Glitches on my servos (USM)

[Raythegreat]

Hi,
I flew today and I'm almost sure my plane got servo glitches or (USM) unintentional servo movements.
The plane rolled suddently about 20 degrees. Quick correction after that. BUT it happened 3 to 4 times in flight. I have digital servos so they are fast to respond to glitches.
SO:
I have a Taranis X9D+ with JXT_ACCST_2.1.0_FCC.frk, and the latest openTX firmware.

Is it normal to get USM's with ACCST V2 I updated in 2021?

Will the UNI-RX fix this issue?

Thank you
Ray

[bob195558]

Hi Ray,
This may be the cause for your servo glitches.
My experience with servo glitches was caused with having the servo wires running close and/or touching each other.
Part of the signal to one servo will jump to the other servo wire causing a servo glitch (electromagnetic radiation).
The fix for me was to move the servo wires further apart or another way do a mod to one of the servo wire to be a shielded wire cable.

[Raythegreat]

Thanks Bob.

In my case I don't think it's caused by servo wires close together.
I suspect more the RF link losing a frame, something like that.
The USM are very sudden and cause a quick unwanted rolls (in this airplane).
I've been suspecting this for a while but now with the digital servos, the USM seems amplified.

I'm going to update to uni Rx for all my receivers.

I have accst v2 that was supposed to solve that problem. Did it solve it?
It does not appear so. Does somebody knows about this?

Ray

[bob195558]

Hi Ray,
I am not up on that stuff, maybe Mike B. will see it and have an answer.

[MikeB]

The USM problem only occurred very rarely, I would be very surprised if it happened several times in one flight, and the V2 firmware does fix it.
UNI will give you an extra telemetry value, VFR, which will indicate if you are losing a lot of RF packets.

What power supply are you using? Digital servos can take a lot of current, so you could be getting a significant dip in the power supply voltage to the Rx.

Mike

[Raythegreat]

Hi Mike,

Thank you for your response.
I have an AeroStar 60A ESC with a built in 5V 5Amps BEC. Battery is 3300mAh 4S.
I checked the Rx voltage and it's a solid 5 volts even when I move all my four digital servos together (2 AIL, RUD, ELE).
So there is no voltage drop on the RX. I determined that with the telemetry Rxbt sensor using the + and - too.

The problem I saw is a sudden movement of the roll axis (bank of about 20 degrees very suddenly). It does not happen often but it did at least 2 times in my last flight. It cannot be the wind since all other flyers with small foamies were stable.
I don't believe there is crosstalk between the servo wires since the problem is not permanent but sudden.
Crosstalk usually causes erratic and permanent small glitches on servos (which I don't have).
Servos are very stable when I see the plane near me on the ground. But in the air, it's difficult to know exactly how the servos are behaving.

So if I understand well what you said is: Because I have ACCST V2, the FrSky known unwanted servo movements are not possible? Right?

Thank you
Ray

[MikeB]

ACCST V2 certainly removes the known USB bug.
If I understand correctly only Ail is glitching? Are the Ail servos on long cables?
If only Ail is glitching, then I would think it unlikely to be caused by a RF problem.
Do the servo cables run close to the Rx antennae?

Mike

[Raythegreat]

Mike,

Thanks again Mike,

From what I have experienced with this plane (only 4 flights), only glitches were seen on the roll axis.
A glitch on the rudder is pretty hard to see. As for the elevator, I don't think I saw something as bad as the roll problem.
Please also note I only see that behavior with this plane that has digital servos, not on all my other planes with analog servos.

You are correct, the two AIL have longer cables (12 inches inside the fuselage) and do pass near one of the antenna and Sport wire also but not near the tips.
These antenna are supposed to be shielded except for the tips, so I assumed they cannot cause crosstalk. True?
The receiver is a FrSky X6R.

So at this point, it seems crosstalk is the only explanation after you confirm RF is unlikely. If RF is unlikely, I'm glad.
I will try to put servo cables even closer to the antenna so that I may reproduce on the ground what I saw in the sky.

So Bob also might be right about servo wires. I have nothing else at this point. My first suspect was RF.
I still consider RF to be a suspect because of the way the problem occurs; very suddenly. Why very sudden if servo wires crosstalk?

You've been a great help Mike and Bob.

Thank you,
Ray

[Kilrah]

Would still suspect power. Not at the receiver but at the servo with the length of the wire.
Digital servos can draw current in huge spikes, which over some length and if the wire is too thin can cause significant voltage drop, which can make them glitch/reboot.
A multimeter would also be way too slow to measure those.

[Raythegreat]

Hi Kilrah,

Yes good idea as always. I should check the voltage at the servo itself with a scope
Unfortunately I don't have a real scope. Maybe I can use the Arduino made scope.

It's going to be difficult to check the voltage at the servo itself since wires are inside the wing.
I would need to open the wing, make a jig to intercept the signal. Not an easy project.

Picture attached of my Pattern Plane.

Unless you have a better way, please let me know.

Thanks
Ray

IMG_20221230_172331.jpg

(718.04 KiB) Not downloaded yet

[bob195558]

Maybe if you have an other same servo you could use it to do the test outside of the plane wing.

[Raythegreat]

Hi Bob,

Unfortunately, I don't have another servo like these ones.
It's a complicated job to de-install a servo and make a jig to intercept the data signal.
This plane is grounded until I have more info.

So according to Mike B, RF ACCST V2 should not be the culprit unless servo cables contaminate the Rx receiver.
I will fix all the wires issues, placements, add use twisted cables.

One important remark concerning my problem.

The snap roll occurred when the attitude of the airplane was straight (levelled), so I eliminated the possibility of a sudden stall.
This plane was built from a kit and the it might be one of the leading edges not perfectly symmetric (the same) between each wings.

My question is: If the snap roll of approx 20 degrees bank occurred very suddenly (and 2 times), could the power still be an issue?
If power is the issue, why wouldn't I have more erratic behaviors instead of a clean flight except 2 snap rolls?

Thanks,
Ray