Loiter mode(and altitude hold) stuttering, how to tune? (Flightvideo)

[HowDoIfly]

The RTL and Loiter issues support the compass is mounted/configured wrongly. According to the pic I linked to, you have it back to front.

[TheEdge]

The RTL and Loiter issues support the compass is mounted/configured wrongly. According to the pic I linked to, you have it back to front.

Yes, I would also say that this is an issue with compass/gps, I just don';t know what..
According to the picture I linked to, I had it mounted correctly, but have also tried with rotating it like in the picture you linked to, still the same..

Starting to think that this compas/gps thing is junk..

[HowDoIfly]

Yes, I would also say that this is an issue with compass/gps, I just don';t know what..
According to the picture I linked to, I had it mounted correctly, but have also tried with rotating it like in the picture you linked to, still the same..

What APM firmware version are you using? I just caught one instruction that claims that unit is not compatible with APM:Copter 2.9.1b and earlier

[TheEdge]

What APM firmware version are you using? I just caught one instruction that claims that unit is not compatible with APM:Copter 2.9.1b and earlier

It is arducopter 3.1.5.

When I connect to terminal it says:

Init Arducopter V3.1.5 (3c57e771)

Free RAM: 1618
FW Ver: 120



So, what should I do to set up the compass correctly.

There are several actions that can be done that involves the compass, I am not sure of in what order to do the steps..

I mean, you can enter magnetic declination, you can calibrate the compass, you run the compassmot and you can set the orientation.. So, in what order should I do these steps? or is there something else I should do?

[HowDoIfly]

According to the guide, you plug in and mount the GPS with compass, then set the declination (we use a negative number) then select the orientation correction (which yours appears to be roll 180 if the compass is underneath) and then the calibration. That is the minimum you must do.

[TheEdge]

According to the guide, you plug in and mount the GPS with compass, then set the declination (we use a negative number) then select the orientation correction (which yours appears to be roll 180 if the compass is underneath) and then the calibration. That is the minimum you must do.

This is what I get when I choose my place on the declination website:
You clicked here:
Latitude: 61° 11'; 16.8" N
Longitude: 7° 1'; 7.3" E
Magnetic declination: 0° 49'; EAST
Declination is POSITIVE
Inclination: 73° 24';
Magnetic field strength: 51238.5 nT

I';ve got Positive declination so I figure I have to enter 0 in degrees and 49 minutes, is that right??

In the Flightdata where you see the model in the map there is a red line indicating Current Heading. In my twisted mind, this red line should point in the same direction as the front of the quad at all times, right?
So that when I rotate the quadcopter clockwise, this red line should also rotate clockwise, always following the front of the quad..??
At least this is what my criusboard did, but the apm does not do that.. It rotates opposite of what my quad does, and hence does not point in the same direction as the front of my quadcopter. This can';t be right..??

[HowDoIfly]

Yes, your mag dec is 0.49

In mission planner, if you look at the compass bearing, it should show your quad';s magnetic heading. If you know your bearings, point the quad at a known quadrant (N,S,E,W) and check it shows that heading (for me, I';m sitting at my computer facing roughly north). Yaw the quad clockwise and check the heading reading follows suit

N->E
E->S
S->W
W->N

etc. It might be inaccurate if there';s nearby interference but it should show if it';s going round the right way at least.

[TheEdge]

Yes, your mag dec is 0.49

In mission planner, if you look at the compass bearing, it should show your quad';s magnetic heading. If you know your bearings, point the quad at a known quadrant (N,S,E,W) and check it shows that heading (for me, I';m sitting at my computer facing roughly north). Yaw the quad clockwise and check the heading reading follows suit

N->E
E->S
S->W
W->N

etc. It might be inaccurate if there';s nearby interference but it should show if it';s going round the right way at least.

Yes, but the problem is not that it is a little bit inaccurate, the problem is that it works exactly as if it is upside down, which it in fact is since it is mounted underneath the gps.
That is why the rotation is opposite of what is real, and as long as it is like that, loiter and rtl will not work..

I tried just for the heck of it to mount the gps/compass unit upside down, and the thing flies like a dream in all modes.. Now also Loiter and RTL works, and compass heading shows correctly in mission planner.

But this is not ideal since the gps is now underneath the compass and pointing to the ground. It still works, but I guess the gps reception is not as good as it would have been with the gps on top..

So, I can very well fly the thing as it is now, but there must be a setting somewhere in the firmware that will compensate for the stupid compass being upside down..???
Roll 180 obviously does not do the trick..

Mounted this way it works

[HowDoIfly]

If you examine the magnometer, there should be a white dot in one corner. Reference the part number for the spec sheet which will tell you it';s direction of axis. The GPS antenna needs to be right way up (facing the sky), otherwise it';s orientation doesn';t matter.

When you';ve worked out the orientation of your compass, there';s a whole host of dropdown options in mission planner to flip axis around to your hearts content.

[TheEdge]

If you examine the magnometer, there should be a white dot in one corner. Reference the part number for the spec sheet which will tell you it';s direction of axis. The GPS antenna needs to be right way up (facing the sky), otherwise it';s orientation doesn';t matter.

When you';ve worked out the orientation of your compass, there';s a whole host of dropdown options in mission planner to flip axis around to your hearts content.

Well, the white dot is in the front left corner. Mounted like that the heading is correct. So if I flip it over sideways, the white dot will be in the front right corner, but upside down compared to now. Heading should still be correct, but y-axis would sort of need to be inverted.. I don';t know what setting I would have to change to give the correct function if I flip it over.. Rotation Roll 180 seems logic to me, but I tried with that setting before I flipped it upside down, so I don';t know..

[HowDoIfly]

Well, the white dot is in the front left corner. Mounted like that the heading is correct. So if I flip it over sideways, the white dot will be in the front right corner, but upside down compared to now. Heading should still be correct, but y-axis would sort of need to be inverted.. I don';t know what setting I would have to change to give the correct function if I flip it over.. Rotation Roll 180 seems logic to me, but I tried with that setting before I flipped it upside down, so I don';t know..

I don';t want to bale out on you but I';ve not had to address this issue, it';s just stuff I';ve picked up from other people';s threads.

All I can offer is that mine has white dot in front right and needs no orientation adjustments. For it to be in the same position as yours, I would have Yaw 270degrees. That';s before we flip it over.

Can you read the part number? Perhaps we can find the spec sheet.

[TheEdge]

Yes, I guess we are a bit off topic since I started out with a crius and megapirate, but thank you for staying.

I looked in the datasheet yesterday but did not find anything there that helped me.
Maybe you have better luck. Here';s a closeup of the chip.


When I flip the unit over so that the gps face the right way up, also I have the dot in the front right corner(pointing to the ground since it is on the underside of the gps), so I guess I could use same settings as you??

[HowDoIfly]

Yes, I guess we are a bit off topic since I started out with a crius and megapirate, but thank you for staying.

I looked in the datasheet yesterday but did not find anything there that helped me.
Maybe you have better luck. Here';s a closeup of the chip.


When I flip the unit over so that the gps face the right way up, also I have the dot in the front right corner(pointing to the ground since it is on the underside of the gps), so I guess I could use same settings as you??

As I mentioned before, we don';t care about GPS orientation as long as it';s antenna faces upwards. We need to identify the magnetometer chip and it';s orientation on the board.

[TheEdge]

As I mentioned before, we don';t care about GPS orientation as long as it';s antenna faces upwards. We need to identify the magnetometer chip and it';s orientation on the board.

Yes, I know. But since it is working correctly now just being mounted upside down, one would not really need to know the actual orientation just figure what change needs to be done when the chip is flipped over sideways. I guess the magnetometer chip is the big one in the picture, the one marked with a white dot in one corner. The white dot is as I have mentioned front left now. If I flip it over so that the gps is on the top the white dot will be front right and the corresponding change in the config needs to be done.

What setting in the missin planner will correspond to flipping the chip sideways?

[HowDoIfly]

one would not really need to know the actual orientation

I have to disagree 

The big chip is the ublox gps, not the magnetometer. I';m beginning to suspect your compass is actually on top (opposite the LED) and if we get it pointing forward, you just set it as internal, no orientation correction needed.

[TheEdge]

The big chip is the ublox gps, not the magnetometer.

I agree, this is the chip for the gps.

Since this one clearly has a white dot in one corner(on the board though) I thought you ment this chip, sorry for the confusion 

When I look at the picture(I am currently at work and do not have the compass with me) I see some smaller chips with white dots marked in the board in one corner, maybe the one with 8 legs in the upper left corner in the picture is the magnetometer chip? I will have a closer look at it when I get home and try to identify the magnetometer chip.
I have marked the chips in this picture

[TheEdge]

Just flipped it over sideways and changed compass to rotation_none. It is correct now, with the gps on top.

Thank you for all contributions.

I might do some more attempts with the Criusboard maybe on a miniquad or something, so maybe I';ll have some more questions here later on topic..

[HowDoIfly]

Just flipped it over sideways and changed compass to rotation_none. It is correct now, with the gps on top.

Thank you for all contributions.

I might do some more attempts with the Criusboard maybe on a miniquad or something, so maybe I';ll have some more questions here later on topic..

Glad we got there in the end.

[TheEdge]

If you get a duff one it is not going to be anything other than bad luck.  That said, of the several I';ve bought on eBay over the past 12 months I';ve had 100% success so far.

Well, I guess I do not have much luck with these things then, I probably got a duff one..
Was flying around just fine in perfectly stable and calm loiter, suddenly the stupid thing starts falling, and even with full throttle it just falls(not like the motors stopped or anything, it';s just in a hurry to crashland  into the ground..) Pick the thing up and fly it again, again in perfectly stable loiter for a couple of minutes, and then the excact same thing happens again.

It';s in the trash now along with the rest of my rc equipment. I am just not ment to fly..

[Hands0n]

Noooooo! Don';t do that. There will always be an explanation for what went wrong. Crashing is not inevitable but an ever present chance. What you have just described sounds like something else completely, not the flight controller but perhaps the RX or even TX Throttle input to the FC. If it were a fault associated with a single motor if expect it to tip or flip over, not for all motors to slow down.

I';m sure that working through this latest problem with folk on here will restore you to flight again.

I';ve had "suicidal" aircraft in this one year of experience and learned that all the faults can be tracked down by a process of elimination.

Sent from my Nexus 5 using Tapatalk

[TheEdge]

Noooooo! Don';t do that. There will always be an explanation for what went wrong. Crashing is not inevitable but an ever present chance. What you have just described sounds like something else completely, not the flight controller but perhaps the RX or even TX Throttle input to the FC. If it were a fault associated with a single motor if expect it to tip or flip over, not for all motors to slow down.

I';m sure that working through this latest problem with folk on here will restore you to flight again.

I';ve had "suicidal" aircraft in this one year of experience and learned that all the faults can be tracked down by a process of elimination.

Sent from my Nexus 5 using Tapatalk

Well, I didn';t really throw it in the garbage, but I threw it in a pile and needed some time away from the thing as I was getting so tired of always testing/troubleshooting/fixing/fabricating/waiting for stuff to arrive/reading forums, instead of flying.
I just downloaded the logfiles from the apm in case they can give any clue of what happened.
https://dl.dropboxusercontent.com/u/98131243/quadcopter/logs.zip
2014-06-28 11-58-17 is the one flight when the crashlandings happened.



I had some flights with the thing the weekend before and everything was working.
Before the last flight(when the two crashlandings happened) I mounted the mobius action cam and a china 5,8Ghz transmitter on the quad for fpv use if that could have anything to do with the crashlandings..

[Hands0n]

In away from anything that would let me see your log. One of the others may be able to feedback.

I';d say that one approach would be to reverse any recent changes. So maybe remove or disable the power to the VTX. That may not be the cause of course but it is one notable change.

Once you get these things stable they tend to stay that way unless something changes or breaks (I had a prop failure just before coming away on holiday). But before that my Disco was flying flawlessly. So there is no reason you can';t achieve same.

Sent from my Nexus 5 using Tapatalk

[HowDoIfly]

Sorry I';ve not read the thread for details but I can run through the log for you.

From an APM perspective (megapirate might do things different by default)...

The crashlandings log...

Pre-Arm checks are disabled. (not advisable)

Only 5 sats and high HDOP (3.2-5.6) reported.

While this probably doesn';t help pinpoint the cause of your crashes, they are worth noting and fixing as a matter of course.

ETA: Do you have anything commanded through Channel 7? I notice it cycles rapidly between "normal" and "low" occasionally.

[TheEdge]

Sorry I';ve not read the thread for details but I can run through the log for you.

From an APM perspective (megapirate might do things different by default)...

The crashlandings log...

Pre-Arm checks are disabled. (not advisable)

Only 5 sats and high HDOP (3.2-5.6) reported.

While this probably doesn';t help pinpoint the cause of your crashes, they are worth noting and fixing as a matter of course.

ETA: Do you have anything commanded through Channel 7? I notice it cycles rapidly between "normal" and "low" occasionally.

Thank you for reading the log. I have not found a good way of reading the logs yet.
Yes, I disabled the prearm checks just to get started, cause I was doing the initial tests inside and had trouble getting gps lock inside, so I turned it of to be able to arm it for testing. Forgot to reenable it.

We(several quadcopter pilots here) have had some issues with the number of usable gps. So since I was in loitermode  both times I crashed, it might be gps related. Have just never had something like this happened before, so it might also have something to do with the fact that I mounted my mobius camera and a vtx just before the flight. Before this the thing was flying very nice, so I thought I was finally ready to go fpv with it.

I am not sure about channel 7 but I have the Turnigy 9x set up with a progmix that combines a 2 and a 3 way switch to give me 6 flightmodes. Flightmodes is connected to channel 5 but I noticed that I was not able to make use of all 6 positions untill I connected a cable from the receiver to channel 7. When I flick one of the flightmode switches, the value of channel 7 changes. Does the activity of channel 7 you question correspond with the change of flightmodes?
Or maybe it could be that either the camera or the vtx creates some disturbance on channel 7..

I have not fully understood the way the progmix of the turnigy 9x is working, so I might have done something strange there.  I was hoping that the progmix would need only channel 5 for all 6 flightmodes so that I could use channel 6,7 and 8 for other things, but I might have it set up wrong.

The receiver is not currently connected, so I am not sure exactly how I had it connected and what went to channel 7.

[HowDoIfly]

Thank you for reading the log. I have not found a good way of reading the logs yet.
Yes, I disabled the prearm checks just to get started, cause I was doing the initial tests inside and had trouble getting gps lock inside, so I turned it of to be able to arm it for testing. Forgot to reenable it.

We(several quadcopter pilots here) have had some issues with the number of usable gps. So since I was in loitermode  both times I crashed, it might be gps related. Have just never had something like this happened before, so it might also have something to do with the fact that I mounted my mobius camera and a vtx just before the flight. Before this the thing was flying very nice, so I thought I was finally ready to go fpv with it.

I am not sure about channel 7 but I have the Turnigy 9x set up with a progmix that combines a 2 and a 3 way switch to give me 6 flightmodes. Flightmodes is connected to channel 5 but I noticed that I was not able to make use of all 6 positions untill I connected a cable from the receiver to channel 7. When I flick one of the flightmode switches, the value of channel 7 changes. Does the activity of channel 7 you question correspond with the change of flightmodes?
Or maybe it could be that either the camera or the vtx creates some disturbance on channel 7..

I have not fully understood the way the progmix of the turnigy 9x is working, so I might have done something strange there.  I was hoping that the progmix would need only channel 5 for all 6 flightmodes so that I could use channel 6,7 and 8 for other things, but I might have it set up wrong.

The receiver is not currently connected, so I am not sure exactly how I had it connected and what went to channel 7.

Out of the logs in your zip attachment, are any of them prior to adding the camera and vtx? If you can pinpoint one, I can compare SatCount and HDOP.

I';m still learning T9X settings but if you go into the menu you can see which additional channels (5-9) are enabled and what they';re assigned to. I don';t know how to work it out (or hazard a guess) from the logs, tbh.

[TheEdge]

Out of the logs in your zip attachment, are any of them prior to adding the camera and vtx? If you can pinpoint one, I can compare SatCount and HDOP.

I';m still learning T9X settings but if you go into the menu you can see which additional channels (5-9) are enabled and what they';re assigned to. I don';t know how to work it out (or hazard a guess) from the logs, tbh.

In the zip-file only 2014-06-28 11-58-17.log is with the cam and vtx mounted, the three others are without.

Ch5 Gear
Ch6 n u l l  (I used to do some channel 6 in flight tuning with Hov thr pot)
Ch7 Hov Thro
Ch8 PiT Trim
Ch9 Hov Pit

I use a combination of F.mode and the gear switch for flightmodes
The flightmode mixes was set up so long ago that I do not remember how, but I followed a tutorial I found somehwere. I think Mix1,2 and 3 is used for flightmodes.

I think that the Gear switch is manipulating channel 7 the way I had it connected.
If I would flick the gear switch while in Loiter mode, it would have changed to Alt.Hold mod.
I am pretty sure I did not use the gear switch during the flight where I crashed, but maybe interferance from the vtx could have manipulated ch7 so that it would change to alt.hold
If that is what happened, it still should not have crashed with full throttle.

[HowDoIfly]

Sorry, just got back from an errand, I';ll work my back through the logs (from crashlanding one) and see if the Sat stuff is different.

No worries about Ch7 if it';s unused. It';s just showing as flipping between two ranges (low and normal) so guess the dial is sitting just on the cusp.

brb

[HowDoIfly]

I';m working my way back from the crashlanding log. The previous is also showing 5 (at most) Sats and a slightly better but far from ideal HDOP of just under 3

Also notice this

22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Throttle dropped below FS_THR_VALUE meaning likely loss of contact between RX/TX.
22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Current Distance from Launch is 18.58m ~ 60.95ft
22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Current Direction of UAV is 131 degrees (0 = north).
22/06/2014 16:25:08 - 011300: THROTTLE ERROR: Throttle error resolve meaning likely RX/TX contact restored.

[HowDoIfly]

Pretty much the same for the other two logs, low Sat count and too high HDOP. How is your GPS mounted and do you wait for it to get updated data or just arm once you get a 3D fix?

[TheEdge]

Pretty much the same for the other two logs, low Sat count and too high HDOP. How is your GPS mounted and do you wait for it to get updated data or just arm once you get a 3D fix?

I place the quadcopter on the "launchpad" and connect the battery, The blue led indicating 3d fix comes on very quickly, I leave it there for 3 minutes with the battery connected before I arm the motors and takeoff. I always do this unless I am in the middle of quickly testing something that does not require gps/compass.

So, the logs show low gps count,  and I know from two other quadcopters that use to fly here that they sometimes also have low gps count. It is probably a difficult area for gps reception rather than the way the gps is mounted. The stand for the gps broke off due to my frustration after the crash, so it is currently not mounted and I do not have a picture of the way it was mounted during that flight.
But this picture show how it was mounted, except that I flipped the compass/gps unit sideways so that the gps is on top like it is supposed to be.
If I lost 3d fix during loiter, would that explain the crash? I mean, would no 3dfix in loiter keep the qudcopter from rising when I apply full throttle?

I';m working my way back from the crashlanding log. The previous is also showing 5 (at most) Sats and a slightly better but far from ideal HDOP of just under 3

Also notice this

22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Throttle dropped below FS_THR_VALUE meaning likely loss of contact between RX/TX.
22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Current Distance from Launch is 18.58m ~ 60.95ft
22/06/2014 16:24:53 - 010729: THROTTLE ERROR: Current Direction of UAV is 131 degrees (0 = north).
22/06/2014 16:25:08 - 011300: THROTTLE ERROR: Throttle error resolve meaning likely RX/TX contact restored.

The connection loss with tx is from a log dated 22nd of june, that was not the log from the crash.
At one time I did test the failsafe by turning of the tx during flight, and I think that is what you see in the log. The log from the crashflight is 2014-06-28 11-58-17.log

Sorry, just got back from an errand, I';ll work my back through the logs (from crashlanding one) and see if the Sat stuff is different.

No worries about Ch7 if it';s unused. It';s just showing as flipping between two ranges (low and normal) so guess the dial is sitting just on the cusp.

brb

Chanel 7 is not unused I guess as the gear switch manipulates chanel 7