Frsky problem.

[takilara]

Interestingly enough, I noticed that when I was doing the review of the FASST compatible receiver, the servos which were most jerky with the FrSky appeared just as jerky with a genuine Futaba FASST transmitter and receiver.

I'm continuing to investigate.

Was these the savox?

the align's i have (DS610) seem to be rebranded savox's. I also have a DS650, but that is on the tail, and going through a gyro, so i can't say if that acts the same, never tried it on another channel.
Looking around on forums, they have gotten mixed reception. They have real good stats, but not everyone seem happy with them. I haven't flown them enough to have anything negative to say about them.

[RCModelReviews]

I tried the Savox and the XCore (coreless digital) servos with similar results.

[Allan Warner]

Hi,
I've re posted this, with some editing and additional info, from another thread as this is a more suitable location.
I'm not happy. I purchased a telemetry Tx and rx set plus additional normal 8ch rx(has a 'HV' label on side of case). Used it in my Futaba FF8, rx battery A123 cells or NiMh . Using either rx the system has a distinct dislike for certain servos by causing a very pronounced cogging/notchy movement instead of a smooth movement of the servo o/p disc in response to the tx stick movement. There was no jitter at neutral or steady stick position. I would not fly using the first 3 listed as it would be like having a sticky/binding control. Although, having read this thread, it may be that 3D style of flying, banging the sticks about, may be be less noticable.
Servos not liked:
1/ futaba S3001
2/ futaba S3004
3/ futaba S3010
4/ supertec naro hp
5/ gws pico
6/ ripmax sd200
7/dimond d250
8/ hitec hs50
9/ towardpro mg996r(digi) - slight cogging
seamed to work OK with the following:
1/hitec hs 60
2/hitec hs 645
3/ hitec hs 5645(digi)
4/ savox sc-0251(digi) Had a much higher whine than normal
5/ power hd 9150(digi)
6/ futaba s9001
7/ futaba fd30m(old but unused/plug changed to fit)
I gave up at this point even though there were more servos to try. All the servos were new. They work fine on other systems (2.4 Spektrum and on 35 - futaba, hitec, gws, jeti) with either power supplies. It was on all 4 primary channels(didn't see the point in checking further).
Is this a hardware fault/software fault or a design fault. Either way it's not fit for purpose IMO. I paid a total of UK£80 for these items and if the issue cannot be resolved or I get a refund then their in the bin. There's something not right and I am not prepared to risk a large'ish plane or an injury to someone. Anyhow, I suspect my BMFA insurance maybe invalide, if I did fly, as I am responsible for ensuring that it is safe to fly and there is some doubt.
I've posted a version of this on the Giantcod(the sellers) forum and hopefully I'll get a definitive answer from somewhere.

[RCModelReviews]

I'm looking at why some jitter and some don't -- it's a *very* strange situation.

I'm going to try running the pulse through a signal buffer and see if it might be voltage-related (though it shouldn't be).

Whatever the cause, I'm pretty sure that if it does turn out to be a problem, FrSky will sort it -- they've been amazingly good in comparison to the likes of Corona, et al when it comes to customer service.

But I'll keep folks appraised as to what I find.

[Allan Warner]

Further info: as previously stated used 4 & 5 cell NiMh packs and A123 pack - no change. Have noticed that the speed of movement of the stick(rate of change of positional information data)has an effect. Slow or very fast movement the effect is greatly reduced. Intermediate speed (like you were doing a string of flowing aeros) has the worst effect. I think it unlikely its a hardware fault on two rx and I feel it would effect all servos equally. I'm swaying towards a software/design fault - possible clash of rx signal processing rate or trasmission rate with rate of change of positional data??? With slow/small stick movements there is time for everything to take place in an ordered maner. With fast/large stick movements the servo goes into full switch on mode to drive the feedback pot to its required position as quickly as posible(latency/delay of the system is mainly a function of the servo). It's the intermediate rate of change that appears to give loss of data(frames) and the coggy o/p. The fact that some servos appear to work OK could be the design of the servo amp (deadband, digi etc) masking the effect. I've posted this on other forums to see what I get back.

[RCModelReviews]

Further info: as previously stated used 4 & 5 cell NiMh packs and A123 pack - no change. Have noticed that the speed of movement of the stick(rate of change of positional information data)has an effect. Slow or very fast movement the effect is greatly reduced. Intermediate speed (like you were doing a string of flowing aeros) has the worst effect.

What you're probably seeing there is the effect of the data transmission packets not being synchronized to the frame rate of the transmitter.

That means that when you're moving the stick at a moderate rate, the difference between each consecutive servo pulse is not consistent.

It's called jitter and is intrinsic to all FHSS systems to some degree.

There's no affect on resolution and in flight it produces no discernible effect but it does make servo movement appear to be "notchy". I'll try to draw up a diagram to show what I mean because it can be a little hard to explain in words.

[Allan Warner]

I understand, but it is not slight and I would not care to fly using the first three of the affected servo types. Never seen anything like this with my Spektrum modules in my Futaba FF9 and FF9 super. No one would accept a car that could only be driven smoothly flat out or at tickover. The normal cruising rev range being plagued by missfiring, although it would still get you about. I'm sure you follow my analagy.

[Allan Warner]

I retired 14 years back and only have a DVM as test equipment plus observation and deductive reasoning. Is the following reasoning sound? To reduce/eliminate the cogging effect would require some/all of the following:
I/ Increase the sampling rate of the PPM pulse train by the Tx modul.
2/ Increase the rate of frequency hoping.
3/ Increase the rate of signal processing in the rx
This, I would think, may require components/ic's/micros with better specs therefore costing more money and increasing the price of the system making it less competative. Is the cogging effect a symptom of being built down to a price.

[mr.sneezy]

Hi Allan. I see you've been cross posting this topic in a few forums...
Keep in mind that most users don't see any problem with this in the air, and I suspect you won't either if you try. That said, the solution may involve additional software latency between stick movements and RX servo output. It may well be why some other brands are much laggier. I guess what I'm saying is with the current chipsets you may well be able to make it smoother, but will end up with latency and a 'disconnected' feeling while flying.
Yet another way of expressing my opinion is 'be careful what you wish for'.

[Allan Warner]

I'm coming round to wishing for a full refund!