search

iNavFlight / inav

INAV: Navigation-enabled flight control software
https://inavflight.github.io
GNU General Public License v3.0
3.12k stars 1.47k forks source link

SP Racing F4 EVO consistently erratic GPS hold until powered cycled #2687

Closed VoicOfReason closed 6 years ago

VoicOfReason commented 6 years ago

I ‘m not sure if this is an iNav issue/bug, but I recently switched from the Omnibus F4 board to a SP Racing F4 EVO because the EVO has more UARTs. The Omnibus hovered stable in with GPS modes on every flight even when satellite count was fairly low. But, with identical hardware and settings, my quad is erratic during GPS hover. Sometimes when I lift off it hovers just fine. Other times, it swill start to drift badly, over correct, and start heading for the ground before I turn off GPS. Today, I let is acquire 17 satellites and it all worked perfectly for 25 minutes of flight time including disconnecting battery several times to test. So, does it sound like a lack of # of satellites (before it has time to acquire more), or a bug, since my hardware is identical with Omnibus setup?

Here's my setup: iNav 1.8 Sp Racing F4 Evo BN-880 GPS, UBLOX-7 setting 3 UARTS set to MSP (1 for USB, 1 for DJI Ocusync, 1 for Mobile Flight telemetry via TBS Crossfire Micro v2), 1 UART for serial Rx TBS Crossfire Micro V2

Compass calibrated multiple times, orientation correct

GPS/compass is isolated from rest of the quad really well. No other antennas nearby and copper tape beneath it to shield.

krzysztofmatula commented 6 years ago

Just to rule out simple things - the mag is operating well, has correct orientation set and is calibrated, right?

VoicOfReason commented 6 years ago

Yes, east is 90 degrees, and I recalibrated Twice. But of course it also flies very well after seemingly a longer than normal warmup. If the mag wasn’t calibrated I would guess that it wouldn’t occasionally fly well by accident.

On Thu, Jan 25, 2018 at 6:21 AM krzysztofmatula notifications@github.com wrote:

Just to rule out simple things - the mag is operating well, has correct orientation set and is calibrated, right?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/iNavFlight/inav/issues/2687#issuecomment-360451199, or mute the thread https://github.com/notifications/unsubscribe-auth/AAq5OciONl2RlAzpAed1MhToO1Lzom7_ks5tOHG7gaJpZM4RsPrt .

VoicOfReason commented 6 years ago

FLIGHT #2: The same problem occurred after 17 satellite locks today so I landed and noticed that in Mobile Flight (iPhone connected via TBS Crossfire) East was no longer 90 degrees. So, I re-calibrated the compass and now GPS works fine again. So, it appears that every time I fly I have to re-calibrate my MAG. I never had to do this before.

Is having to calibrate for every flight normal for some people?

I have copper tape beneath the compass for shielding and it is very far away from all other antennas and FC.

VoicOfReason commented 6 years ago

Is it possible the FC isn't saving the MAG calibration settings long-term?

shellixyz commented 6 years ago

It's probably not the issue. You can check the mag calibration it is magzero_x, magzero_y and magzero_z.

VoicOfReason commented 6 years ago

FLIGHT #3: I flew today. East was 90 degrees, so compass was fine, I'm guessing, since I also calibrated again. I waited for 17 satellites and took off. Turning on GPS hold, it started flying off at an angle into the sunset. My satellite count stayed high, and I flew to some other locations and kept trying GPS hold but it acted erratic each time. So, I landed, unplugged the battery and waited 5 seconds. Then, I plugged it back in, waited 10 seconds until it was ready to fly again and took off. From this point onward, it hovered with GPS on perfectly, in 40mph winds. So, this seems like a software issue to me. If it was hardware it wouldn't SOMETIMES work. And, it now seems to be fixable by power cycling my drone.

Any thoughts/ideas? This is a serious problem.

VoicOfReason commented 6 years ago

FLIGHT #4: Went flying again later today after replacing the GPS with a new, different GPS. Calibrated compass. Did the same thing: flew erratic in GPS hold mode with 17 sats until I power cycled and then it flew perfect.

VoicOfReason commented 6 years ago

FLIGHT #5: Waited 20 minutes and flew again after changing baud rate of GPS to 115k. Just as before, GPS hold was erratic at first just like all the other times. I unplugged power for 2 seconds and reconnected and GPS hold was still erratic. So, I landed and unplugged power for 10 seconds. That fixed it. So, it appears that removing power from the system for 10 seconds after initial startup fixes the problem and then it flies as it should. What's going on?

VoicOfReason commented 6 years ago

FLIGHT #6: Flew today after removing the copper foil beneath the GPS (maybe causing problems??). It behaved exactly as before. It quickly locked onto over 12 satellites. I took off, turned on GPS hold and it started banking and drifting hard like it I was pushing the the pitch forward. I landed and unplugged for 10 seconds, turned back on, flew and turned on GPS hold immediately and it held position perfectly from that point onward. Nothing else changed.

I'm waiting about 20 minutes between flights before the issue returns.

shellixyz commented 6 years ago

It would probably help to analyze the backbox logs. The mag values, GPS coordinates and number of sats are all recorded.

VoicOfReason commented 6 years ago

OK. I'll install an SD card and report back soon.

VoicOfReason commented 6 years ago

Trying to setup SD card blackbox logging. Formatted card with SD Card Formatter but can't get past "fatal SD card error". Anyways....

FLIGHT #7: Today I flew with 17 satellites. It hovered perfectly and did a successful RTH. Then, I flew over to the other side of my yard 15 feet away and turned GPS hold on again. This time it flew up into the air full speed. I quickly turned off GPS hold and descended. I flew out a little ways and hit RTH and it just shot straight up into the air again full throttle. I turned off RTH and came back and did a test hover. I noticed that when I pointed straight North, it hovered stable but when I starting turning the quad East, it show up into the air at full throttle. So, this is all baffling to me. Like I said everything was working perfectly with the same hardware before I changed flight controllers. Same hardware.

VoicOfReason commented 6 years ago

FLIGHT #8: Quad has become uncontrollable in GPS modes. It sometimes hovers fine if pointing any direction but straight East. As soon as it points East, it shoots into the sky like a rocket on its own. This includes GPS alt/pos hold modes as well as RTH.

Still trying to work out SD card issue so can't upload blackbox yet...

shellixyz commented 6 years ago

You don't need special software for formatting the SD card. Did you try with the windows formatting utility ?

krzysztofmatula commented 6 years ago

"Shooting into the sky" may be caused by accelerometer resonance caused by hard-mounting the controller to the frame (or bad damping). It is clearly visible in the log.

VoicOfReason commented 6 years ago

The FC is soft mounted and so are all the motors...

I can't view the log yet. I just got it to finally write to the SD card but it only recorded 5 seconds of data and rest of the file is garbage.

VoicOfReason commented 6 years ago

I've been trying to get a fully working iNav quad working since the summer of 2016. If isn't one thing then it's another. Went outside to finally do a flight with blackbox log recording and now after absolutely no changes to my quad it wobbles out of control (regardless of GPS mode) and falls to the ground........... sigh......

VoicOfReason commented 6 years ago

OK! Here's the first blackbox log from today. The behavior here is not as terribly erratic as it has been, but when I turned the quad it did start to drift hard and fly off in another direction. I will post more logs this week...

LOG00002.TXT

VoicOfReason commented 6 years ago

Is it possible the having 2 MSP ports functioning at once is taxing the FC and causing erratic behavior? I have 1 MSP for DJI Goggles RE (just reads voltage) and another MSP for Mobile Flight. Anyone?

shellixyz commented 6 years ago

Probably not. And with the F4 MCU even more unlikely. Anyway you can check the CPU usage in the configurator.

VoicOfReason commented 6 years ago

Yeah, it's only at 15% CPU. I flew again yesterday and tried to do a log but suddenly it's flying perfectly. I don't get it. Here's the erratic behavior summary.

Sometimes:

  1. It flies perfect
  2. It shoots to the sky when facing East with GPS hold on
  3. It drifts hard and starts to fly away with GPS hold on
  4. It wobbles uncontrollably when blackbox refresh rate is set below 100%

I haven't been able to get the erratic behavior on blackbox yet. It's like trying to photograph Big Foot.

Acanevaex commented 6 years ago

Can you use Mission Planner on a phone connected to quad? You could see if gps position on the map is drifting. Im having position hold issues as well. My 450 with SpF4 Neo and naza proto m8n will hold well for 5 or 10 seconds then start doing circles gradually building speed and spiraling wider and wider out. I am within half mile of cell tower so that may be a factor. But try the mission planner or convert your log into the google map gps output. Im still trying to figure that out.

stale[bot] commented 6 years ago

This issue / pull request has been automatically marked as stale because it has not had any activity in 60 days. The resources of the INAV team are limited, and so we are asking for your help. This issue / pull request will be closed if no further activity occurs within two weeks.

stale[bot] commented 6 years ago

Automatically closing as inactive.