What's with the toilet bowl?

[srbell]

I've been flying my Carbon Scarab with a WooKong M, flies great.... Except in GPS mode. ATTi is rock solid, but GPS mode does a counter clockwise circle about 10 to 15'. I've got the GPS mounted far from anything that would cause interference. How do you get rid of this behavior?

 

[Tomstoy2]

Hmmm, re-vistit your cog settings, maybe?

 

[srbell]

Hmmm, re-vistit your cog settings, maybe?

Checked that already...

 

[matwelli]

does it use a magnetometer ? if thats out by 5 deg it will cause circling (did on my ardupilot) as the multi trying to get back to the set location, will keep missing it

 

[srbell]

Well, it does dog track a bit when flying forward in GPS mode. I guess the compass could be out a bit. I'll try rotating and see if that makes a difference.

 

[robvree]

I have found the whirlpool effect is definitely directly related to COG settings.

 

[srbell]

I got the cog front to back and side to side accurately, but maybe my cog top to bottom could be out. I guessed as it's not that easy to find the balance point top to bottom.

 

[vislaw]

There have been a couple of threads dealing with resolving the issue. Search for "toilet bowl" (be sure and spell bowl correctly - not bowel as in your thread title

I don't think there's a consensus yet on how to solve the problem. One person said he fixed it by doing the compass calibration before each flight (something that doesn't really make sense but he swore by it). Others have had success by moving the gps/compass. Some people report success by changing their COG settings while others have said it didn't make a difference for them. Some people have upped their gains to resolve it. It would be nice if we could get it figured out as I have been dealing with it on my AD6. I've moved my compass/GPS and will be doing some tests this weekend.

 

[Bartman]

just curious about something, the MK will also do this if the GPS gain isn't set close enough to what is appropriate for the winds aloft. but....without changing the GPS gain and with a little stick input you can stop the toilet bowl efect and it will be steady for a minute or two before it starts up again. you basically counter the toilet bowl pattern with opposite stick inputs to steady it and get it to settle down. does this work with the DJI controller?

 

[DucktileMedia]

I have been told by a few people that toilet bowling is directly caused by gains and inaccurate GPS location measurements. I have dialed mine in but I swear it depends where I fly. Today it flew great, yesterday it was a 30' circle. No changes. I am wondering if I need to use the longest rod that came with the GPS. i would prefer to have it as low as possible so it fits in my car better and is less likely to break if it tips over. Does anyone know why they give you 3 sizes to choose from?

There are so many variables from one WKM to the next. Ken has his settings twice as high as mine at 160-175. If I set mine at 250 the heli will go ape **** in the air and I can barely control it.

 

[matwelli]

i still reckon you need to look at the mag, without that working 100% GPS wont work, as GPS can not tell the flight controller physically what direction your multi is pointing. if the mag is out it will try to reach the desired location, but be flying 5-10 degrees off, over shoot, try to compensate, rinse repeat, in a big circle. from my experiance with ardupilot, gps gain only caused hunting back and forwards (esp with wind) and not in big circles

 

[RTRyder]

just curious about something, the MK will also do this if the GPS gain isn't set close enough to what is appropriate for the winds aloft. but....without changing the GPS gain and with a little stick input you can stop the toilet bowl efect and it will be steady for a minute or two before it starts up again. you basically counter the toilet bowl pattern with opposite stick inputs to steady it and get it to settle down. does this work with the DJI controller?

Yes, it does. Today I had the medium size quad I built for testing at the field and it was doing the TB thing in GPS mode which is odd because so far it's been rock solid testing it at the house. A bit of cyclic stick pressure will get it to stop and keeping the pressure on will make it hold fairly well but as soon as you let off it starts the circling routine, starting small and progressing into ever larger circles. Maybe today was just a bad day for GPS signal strength???

What bothered me most was I sent the quad about 30 yards downrange and triggered a failsafe, it rose to the correct height turned towards me, flew back, overshot the takeoff point, did 1 1/3 hard fast circles in reverse then shot off to it's left rear, my right as I was seeing it nose in. I had to very quickly go to manual and back to ATTI to stop it as it was heading off to the right at a very fast pace and would have been in the trees in short order if I hadn't stopped it. I've NEVER had a problem with this WKM and GPS before this, so I have to wonder what the heck is going on. Granted I haven't had a chance to get out to a big field and use RTH and GPS position hold since I built this frame and upgraded the WKM firmware to 5.02, so either one or both of those things may have contributed to the problem.

I'll be testing over the next couple of days to see if it was just a fluke or if there's something else going on.

Ken

 

[srbell]

I have to call BS on myself lol. When I said I've checked the cog, what I really meant was I had accurately measured the gps and imu. I didn't recheck it though. Had the unit plugged into the computer last night and was shocked to see the xyz measurements were nothing like I had entered! I'm thinking either a firmware upgrade tossed the settings or my 6yo decided to play a joke on me : ) As soon as the sun's up I'll give it a test flight and see if it's fixed. Thanks for all the input! Oh, being a newbie I'm assuming the cog settings are the xyz measurements for the imu and gps right?

 

[Siteline]

Yes, it does. Today I had the medium size quad I built for testing at the field and it was doing the TB thing in GPS mode which is odd because so far it's been rock solid testing it at the house. A bit of cyclic stick pressure will get it to stop and keeping the pressure on will make it hold fairly well but as soon as you let off it starts the circling routine, starting small and progressing into ever larger circles. Maybe today was just a bad day for GPS signal strength???

What bothered me most was I sent the quad about 30 yards downrange and triggered a failsafe, it rose to the correct height turned towards me, flew back, overshot the takeoff point, did 1 1/3 hard fast circles in reverse then shot off to it's left rear, my right as I was seeing it nose in. I had to very quickly go to manual and back to ATTI to stop it as it was heading off to the right at a very fast pace and would have been in the trees in short order if I hadn't stopped it. I've NEVER had a problem with this WKM and GPS before this, so I have to wonder what the heck is going on. Granted I haven't had a chance to get out to a big field and use RTH and GPS position hold since I built this frame and upgraded the WKM firmware to 5.02, so either one or both of those things may have contributed to the problem.

I'll be testing over the next couple of days to see if it was just a fluke or if there's something else going on.

Ken

RT What you described here is exactly what happened to us. This was our first try at failsafe which led to a crash. needless to say there is a bit of trust building to do.

 

[llbr22]

just curious about something, the MK will also do this if the GPS gain isn't set close enough to what is appropriate for the winds aloft. but....without changing the GPS gain and with a little stick input you can stop the toilet bowl efect and it will be steady for a minute or two before it starts up again. you basically counter the toilet bowl pattern with opposite stick inputs to steady it and get it to settle down. does this work with the DJI controller?

I find it a MUST to fly the bird into a perfectly still position before enabling GPS hold. Then yes, up the gain if needed for the wind conditions. But I do find if I fly 'into' a GPS position, then it will most likely start to circle. Also, if it is circling you can usually counter it again by using the sticks.

The WKM is NOT an autonomous controller for sure. Nor do i think it should be treated as such. Friday when I was filming my buddy fly his rc heli, not only did I need to stay out of the flight line, but I was 20' from power lines. Even though I was in GPS hold, There was no way I was going to trust taking my eyes off of it for more than a second or two.

 

[DennyR]

Ken if your GPS has had a crash you need to send it back. I have one that got damaged and it will fly ok but not where you want it to go autonomously.

 

[RTRyder]

Ken if your GPS has had a crash you need to send it back. I have one that got damaged and it will fly ok but not where you want it to go autonomously.

I don't believe it has anything to do with the GPS needing attention, it worked perfectly in wind here at the house holding position to within a couple feet for as long as I wanted to let it stay there, something else is going on here. It's not the TB effect that has me concerned, it's more the behavior of RTH that was flawless in previous versions of the firmware, now it appears to be not working right and I'm not the only one that has had it go wonky.

This frame also has a datalink with Click and Go active but I wasn't using it at the time. I have to believe that DJI has thoroughly tested the datalink for interference with the GPS and other modules so I doubt that had anything to do with it. There seem to be some issues with the current firmware release so I'm waiting to see what the next one brings. In the meantime I'm going to have a look at the IMU mounting and recalibrate the compass to see if that does anything. It would not surprise me if it works perfectly once again here at the house, even though the flying field is just a big open space it was once an active sand and gravel operation and there's no telling what might be buried there that could cause compass interference, like maybe an old dozer or other large metal object...

Ken

 

[robvree]

I have to call BS on myself lol. When I said I've checked the cog, what I really meant was I had accurately measured the gps and imu. I didn't recheck it though. Had the unit plugged into the computer last night and was shocked to see the xyz measurements were nothing like I had entered! I'm thinking either a firmware upgrade tossed the settings or my 6yo decided to play a joke on me : ) As soon as the sun's up I'll give it a test flight and see if it's fixed. Thanks for all the input! Oh, being a newbie I'm assuming the cog settings are the xyz measurements for the imu and gps right?

Yes, the xyz measurements are for the IMU and GPS in relation to the crafts COG. You really must take some time to calculate your COG, don't assume it is in the center. I found the most important variable for the GPS is the Z axis. If you are carrying a landing gear, batteries and a camera below the center plates, your COG will most likely be a good distance under the plates. Mine is about 7 cm under the bottom plate. This will make your z settings for both the IMU and GPS a NEGATIVE VALUE.

The bummer part of all this is that I have different battery setups (1 - 6200mah, 1 - 5000mah or 2 x 5000mah) and the software only allows you to use one value. When I switch to different batteries my COG changes.

 

[RTRyder]

I find it a MUST to fly the bird into a perfectly still position before enabling GPS hold. Then yes, up the gain if needed for the wind conditions. But I do find if I fly 'into' a GPS position, then it will most likely start to circle. Also, if it is circling you can usually counter it again by using the sticks.

The WKM is NOT an autonomous controller for sure. Nor do i think it should be treated as such. Friday when I was filming my buddy fly his rc heli, not only did I need to stay out of the flight line, but I was 20' from power lines. Even though I was in GPS hold, There was no way I was going to trust taking my eyes off of it for more than a second or two.

It's odd because back at the middle of last summer I did this video using my Droidworx AD6 on MK electronics. I basically parked it in position hold and altitude hold at about 225 feet vertically and let it sit there for about 5 or 6 minutes. The back and forth yawing is a compass board that was about to go bad as I found out afterwards, it died completely two days later never powering up again but that's a subject for another thread. I forget what version of firmware it had at the time but the GPS position hold was really solid and so was altitude hold, both seem to have gone downhill in the firmware versions since or perhaps there's just more GPS interference lately. In any case I used to have this kind of GPS hold with the quad that gave me problems yesterday...

Ken

Ken

 

[Tomstoy2]

Wow, nitro smoke,,, God help me, I LOVE it! No wonder everybody hated us,,,,,

Nice vid, by the way!