Independent tests prove lack of frequency hopping with XPS

[RH1N0]

	Quote: Originally Posted by XJet
	These devices are ony useful if the receiver stops outputting pulses when the signal is lost or interference is received -- a lot of the DSP-based PPM receivers just go into a default "last position hold" for quite a while before they stop outputting pulses. I'll check this cheap Chinese 2.4GHz set and see what it does.

I've noticed that JR PCM receivers change frame rate when they are in fail safe. It's only a slight change and not enough to upset the servos but it is a useful way to detect that the receiver is in failsafe when the servo pulse duration is set to hold.

I'm interested to know if any of the 2.4 systems do this as well. Worth a look when you test the system...

[3DDevil]

XJET... when you tested the IMAX 9x system, did you notice that the plane was way more sensitive than when flown on PPM or PCM? I found this when I first flew my plane on XPS. I really liked the extra responsiveness...They recon its because with 2.4Ghz you get a lot more signals/sec to the servos?

That was until my XPS locked out and the plane crashed!!....

[XJet]

	Quote: Originally Posted by 3DDevil
	XJET... when you tested the IMAX 9x system, did you notice that the plane was way more sensitive than when flown on PPM or PCM? I found this when I first flew my plane on XPS. I really liked the extra responsiveness...They recon its because with 2.4Ghz you get a lot more signals/sec to the servos?

I suspect that, much as folks claim "but it's true", this talk of "locked in" feeling and extra responsiveness is more placebo effect than anything.

Why?

Well no matter how fast your transmitter and receiver communicate, the fastest your servos can accept data is once every 20mS (50 times a second). That means there's an average latency of 10mS between when you move a stick and the servo becomes aware of that move.

Since a module-based 2.4GHz system can't send the data any faster than it's being created by the transmitter, it can't produce a faster response-time than the old "dumb" FM/PPM system.

Any module-based 2.4GHz system is actually going to introduce *extra* latency because it has to process each frame of data (1.5mS x num of channels) and convert it into a set of digital values, then send it to the receiver. Once the receiver gets that data, it then has to convert the digital values into a pulse for each servo.

The *fastest* and most responsive of all radio systems is actually the good old FM/PPM setup with a "dumb" (ie: no DSP or other computer stuff) receiver. With these old systems, the average delay between when a stick is moved and the servo receives the new position data is still still 10mS but the signal is passed on to the servo immediately, without the additional latency of processing or frame-reconstruction. In fact, the delay between when a pulse enters the transmitter module and when the same change appears at the servo lead on an old FM/PPM system is measured in *microseconds* rather than the milliseconds we associate with PCM, DSP or module-based 2.4GHz systems.

In fact I noticed a definite reduction in latency when switching from a Berg 5/II in my Katana profile to a plain old "dumb" FM/PPM receiver -- because the Bergs (like most DSP receivers) actually tend to buffer a whole frame before they output it. The difference *was* noticeable when flying. I've noticed no such difference when comparing my JR 9XII (9303) PCM to either the XPS or this cheap Chinese 2.4GHz gear.

The latency of an old FM/PPM system is also a whole lot better (a frame less delay) than a PCM system too -- but I don't recall folks saying that their FM gear felt more responsive or "locked in" than their PCM systems.

And let's not forget the weakest link of all -- the mechanical parts of the servos. Most servos are rated somewhere between 100mS and 250mS transit-time. This makes a delay (or improvement) of a few mS in latency elsewhere in the system seem rather insignificant.

If people are finding their 2.4GHz module-based systems feel more responsive then just what were they using before?

[KrisW]

"It's so much faster and responsive.. I can FEEL the difference ! ! !"

Ahem. . Slick Marketing at it's best.

[3DDevil]

Xjet,

I have to disagree with you on the 'responsivness' feel factor of the XPS. In the end its one of the only things I miss from it... I can tell you I definately noticed the plane was way more responsive. My low rates became like my high and my high became ultra high. I cannot tell you why, but I know what I felt. Other guys at my club found the same thing with XPS. One guy was even flying an old warbird as a test bed and he even found this plane to be more responsive and twitchy.

I was using a JR 9XV2 with PCM RS77S Rx before the XPS. I then used the XPS module and 8Ch Rx. I changed nothing in the plane settings on the radio before I first flew with XPS and beleive me it was way more touchy and needed to be tuned down in the radio for the second flight. As I said. I dont know why, but I know what I felt. I also learnt what I didnt feel.... and that was a response to stick input, just before the Extra hit the ground, destroyed itself and became another XPS statistic!!...

[MalaysianFlyer]

	Quote: Originally Posted by KrisW
	"It's so much faster and responsive.. I can FEEL the difference ! ! !" Ahem. . Slick Marketing at it's best.

I had one 14MZ owner come up to me and start going on and on about how fast the response was on the radio with G3 PCM, and how it was faster even when set back to the old PCM1024 mode (god only knows how he figured that), and kept insisting that it was faster when I disagreed. He finally shut up when he realised that I was using an FX-40 and decides to exit with what dignity he has left, then he comes back 5 mins later and starts telling me about latency test results that someone had emailed to all the local flyers that shows how fast the 14MZ is compared to other radios. (He did not realise that the person who had emailed the latency test out to the local flyers was actually ME!!!! )

So, I just told him that it must be great to have 200ms latency and can you please show me your hands on the sticks moving to five discrete positions in one second, to enjoy this speed LOL

Strangely enough he doesnt talk to me anymore LOL (thank god)

Every radio manufacturer be it JR FUTABA XPS SPECTRUM XYZ DEF ABC and CR*P will always have to claim there is something that makes it worthwhile for us to spend $$$$ and more $$$$ hence the "faster response time ploy" "increased resolution ploy" etc etc

Interestingly enough, one of the test on RCgroups showed a 10yr old Airtronics stylus as having a lower latency than a 14MZ OWWWW my FX-40 is now a piece of crap as it cant beat a 10yr old radio!!!!

Sometimes ya just got to laugh..... then go fly and enjoy yourself!!!!

[Julez]

Coming back to the original topic of this thread:
It seems XJet was not fast enough, someone else did the tests before him.
Nevertheless, these tests prove his opinion to be correct:

http://www.rcgroups.com/forums/showthread.php?t=814276

[bob_nj]

Incredible and scary at the same time.

You did just take all the fun out of this thread though

[ricoalonso]

Julez,
I read the thread and what I understood is that even the notion of XPS going to 'frequency hop' if the noise is slowly introduce seems to be not true also??

[jonkoppisch]

..

[Julez]

	Quote: Originally Posted by ricoalonso
	Julez, I read the thread and what I understood is that even the notion of XPS going to 'frequency hop' if the noise is slowly introduce seems to be not true also??

This is the way I understood it as well.

[XJet]

	Quote: Originally Posted by 3DDevil
	Xjet, I have to disagree with you on the 'responsivness' feel factor of the XPS. In the end its one of the only things I miss from it... I can tell you I definately noticed the plane was way more responsive. My low rates became like my high and my high became ultra high.

Did you compare the actual servo throws with the FM setup and the XPS setup?

It *is* possible that XPS is simply delivering greater servo throws per degree of stick movement.

The latency of XPS will always be greater than the latency of a dumb FM/PPM setup, that's a fact, so if you're experiencing greater sensitivity then the XPS is not accurately representing your stick positions.

[XJet]

	Quote: Originally Posted by ricoalonso
	Julez, I read the thread and what I understood is that even the notion of XPS going to 'frequency hop' if the noise is slowly introduce seems to be not true also??

That would appear to be the conclusion reached by the tester - and it does agree with my own informal tests.

I was willing to give JD the benefit of the doubt (due to the unscientific nature of my initial testing) but it seems that the entire frequency hopping abilities of the XPS are now in doubt.

If this is the case, it certainly puts XPS well behind Spektrum/JR and Futaba in terms of the resilience and robustness of its radio-link.

Those folks who are happy to place their giant-scale planes and the lives of bystanders in the hands of a system that (it would appear) can be shot down by interference on a single part of the 2.4GHz band are free to do so but it's not something I would do.

True, it's no worse than 72MHz -- but it's no better either. XPS remains a viable 2.4GHz option, but certainly not the best one when you're flying expensive and/or dangerous models.

[sweetpea]

The only thing the test is missing is comparison of FASST and Spektrum on the same test to verify that they are actually more robust and unaffected by the same situation.

Until that is done it is all speculation on our parts.

My opionion is that they will pass...but I'd like it to be 100% proven.

[jonkoppisch]

According to the name of the thread... " Independent tests prove lack of frequency hopping with XPS" I'd say that there's no longer any speculation...

I believe that spektrum would be effected as well but have 2X's the chance to survive over xps unless the entire 2.4 band is swamped then none of them would survive. Futaba is looking like the overall best solution but..