G-RX8 + telemetry + PWM mode = continuous digital servo chatter?

[Landru]

Greetings,

Can anyone solve a mystery for me?

I have 4 G-RX8 receivers. All of them cause continuous servo chatter in my MKS and KST digital servos when telemetry is on and the receiver is in PWM mode. In S-BUS mode, the chatter disappears (apart from the odd momentary noise). Other receiver and transmitter settings such as frame rate and firmware have no effect.

The problem can reproduced consistenly with simply the receiver, a 5V dc power source, the servos, and a barebones OTX model profile in my X10S or QX7 transmitter.

Any ideas?

Best,
Andrew

 

[Konrad]

This is normal. In PWM as a result of the RX clock speed there is a lot of time for the servo to wander. The next clock cycle (pulse) trigger the amp to reposition the servo. With Digital SBUS the amp is powered for a much longer time meaning that the servo can't move much (dead band) so the chatter is much less. Chatter is the actually movement in the gear train verse amp buzz (power draw of the servo).

With telemetry on, part of the band width is used for this telemetry data rather than the whole band width being used for PWM data. So the servos look to be responding a bit rougher. There is also the fact that with telemetry on there is another source of RF noise that the servo (leads) are picking up.

Remember that Analog is using the shape of the pulse to position the servo. Any noise on the signal wire will change this shape. The digital amp servo can detect this change with a higher resolution than the classic analog amp servo. SBus uses a digital numeric code and is not effected by analog noise.

 

[Landru]

Thanks, Konrad, for the detailed and helpful explanation.

Forgive me if I'm being slow but one thing is still unclear to me: In S-BUS mode, channels 1-6 still output a PWM signal, not digital. Yet, there is no chatter on those channels even with telemetry on. If chatter can be removed in the case of those PWM signals, why not in PWM mode?

In any case, is it safe to fly a model with that chatter? The servos vibrate vigorously.

Regards,
Andrew

 

[Konrad]

Not sure but I think it has to do with the refresh rate (9ms vs 18ms). This why it is recommended that you use digital servos with these RX even in PWM mode.

"Vibrate Vigorously" You and I might have a different definition of vigorously. A non filtered video might be best to answer that question.

 

[Landru]

Yes, could be a difference of opinion. I'll post a video asap. When the servos are mounted in a model, the plane sounds like a popcorn machine chattering away.

FWIW I've tried the various settings that affect the 9ms vs 18ms refresh rate. You can set it in the RX of course, and, with OTX, you can also affect in the TX with the channel range (e.g. channels 1-8 gives 9ms; channels 1-16 gives 18ms). None made any difference to the chatter.

Thanks again for your help.

 

[Konrad]

Well, you have now exceed the knowledge of my pay grade.

I'd need to put the RX on a scope.

Good luck,
Konrad

P.S.
I've flown 4, G-RX8 receivers* in this bird using PWM 18ms. I haven't had a problem and you will note that I have long leads to the rear servos. And yes the digital KST servos do like to talk to me.
https://forum.alofthobbies.com/inde...eter-glider-build-thread-on-alofts-forum.125/

* I had an operator induced issue. I used this ship as a test bed for my G-RX8 receivers.
https://forum.alofthobbies.com/inde...eceiver-is-not-ready-for-the-mass-market.227/

 

[Landru]

LOL I got caught by the same channel 7 S-BUS gotcha when setting up my first G-RX8. As a last resort, I invoked the RTFMS protocol. And, as usual, it provided both useful information and the opportunity for a face palm.

If you ever put the G-RX8 on your scope, I'd be curious to know what you find.

I'll make an effort to post a video later today as mentioned. Then you can tell me how loud my servos talk compared to yours. ;-)

Cheers,
Andrew

 

[Landru]

A couple of videos:

PWM mode

Note: As you see in the video, I forgot to remove the jumper from the signal pin of channels 1 and 2 before recording the video. The same chatter occurs without the jumper.

S-BUS mode

 

[Konrad]

WOW, that is dramatic!

My KST X08 v5 don't make that much noise. I do have a set of MKS 6100HV. I'll need to try to duplicate what you're getting. I'll get back to you with my findings.

.rtf files - I hate reading and understanding them. It just just proves what I don't know!

 

[Landru]

Yes, I thought it was pretty bad, too.

When the servos are out of the model, the noise is less dramatic. However, the servos do still chatter continously and vibrate quite obviously to the touch.

I'm very curious to know if you can reproduce the problem with your 6100HV.

And if Wayne or anyone else reading here has any thoughts, please chime in.

I've also emailed FrSky for support. However, my past experience is that they never respond to emails.

 

[Konrad]

OK. Per my manual with the RX in PWM mode Ch1-Ch4 are high precision (<5ms), Ch5-Ch8 are ordinary precision. My manual does not indicate that this can be changed with jumper.

Testing with both MKS HV 6100 and KST X08 plus servos I get the same results. In Sbus mode all servos are quiet. (Like in your video).
In PWM mode Ch1 - Ch4 are quiet, but the ordinary precision channels Ch5 - Ch8 are as noisy as yours in your video. (Not as loud as mine are not in an amplifying glass fuselage)

The outputs Ch5 - Ch8 to my eye aren't moving with this noise, just like in your video.

If your Ch1 - Ch4 are quiet and Ch5 - Ch8 are noisy I think you are fine.

 

[Konrad]

Going by the on line manual with jumpers across Ch5 - Ch6 should put the RX in High Speed mode.
https://www.frsky-rc.com/wp-content/uploads/Downloads/Manual/G-RX8/G-RX8- manual.pdf

In this configuration (High Speed mode) and in PWM I saw no change is the servo noise print. CH1-Ch4 quiet and Ch5- Ch8 noisy. So by my testing HS (9ms) and FS (18ms) makes no audible differents. But there is a definate difference in the noise from the high precision channels (Ch1-Ch4) and the ordinary precision channels (Ch5 -Ch8) while in PWM mode.

All the best,
Konrad.

PS
All tests were done using a X9D TX running 2.2 OpenTX and the RX was powered by a 2 cell Life battery.

PSS
I just repeated these test on 2 more RX G-RX8 with the same results. I think you don't have an issue with the hardware as I have flown all 3 tested RX at great range in my 3.2m Samsara glider. Now the manual could be made a bit clearer (nothing new here!)

 

[Landru]

Absolutely brilliant!

I think you have nailed it. Another case of RTFMS.

I'll confirm your findings asap, possibly this evening if I can manage it.

Thank you for your careful and thorough investigation!!

With thanks,
Andrew

 

[Konrad]

You are more than welcome.

Not sure you got snagged by the RTF file. The manual is not clear as to what to expect from "precision" or high speed modes. I have to give you a lot of credit for noticing. I for one didn't catch this. Because you asked, I now have a much better understanding of the RX's characteristics. So thanks for asking the question.

Now I have to figure out how to bill FrSky for this customer support!

All the best,
Konrad

 

[Landru]

Your support is much better than any I've received from FrSky. In seriousness, kudos to you.

I can't see how this issue hasn't snagged other G-RX8 users. Aparantly, a lot are using them as 4-channel RX for DLG. Maybe that's a factor.

I'll be back in touch asap.

Best,
Andrew

 

[Konrad]

...I can't see how this issue hasn't snagged other G-RX8 users. Aparantly, a lot are using them as 4-channel RX for DLG. Maybe that's a factor.

I'll be back in touch asap.

Best,
Andrew

Well, I fly mostly 7 channel E-Soars. The first four channel are mapped for my primary flight controls. Ch5 and Ch6 are flaps and Ch7 is motor. Now I have to admit that because of the large movement of the flaps most of my ships have always had noisy flap servos. And as most of us are use to Digital servos making noise we probably didn't notice the difference between Ch1-Ch4 and the greater noise from Ch5 and Ch6. As the ESC (Ch7) has no motor in doesn't make noise.

Now I will say I don't know why CH-7 is dead in SBus. I'm a little disappointed in this RX as I wanted to use a redundancy feature of this RX. But one only gets this in SBus mode. Now as I want that 7th channel I can't use Sbus without using Sbus servos. Not practical, as so few servos are SBus compliant. So no redundancy feature for me?. Today my go to RX is the R-X8R pro. It has all I want in a glider RX and none of the gadgets.
https://alofthobbies.com/frsky-rx8r-pro.html

 

[Konrad]

Andrew,
Have you had a chance to verify the noise profile between Ch1-Ch4 and Ch5-Ch8 with your receivers?

All the best,
Konrad

 

[Landru]

Hey Konrad,

I squeezed in some testing this morning before leaving for work.

The Good
Not surprisingly, I confirmed your findings for PWM mode. Specifically, in PWM mode with telemetry on, my digital servos on channels 1-4 were completely silent, while those on channels 5-8 chattered continuously. Here was my set-up:

TX
QX7 running OTX 2.2.2

RX
Hardware: G-RX8 V1.2
Firmware: 170929
Binding: PWM mode (jumper on 1-2) with 'Telemetry on channels 1-8' selected in TX
Power source: 5.0V linear BEC on CC Phoenix 25 ESC

Servos
MKS DS-6100, MKS DS-450, KST X08+

The Not-so-good
When I tried S-BUS mode, I got an unpleasant surprise. Channels 5 and 6 were silent, but channel 1 had continuous servo chatter and channel 2 had continuous servo buzz. That result was with the same equipment listed above with the following changes:

RX
Binding: S-BUS mode, FS framerate (no jumpers), with 'Telemetry on channels 1-8' selected in TX (which reportedly results in a 18ms framerate in TX)

I was running out of time for thorough testing but did find that I could only eliminate chatter and buzz with the following set-up:

TX
As above

RX
Hardware: G-RX8 V1.0 (I don't know if the hardware version is significant)
Firmware: 170929
Binding: S-BUS mode, FS framerate (no jumpers), with ***'Telemetry on channels 1-16'*** selected in TX (which reportedly results in a ***18ms*** framerate in TX)

Conclusions???
At this point, I'd want to do thorough testing of all the possible permutations before attempting a firm conclusion. However, like you, I'm growing more disappointed with this receiver.

And like you, my main application is 7-channel e-gliders. Continuous digital servo chatter or buzz is not something I want live with. The obvious workaround of using S-BUS mode with S-BUS converter(s) and the receiver bound for 18ms framerate in the transmitter begs the question of why not use a different receiver entirely (as you suggested). The bad altitude drift in the G-RX8 vario adds more motivation in that direction.

For curiosity's sake if nothing else, I will do more testing. I'm likely to get to that on Friday evening or perhaps on the weekend.

Meanwhile, very interested to hear any further thoughts you may have.

Best,
Andrew

Edited to correct mistakes in framerates info.

 

[Konrad]

Are you transmitting 8 channels?

Sorry I have no idea about Ch1or Ch2 in SBus mode.

Do the other Rx's have this Ch1-Ch2 issue?

 

[Landru]

Only transmitting channels 1-8 as I recall. That's another possible factor I forgot about.

In any case, the two RX I tested suffered from the S-BUS chatter and/or buzz on channels 1 and 2.

At this point, I think more careful testing of all possible permutations is needed. There are quite a few! As things stand, I think I have a combination that eliminates chatter, but I'm not sure if it's the only possible solution or of which factors are critical.

Flying with the receiver is the goal but this little mystery has become a project in itself.