Wookong-M GPS Hold Broken? Troubleshoot?

[stj5353]

Hi Guys,

Because you may want to know, first, my setup:
Tarot 100MM Hex
DJI WKM - 5.26 Firm (Latest)
DJI IOSDII - 3.01 Firm (Latest)
~14 Lbs
Futaba 14SG
10K mAh 6S
370 Kv Motors x6
1655 t-style props, perfect balance on each


I'm having horrible luck with my new Hex. I bought a 1000MM Tarot hex frame. Love it, great quality, rigid, light. Very well constructed. I thought I would have no problems mating it with a DJI Wookong-M. I didn't want to go for the latest/greatest A2 since I thought the WKM-M would have more miles on it and more bugs worked out. But every time I've taken it out I've had troubles and rough landings. Today I attempted to start from a clean slate. I meticulously balanced all of my props, both wing and hub balance. They are all balanced perfectly, so they should not be causing any vibrations over and above the normal. Here are some important points I want to call out:

* I mention prop balancing because I take seriously the quality of the flight is highly affected by vibrations. Each of my props will hold any position from 0deg to 180deg on the balancer you put it in.

* My IMU is mounted center-mass, and I've calculated the offset from COG as best I can. I have double checked the math and used negative values for the offsets when necessary. I'm 99.99% sure I've got the IMU configured correctly and mounted properly, via velcro double-sided tape. This isolates it from any high-frequency vibrations and still holds it very firmly from moving in flight.

* The GPS/Compass is mounted properly in the right orientation. It has no interference and calibrates successfully, which I've repeated on a couple occasions.

* I power it on and let it get GPS lock +6 sattellites every time I've ever flown. It will sit idle for a couple minutes because I want it knowing home point.

* The LED flashes purple once I'm ready for flight, and it stays purple flashes for the best of my knowledge. I've never seen it flash any different color, like white during flight, although I couldn't tell you exactly the pattern.

* All of my gains are about 150% give or take 5% or so.

* My hex is heavy, its about 15lbs loaded. I'm flying a test load now, luckily, not my actual camera and gimbal, but I attached below the frame and balanced COG as best I can mirroring what the camera/gimbal load will be.

When I fly, I take off in Atti mode. Altitude hold is not perfect, but it's not far off either. My altitude gain is about 140%. I haven't played with this gain yet as I'm still trying to figure out GPS hold and altitude hold is enough to keep it within 1M or so.

I will switch into GPS mode and it will drift seemingly exactly like Atti does. I can't seem to find any delta in hovering between atti and gps. The weather was a bit windy, but not bad. Maybe 3mph gusting to 5mph. It was a steady nice breeze.

I don't know why this is happening. The logs I've looked at in the IOSD viewer really don't help me much. I'm not sure the scale or what the numbers are representing exactly, but I do see the mode toggling between GPS and ATTI, as I would expect representing my input on the controller.

Do you guys have any suggestions as to what I can do to track this issue down? Or what I could look into to resolve? I'm scared shirtless to take off in any breeze since I feel like it's going to walk away from me. I don't want to rely on GPS hold, but I will use features like home-lock which will depend on GPS to govern pitch/roll commands. I have not tried these modes yet, since I have the gains mapped to x2/x3 to adjust in flight.

Thanks guys for any help. I'm feeling very confused and incredibly frustrated. I spent a lot for the WKM and I feel like it's providing none of the values it should.

Thanks again.

 

[ACP]

Do you have video feed setup yet? If so look to see if it is actually going into GPS (Bottom left of the screen). We have a Wkm which constantly drops out GPS back to Atti without any indication on LED.
I have been told there was a bad batch of GPS units which a certain suppler still sells in the UK with this fault on. I'm still trying to have it replaced.

 

[stj5353]

Very interesting. I bought my wkm from aerial media pros, I believe they are in California. I do have video setup but honestly I was not monitoring it, just the led. I was using the iosd flight logs to get the data. I do see it flipping between GPS and attitude, but without the monitor, I couldn't tell you if it was because of me switcing it or if it was on its own. On the same note, I never see GPS satellite counter drop below 10.

I hate even trying to fly this thing now since it has been so unpredictable. But I think you bring up an interesting point and I think I have to go out and see if it is dropping out of GPS frequently without my input. It seems like It's either this, or when in GPS it just fails to do any corrections to position.

I will go out today and do a test flight specifically monitoring the satellite count and flight mode. I will not adjust flight mode after takeoff. Maybe this will shed some light.

Thanks so much for the suggestion.

 

[Tahoe Ed]

First of all. Get rid of the velcro on the IMU. It needs to be hard mounted on the craft using the supplied DJI thin double sided tape. There are internal dampeners in the IMU and you do not need to isolate anything. All you are doing is creating movement that confuses the controller. I think your flying experience will improve tremendously.

 

[stj5353]

Thank you for the suggestion. This seems to be highly controversial. I've read about as many articles saying to rigidly mount it as I've read to mount it isolated from vibration. I think even DJI makes a specific part to mount the imu in a vibraton isolated way on their platforms. My thoughts were to do the same by using velcro. The imu doesn't wiggle or move freely, it is pretty rigidly mounted even though it's on velcro. But maybe it's still worth a shot to try I suppose.

On another note I did fly this afternoon and I noticed in the log one period of time for about 20 seconds it switched out of GPS mode to attitude. I have no explanation for it. Maybe the imu mounting could cause this but I'm not feeling confident this is the root of my problem.

Thanks so much for the suggestions!

 

[ACP]

With the particular system we have, even if we have 10+ sats it still drops out of GPS. If you hover for a while in GPS and it "floats about" for a while then jerks as if it has locked on to something then "floats about again. This is what ours dose.When it jerks this is it going into GPS for a few seconds then drops back to Atti.

NEVER USE VELCRO FOR YOUR IMU

 

[Tahoe Ed]

Actually it is not. I work for DJI and have built many craft over the last 3 years. We always hard mount the IMU. Isolating the IMU from vibration on a separate mount is different from using a thick non - solid mount. However what you do is up to you.

 

[Fat Corgi]

I read this on the Wookong page of the dji wiki...
"The loss of altitude when flying an S800 comes from excessive vibration getting to the IMU (inertial measurement unit). Users should attempt to reduce the vibration of IMU by both reducing the vibration (i.e. prop balancing) as well as dampening the IMU from the vibration (dampening foam, gyro tape, etc)."

 

[ACP]

It is common scene that an IMU would suffer from excessive vibrations. these are sensitive MEMs. You can't expect them to perform perfectly when they are being vibrated.
I have tested this with the A2 as posted in http://www.multirotorforums.com/showthread.php?21630-A2-Problems-Catalogue
the A2 can't cope with certain frequencies and simply gives up.
I wont be doing this test with the WKM as they are the only unity I have left I can semi trust.
I't is unclear whether the A2 IMU has built in anti-vibration but I presume not.
There is also the amplitude of the vibrations to take into account. Dji CAN'T be expected to manufacture a bullet proof system at these prices.
We have to look after the hardware so they look after us.

 

[Fat Corgi]

This is the theory I am working on too. I was going to incorporate one of these in my latest build for the IMU.

In terms of the GPS hold being erratic or non-existent, the only thing you didn't mention in your original post stj5353 is external interference factors. Be they electrical or mineral, I'm always conscious of what things around or below you are present and whether they may affect the performance of the GPS.

 

[baja-king]

Just out of interest I would try downgrading to FW 5.20 as some have had gps issues on FW 5.26...

 

[JLO]

Guys Im flying a CS8 frame with A2 FC I mounted the IMU right in the middle of the copter on a separate smaller CF plate with silicone servo tape, the plate rides on 4 rubber short standoffs I have not have an issue...... yet..... I was wondering if we should do a IMU calibration more often like at the beginning of the flight day just to make sure that the IMU is happy what do you guys think

 

[stj5353]

@Baja, I did downgrade to 5.20 today, I was thinking the same thing and your suggestion made me go ahead and do it. While I lost the 50-waypoint free upgrade, I can say the GPS hold is much better. I was able to adjust gains this time, which I was not able to do before (too much time on the sticks). My gains were much to low on the newer software version. On the older(current), when adjusted properly, they were about 85% increased. Now, was the problem with the newer software just the gains? I'm not sure, probably partly yes, but why was it that the older(current) software allowed me leverage to control the hex and adjust them where the newer release felt just out of control and I never was able to dial it in? I'm not sure the root cause, but I can say that 5.20 is running better. I'm not sure if I'm going to upgrade it back to current with the modified gains or leave it alone. Maybe that's a problem for another day.
@Fat Corgi, I love that little accessory. I'm going to try to track it down and buy one of those. That looks like just what I need to mount the IMU. What a good find.
@ACP, thanks for the guidance. I can't believe some of the testing you've done. You're definitely helping us figure things out. Thanks so much.

 

[stj5353]

Oh, also, I did not notice it drop out of GPS mode today once! 8Sat's or better for the entire 3 flights, GPS was always selected.

 

[baja-king]

Oh, also, I did not notice it drop out of GPS mode today once! 8Sat's or better for the entire 3 flights, GPS was always selected.

That's good news, I would stick with 5.20, you lose the waypoints and the Bluetooth function which is a pain but I have found it reliable

 

[ACP]

Stj, was it locking to GPS?

 

[stj5353]

I'm not 100% sure how to answer that. I believe both times it had the same behavior with regard to gps. But with the older software it seemed to actually do something about it. The led and log indicated >6 satellites both times and I gave it a good 2 minutes of sync up time before flight. The older software never dropped out of mode 0 (gps) where I saw about 5 or 6 dropouts before. Same or similar satellite counts, give or take 1.