll give some more values a try.

I was also testing the USE_ANALOG_AUX feature which seemed to work super fine.
Yesterday I re-read the pull-request and noticed the issues about CPU utilization when running beta filtering.
I did not notice such issue but I was only running the analog aux for rates.

Am I right to assume that the cpu load increase is likely still in the boundaries of acceptable when enabling only one of the 2 analog aux switches?

Edit: Upon checking the code some more it doesn’t look like it would save much cpu usage. However the code could be optimized to skip some stuff by not only making it dependant on #ifdef USE_ANALOG_AUX but the respective rates/angle defines as well.

I’ll probably re-flash with cpu load telemetry debugging these days to confirm


#1616

If you want to do a quick check you can just define the debug option in config.h … I added CPU load to the debug struct. Just be sure to take props off and arm as well as switch through the flight modes you use when running the debugger. Horizon mode is the most load intense followed by angle and then acro. Maybe move it around a little while the motors run. That will give you a good estimate of CPU load in flight.


#1617

Headsup: 5$ bwhoop b03 fcs (without baro) here http://www.boldclash.com/boldclash-bwhoop-b-03-receiver-p54.html?fbclid=IwAR3swcHLEhOC6sQLClaL68UVR-efD5zw4yqdxU4nLzfgZkUESGMLLNplIHo


#1618

wtf - now i am getting it on a second whoop too (e011c fc this time). replaced crystal, replaced antenna but still rssi around 23. When I plug in my third whoop i get 98-100% tho.

I have no idea why it was working fine one battery ago, then i switched quad flew one battery and then tried again the other whoop and suddenly 20ish rssi max problem :open_mouth:

edit: it appears like it makes no difference if an antenna is connected or not


#1619

Really low rssi it typical of antenna problems, range is about 10m with no antenna. Check for anything amiss near the antenna point.


#1620

Yeah I think the trace to the antenna pad or something may have been affected by the frame I am using (carbon/nylon).
I can’t find my multimeter (lost in vacation?) so need to wait for tomorrow for my new one to check on the working board where it should connect to.


#1621

Someone posted a makefile for building it with armcc under linux on git.

I added a comment with instructions how to install armcc with the mdk_lite native license in the comments: https://github.com/NotFastEnuf/NFE_Silverware/pull/9


#1622

There’s some schematics here, might be of help, along with the new multimeter.


#1623

I was gonna say too that maybe the filter cap is knocked off the antenna circuit


#1624

Awesome thank you! I was looking for some schematics! That will definitely help


#1625

Filter cap is still on there but I think the cap is damaged somehow possibly by my frame as there is the cam mount that runs close to it.
I tried heating it up on both sides with the soldering iron but didn’t change anything.

I don’t get any continuity whatsover on the one right next to the antenna at the top of the fc.

I also figured it might be possible to just attach the antenna directly to the XN29L leg and skip the cap but that didn’t help change anything either - neither did removing the cap completely (I figured some noise may not be filtered but it should still reflect somehow on the rssi).


#1626

Maybe a bad xn297 or the caps that sit on the sides of the crystal. That diagram from Ian outlines the whole circuit. You could just gut it all and rebuild it for science.


#1627

The crystal itself has been known to fail too. Toy tx is a source of crystal.


#1628

already replaced the crystal and that’s not it.
Could try replacing the xn297 but I’d rather not scrifice a working board for checking if it works.

I’ll might just chalk them this 2 fcs up as lost or as a future repair project


#1629

Could be an excuse to get a hot air station… That way nothing gets sacraficed and stuff just gets moved around.