Phantom Yaws CCW

[questech]

Hello All.

My Phantom started acting weird a few flights ago.

When I fly GPS and ATTI mode don't work as well as they did... the Phantom hovers and glitches like it's getting hit with radio interference. It will tend to fly around a bit, and if I let it go it will cover a 30' area.

In a hover it will yaw CCW 90degrees and as much as 180 degrees.

During forward flight it will fly better than it hovers, but it will glitch and drop to on side briefly or roll side to side rapidly.

When I am in a hover and use my rudder control to yaw slowly (fast yaw works much, much better) it will fly in a direction instead of remaining stationary.

I have a good idea of what the problem is after my intensive testing, but I'd rather not say at this time so that I don't influence any feedback... I will post my thoughts on this later.

Now here is what I've done to test and troubleshoot the condition:

* I have upgraded the motors to the T-Motor Anti Gravity replacements.
* I swapped out the receiver.
* I swapped out the GPS.
* I swapped out the Naza flight system.
* I checked the continuity of my ESC connectors to make sure that they had a good connection and weren't shorting out.
* I moved to several locations (miles away from each other) to make sure it wasn't radio interference (because the problems look like RF Interference).

** I tied the Phantom down on a bench and ran up the motors to watch (no rpm meter of any sort) the speed parity (more on that in a later thread).

The Phantom acted exactly (or nearly) the same with all of these various components.

So I can comfortably say that it Isn't: * GPS * NAZA Flight System * Radio Receiver * Motors.

No I may be wrong, I am new to the Phantom, but I think I covered a lot of bases.

What do you guys think??

I appreciate and welcome your feedback!

AM

 

[questech]

BTW I failed to mention that I experimented with many, many gain combinations. I returned to default, and now I'm at 125% on everything but Yaw and Vertical.

 

[martcerv]

Id suggest to try re-calibrating your radio first. Second making sure the compass is secure and in the same position it comes in as stock and on the same leg. If you have 3rd party landing gear make sure its on placed in a way that will be the same as the stock gear being a few degrees out will cause issues and this will mostly effect the heading and yaw hold.

 

[questech]

Hi martcerv, thanks for the post.

The compass is in the exact same position it was in when I purchased the Phantom. I double checked by substituting a GPS unit that included the compass in it, and it made no difference.

I failed to mention it but I calibrated the radio multiple times, as well as doing the compass dance before every flight, and doing multiple advanced (IMU?) calibrations within the assist software.

Thanks again!!

AM

 

[Driffill]

Hi martcerv, thanks for the post.

The compass is in the exact same position it was in when I purchased the Phantom. I double checked by substituting a GPS unit that included the compass in it, and it made no difference.

I failed to mention it but I calibrated the radio multiple times, as well as doing the compass dance before every flight, and doing multiple advanced (IMU?) calibrations within the assist software.

Thanks again!!

AM

Can you explain more on how you tested the compass using a different GPS module? The phantom naza is "tuned" to read the compass data from where the compass is mounted in relation to the phantom, using a Naza GPS module made for the 450's etc is likely to give faulty readings to the naza controller.

I'm a little suspect on the compass, but, if I had to go all in and place a bet on what your issues is, I'd guess a "cold" solder join. It will look fine in a continuity test, but once high amps start running thru it, it could cause voltage drop, that would cause un explained turning, tho the naza should fight to correct this (via the compass data!)

 

[martcerv]

I did also notice this a little when trying 8x5 graupner clone carbon props, may be due to the lack of lift as these props could barely lift my setup with FPV gear and gopro. Flight time was half of the stock props and it couldnt hold altitude very well along with some yaw drifting.

 

[questech]

Can you explain more on how you tested the compass using a different GPS module? The phantom naza is "tuned" to read the compass data from where the compass is mounted in relation to the phantom, using a Naza GPS module made for the 450's etc is likely to give faulty readings to the naza controller.

I'm a little suspect on the compass, but, if I had to go all in and place a bet on what your issues is, I'd guess a "cold" solder join. It will look fine in a continuity test, but once high amps start running thru it, it could cause voltage drop, that would cause un explained turning, tho the naza should fight to correct this (via the compass data!)

Good Points....

First of all, the compass.

Yes, I did use a GPS unit from a NAZA-M Lite for a test, actually a bunch of tests, and it flew exactly like the stock compass. In fact other than the glitches I mentioned, like the Phantom randomly turning 90 degrees or so CCW, it flew fine.

I believe that DJI put the Phantom compass where they did to keep it away from other electronics in the cramped interior quarters of the shell.

BTW I flew it yesterday and I noticed that if I Yaw hard right or left and hold it, the Phantom spins nicely in place, but when I use my rudder to yaw slowly, the Phantom fly's away. In fast flight I don't notice any abnormalities, only in slow flight and hover do I have problems.

Now this could mean that the stock compass and position isn't as critical as led to believe, or you could read it as I definitely have a compass problem

My gut feeling is that it isn't a compass problem, for what that's worth.

I'm going to use my brother's transmitter at some point so that I can rule out a transmitter issue. My Futaba T8J is only 3 months old and I take great care of my gear, so if it is defective, it is not from abuse of any kind. I also want to believe that it isn't the transmitter... but we'll see at some point in the future.

Now the cold solder joint thing intrigues me.

I've been soldering for over 30 years, but I don't consider myself a master solderer, so that's worth looking into.

Now where do you think that a suspect cold solder joint might be that could cause this type of problem?

I replaced the motors, so that is a possible place to look, but would that cause these symptoms rather than something more like erratic motor behavior (which this actually could be)??

Now I know that I got the lands hot enough and that heat melted the solder of the tinted wires, so unless I'm missing something (please chime in) that should not be a cold solder joint.

I also swapped the ESCs out (2 of them) so that the red LEDs are on the Pilot's side and the green LEDs are on the passenger's side (I'm a pilot so that red in front and green in back bugged the hell out of me).

Again I used the same technique as with the motors, however I did notice a bunch of flux residue around the joints, but I don't think it was intermixed in the joint.

You mentioned that a cold joint would test out fine for continuity, but that it may be problematic once high amps are cursing through it... How do I test for that?

I have suspected a bad (not burned out, but freaky) ESC all along, so I ordered a couple of replacements.

Could the ESC exhibit random inconsistencies like this rather than just burning out and not working at all?

Oh... Also; I ran a bench test with the Phantom tied down.

I was monitoring RPM discrepancies (without an RPM Meter). I cycled all of the motors with the props attached and noticed something very definitive.

When I would Yaw with the rudder control it seamed like my M3 motor was spinning slower than it's corresponding motor. The opposing motor (M1) created much more thrust when Yawing Right. Now when Yawing Left both M2 & M4 seam to provide the same amount of thrust.

Wow, that was a mouth full

Thanks for the feedback and I would appreciate your response!

Gratefully, ANthony

 

[questech]

I did also notice this a little when trying 8x5 graupner clone carbon props, may be due to the lack of lift as these props could barely lift my setup with FPV gear and gopro. Flight time was half of the stock props and it couldnt hold altitude very well along with some yaw drifting.

That's an interesting point!

I'm using the stock props that are very well balanced, but they have many flights on them and they have been under some stress. They are very flexible to the point of where I wonder how they even get the craft into the air.

I too have FPV and a GoPro but the load has never seemed to be an issue.

I have some Graupner props myself that I think I'll try to see if the stiffness makes a difference.

BTW I upgraded to the T-Motor Anti Gravity motors and I expected more flight time. Well they run smooth as can be, but my flight time wnet from 6 - 8 minutes down to 5 - 6 minutes :-(

I hope that I have a bad ESC and that once I replace it my flight time will go back up to where it should be.

Thanks for the input!

AM

 

[martcerv]

I did also notice this a little when trying 8x5 graupner clone carbon props, may be due to the lack of lift as these props could barely lift my setup with FPV gear and gopro. Flight time was half of the stock props and it couldnt hold altitude very well along with some yaw drifting.

That's an interesting point!

I'm using the stock props that are very well balanced, but they have many flights on them and they have been under some stress. They are very flexible to the point of where I wonder how they even get the craft into the air.

I too have FPV and a GoPro but the load has never seemed to be an issue.

I have some Graupner props myself that I think I'll try to see if the stiffness makes a difference.

BTW I upgraded to the T-Motor Anti Gravity motors and I expected more flight time. Well they run smooth as can bIe, but my flight time wnet from 6 - 8 minutes down to 5 - 6 minutes :-(

I hope that I have a bad ESC and that once I replace it my flight time will go back up to where it should be.

Thanks for the input!

AM

Getting shorter flight times if the only change is the motors is odd, are all the motors spinning as freely? If there is an issue with an ESC this could be causing your problems and best to check this as I have heard of a few guys that had them catch fire after doing a poor solder job. If its using more power then it should and you haven't changed props or added weight then investigating to see which esc is having issues and fixing this will save you a likely crash due to failure in flight if there is one playing up.

The graupner 9 inch is ok but the 8 inch has significantly less lift then a stock dji 8 inch prop so this will give you even shorter flight times. It cut my run time in half when using 8x5 graupner after my motor upgrade but with stock props I got a little better then before and now using larger triblades it does well up to 1.3kg when flying my 3d setup.

 

[questech]

Getting shorter flight times if the only change is the motors is odd, are all the motors spinning as freely? If there is an issue with an ESC this could be causing your problems and best to check this as I have heard of a few guys that had them catch fire after doing a poor solder job. If its using more power then it should and you haven't changed props or added weight then investigating to see which esc is having issues and fixing this will save you a likely crash due to failure in flight if there is one playing up.

Hi...

The motors are in perfect condition. I had the same exact problem with the stock motors, so I think I can safely rule out a motor issue.

I have the thing running pretty smoothly right now after adjusting the gains, but I think I have some more tweaking to do.

The self induced yawing is still happening though. I believe that one of my motors is slowing down a little bit for some unknown reason which is causing the problem.

When I did a bench test and when I yaw right with right rudder control, the M3 motor produces considerably less thrust.

When giving left rudder controlM2 and M4 produce equal thrust.

Now that doesn't necessarily mean anything, or maybe that's the symptom that leads to the answer.

What other ways can I test the ESCs to identify a faulty one?

Thanks!!

AM

 

[martcerv]

If you have v4 firmware you could try the motor test function, this may show one spinning less with the same load as in the test it just gives what I think is a quick and consistent bit of power to each individual motor. With this if there is an esc with issues it may show up with less rotation during the quick burst test. If one clearly shows a difference then changing that esc will hopefully fix your issues, you may also get away with just resoldering the connections to the mainboard and motors from that esc before replacing it to make sure its not just a bad solder joint.

 

[questech]

If you have v4 firmware you could try the motor test function, this may show one spinning less with the same load as in the test it just gives what I think is a quick and consistent bit of power to each individual motor. With this if there is an esc with issues it may show up with less rotation during the quick burst test. If one clearly shows a difference then changing that esc will hopefully fix your issues, you may also get away with just resoldering the connections to the mainboard and motors from that esc before replacing it to make sure its not just a bad solder joint.

Thanks for the feedback!

I received 2 new ESCs and I'm going to replace the suspect ones... we'll see what happens.

Thanks Again!

Anthony

 

[MrMediaGuy]

Curious to know the outcome. My bet is on a bad ESC on M3. SInce you already have the new ESCs you might as well just swap them out. Otherwise you could swap the *motors* on M2 and M3 and repeat the bench test. If M2 then becomes the problem, you know it's the motor. If M3 remains the problem, then it's the ESC or the NAZA. And since you've replaced the NAZA, bets are then on the ESC. Let us know....

 

[questech]

Curious to know the outcome. My bet is on a bad ESC on M3. SInce you already have the new ESCs you might as well just swap them out. Otherwise you could swap the *motors* on M2 and M3 and repeat the bench test. If M2 then becomes the problem, you know it's the motor. If M3 remains the problem, then it's the ESC or the NAZA. And since you've replaced the NAZA, bets are then on the ESC. Let us know....

Hello,

Changed out the ESCs and rewired the phantom to make sure all of the solder connections were good.

Same condition, periodic yaw action.

Same thing happens to my Octocpter with NAZA-M V2.

Yawing is natural if any of the motors slow down or speed up without some correction. I don't know why NAZA doesn't bring the bird back to it's original heading though??

I thought it might be me being a little heavy thumbed on the sticks, but then I took my hands completely off of the transmitter, and it still happened.

I can manually compensate, but isn't that what GPS mode is all about??

I'll keep experimenting... Anthony

 

[martcerv]

When you tested the motors did you do this in manual mode or gps/atti as in manual it should just be a direct response to stick input and not dealing with the naza or compass. This will either show an issue with an esc or wiring if you notice a difference in manual. If the issue is only in gps or atti mode then it makes it a little trickier but at least youd rule out esc's, motors and wiring.

By any chance is your controller not going back to zero when centered on the yaw axis? Are you using a dji stock radio for the phantom or a 3rd party radio that your also using on your octocopter. As this sounds like a strange issue that isnt common with most users as naza does pretty well at yaw hold normally. To have the same rare problem with 2 naza multi's then I suspect it may be due to the only likely common component being the radio.

Does well for me in gps, manual and atti with no yaw drift at all. Only time it did was with the graupner 8x5 props for some strange reason but on hundreds of other flights it has had no issues with all other props s5ock or futaba radio.

 

[questech]

When you tested the motors did you do this in manual mode or gps/atti as in manual it should just be a direct response to stick input and not dealing with the naza or compass. This will either show an issue with an esc or wiring if you notice a difference in manual. If the issue is only in gps or atti mode then it makes it a little trickier but at least youd rule out esc's, motors and wiring.

By any chance is your controller not going back to zero when centered on the yaw axis? Are you using a dji stock radio for the phantom or a 3rd party radio that your also using on your octocopter. As this sounds like a strange issue that isnt common with most users as naza does pretty well at yaw hold normally. To have the same rare problem with 2 naza multi's then I suspect it may be due to the only likely common component being the radio.

Does well for me in gps, manual and atti with no yaw drift at all. Only time it did was with the graupner 8x5 props for some strange reason but on hundreds of other flights it has had no issues with all other props s5ock or futaba radio.

All good questions and points!

First of all I only did flight tests in GPS and ATTI modes. I have only flown these multis in a very close quarters and have not attempted to fly them in manual mode, ever.

I am using a Futaba T8J

Being that I have never flown them in manual mode (and the phantom being my first multi, which I learned to fly on) I don't know how difficult it would be to control without the assistance of the FC... so I certainly don't want to find out in the small space that I'm flying in with so little room for error.

I thought it might be the radio as well (yes same TX for both multis), so I did some testing. Now I didn't have another TX to test so I did the only thing that I could think of, and that was to monitor my inputs in the Assistant app.

Even though this isn't a "Real World Test" I did everything that I could to simulate actual flight conditions.

Looking at the screen, I did not see any deviations that would induce yaw, and the controls returned to center every time. So there was no indication that the transmitter was sending any commands to the FC that were not intentional.

When this unintentional yaw happens it can be slight, in which case a slight opposite control will correct, or it can be more severe and the craft will yaw 90 degrees, at which point I have to take more aggressive action.

I thought it might be my location, so I took it elsewhere, but it acted the same way.

It seems to happen mostly in wind, but if I remember correctly it may have happened once or twice on calm days.

In the wind it appears that when the system compensates to maintain position that the act of spinning up some of the motors to counteract the displacement causes the yaw.

Also, if I recall correctly, it always yaws to the left.

Maybe the same dynamics that were at play when you used the graupner 8x5 props are at play in my situation??

Thanks for helping to keep the topic alive!!

Anthony

 

[martcerv]

I meant when doing a prop off bench test, if you set the control to manual mode then seeing if motors spin differently between left and right yaw or different pitch angles. In gps mode the naza even in the hiuse while not flying may alter power to each motor as may atti mode if its not 100% level. Doing this test in manual will give a true indication of how the control inputs effect the motors rpm.

 

[questech]

I meant when doing a prop off bench test, if you set the control to manual mode then seeing if motors spin differently between left and right yaw or different pitch angles. In gps mode the naza even in the hiuse while not flying may alter power to each motor as may atti mode if its not 100% level. Doing this test in manual will give a true indication of how the control inputs effect the motors rpm.

Oh... Sorry about that.

I did a bench test prior to changing the ESCs. I haven't done another yet, but the swap didn't change anything.

* When on the bench and I use the rudder left control input both M2 & M4 seem to operate at the same RPM and observable thrust (no tachometer or thrust scale used).

* When inputting right rudder control M1 & M3 seem to spin at different RPMs. M3 seems to run slower and produces less thrust.

Based on this test I made sure that I replaced M3 with a new ESC... b ut again that didn't make any difference in flight.

BTW I am 80% sure that I ran that bench test in manual mode, but it's been a while and a lot of water has passed under the bridge so to speak :lol:

 

[OI Photography]

Hi Anthony (et al)...I found this thread while trying to resolve the same issue...mainly wild yaw when giving light control input as well as some other inconsistent instability in general. I had added an aftermarket landing gear, but even after moving the compass back to the stock location and multiple recalibrations of the compass and IMU both I still experience the "drunk driver" turns from light yaw on the stick.

I haven't bothered to try to estimate the output differential of the motors like you have, but I suspect I'd find the same results...however I also suspect that's the symptom more than the cause. Seems to me like the NAZA is having issues regulating power in that situation, since you've pretty much ruled out faulty ESC's.

I have the T-motors on the way just because they're better built than the stock ones (better/quieter bearings, etc) and some carbon fiber props, but I doubt either will help with this issue, and I'm planning more test flights first before installing them.

Since this has been such an elusive problem, I'm trying to use the various telemetry recorded with my Flytrex module ($50) to figure out the source, and initial tests indicate that the NAZA is registering some weird data in relation to compass heading and relative altitude, so that may be the source.

I'll post any further insight I gain on the matter, and I might suggest you try looking at your telemetry as well to see if we can isolate the cause here and hopefully fix it!

Fingers crossed!

-Chris

EDIT: Anthony, what batteries are you using, just the stock ones? It just occurred to me that there's a possibility a substandard batt could cause issues with the current that might result in RPM variances.

 

[questech]

Hi Anthony (et al)...I found this thread while trying to resolve the same issue...mainly wild yaw when giving light control input as well as some other inconsistent instability in general. I had added an aftermarket landing gear, but even after moving the compass back to the stock location and multiple recalibrations of the compass and IMU both I still experience the "drunk driver" turns from light yaw on the stick.

I haven't bothered to try to estimate the output differential of the motors like you have, but I suspect I'd find the same results...however I also suspect that's the symptom more than the cause. Seems to me like the NAZA is having issues regulating power in that situation, since you've pretty much ruled out faulty ESC's.

I have the T-motors on the way just because they're better built than the stock ones (better/quieter bearings, etc) and some carbon fiber props, but I doubt either will help with this issue, and I'm planning more test flights first before installing them.

Since this has been such an elusive problem, I'm trying to use the various telemetry recorded with my Flytrex module ($50) to figure out the source, and initial tests indicate that the NAZA is registering some weird data in relation to compass heading and relative altitude, so that may be the source.

I'll post any further insight I gain on the matter, and I might suggest you try looking at your telemetry as well to see if we can isolate the cause here and hopefully fix it!

Fingers crossed!

-Chris

EDIT: Anthony, what batteries are you using, just the stock ones? It just occurred to me that there's a possibility a substandard batt could cause issues with the current that might result in RPM variances.

Hey Chris, Thanks for the Interest!

AS far as the batteries, I used many different (high quality and lower quality)batteries and there is no change in the behavior.

It seems to happen mostly when correcting when the wind pushes it away from its position. I don't believe (95% + sure) that it has ever yawed on its own when moving in a forward/reverse or left/right direction. It happens when hovering.

BTW I replaced the stock motors with the T-Motor Anti Gravity motors after this started happening, and it didn't make any difference at all.

I think you're on to something with the telemetry, I suspect that NAZA is doing something and it isn't the fault of any other component.

PLEASE share your future findings!

I get a similar, even though to a lesser degree, behavior with my 960mm octocopter using a NAZA-M V2 Upgrade.

I am going to try a different flight controller for the 1200mm octocopter I'm building now.

One thing that helped, at least I think it helped, was to increase the yaw gain. I don't fly the phantom much anymore as I only used it as a test platform, but with the octo, when it's in a hover and starts to yaw, I've learned to be quick (but gentle) on the controls and compensate. It works great, although I have a FC that's supposed to do that, but again, my next tests will be with a different FC.

Anyone have a suggestion for a good FC that has great GPS hold/hover and has a go home and land failsafe mode (or just GREAT GPS Position Hold)?

Thanks Again Chris !!!!

Anthony