Kiwi's XPS test results are up.

[JKos]

> XPS has 12 wide SS channels so may be more resistant to narrow band
> interference, than Spektrum for instance. FASST has very narrow invidividual
> channels, but it sort of simulates one big fat channel that covers the whole
> spectrum.

I would like to correct this statement. While the XPS channels are spaced 5 MHz apart, the bandwidth of the actual signal is 1.6 MHz. Futaba also uses 1.6 MHz of bandwidth. Both Spektrum and the new Airtronics FHSS system use 1 MHz bandwidth signals. All of this is per the FCC test reports on file and publicly available.

And to clear up a few myths; none of our R/C 2.4 GHz systems use collision avoidance nor does Futaba FASST avoid any channels. Yes, their marketing seemed to claim that at first, but they've drastically changed the wording on that "feature."

Also, despite having a narrower bandwidth, Spektrum actually has more processing gain than Futaba; so it's not all about bandwidth either but also the chip to data rate ratio. I'm not sure of the processing gain of XPS and Airtronics.

- John

[KrisW]

	Quote: Originally Posted by Prof. Dr. YoMan
	@Frank: Do you got the URL from rc-network or rcline, where Frank T. measured the power output of the US-DX7 a year or more ago? I cant find it. He clearly measured way more that 500mW peak power, as i remember.

You can have a "peak" of 500mw, but the average signal strength is what matters more. If that peak is an extremely narrow spike in a much wider pulse, looked at with an oscilloscope, then it's kind of irrelevant. However, if that 500mw is across 75% of the pulse width, then it's a substantial amount of power, and more likely to penetrate harsh environments and be "heard" by the receiver.

I do not think I have seen an example of an actual XPS, Spektrum, or FASST information "pulse" shown on the web, yet. During Kiwi's video, we saw an extremely wide signal, that I think reflects an average of the pulse train width during the testing, not an actual single pulse. At the same time, you see a very narrow FASST signal bouncing around the spectrum with these teensy little spikes. Having done a lot of IFF, Radar, TACAN and other Com/Nav work in the past, I expect it would be possible to isolate the information pulse trains to sort out the individual pulses. After all, if I could see the individual IFF pulse trains at 1.03-1.09 ghz with equipment built in the 1960's, I'm sure todays technology allows us to easily pick out pulses in a 2.4ghz fixed frequency pulse train, and analyze them for strength, shape, and any non-square wave characteristics. It would be interesting to see how each manufacturer addresses these items, as well as do a comparative test of pulse strength and width. After all, wider, or taller, will make a signal able to penetrate interference better than narrower or shorter. .

[gareth.ky]

I doubt any of them are still sending an analog pulse train. Its probably all digitally encoded (and possibly encrypted). E.g. On the Spektrum Rx its a Wireless USB chipset. Your in the digital age now, all 01100100011010010110011101101001011101000110000101 101100 now. Why send a pulse train when you can send 11 bits (2^11 == 2048) per channel?

[KrisW]

	Quote: Originally Posted by gareth.ky
	I doubt any of them are still sending an analog pulse train. Its probably all digitally encoded (and possibly encrypted). E.g. On the Spektrum Rx its a Wireless USB chipset. Your in the digital age now, all 01100100011010010110011101101001011101000110000101 101100 now. Why send a pulse train when you can send 11 bits (2^11 == 2048) per channel?

You wold still see the logic pulses being transmitted. A high or low is the logic 1 or 0, so it would be very easy to sort out individual pulses to ascertain their actualy strength. BTW. 1024pcm is only 10-bit, and would be represented by 10 digits, or possibly 12, if that is all they wished to encrypt. They might use 16 bit and use the first 4 as identifiers, then whatever it takes for the rest of the train as information.

A synchronized screen shot would look like:

1 111 11 1 1 11 11 1 11 1 1 111 111 Etc. Etc. Etc.

Or a 16-bit as 1 111 11 1 1 11 11 1 11 1 etc.

But, you would still be able to see the individual pulses.

BTW. . IFF is totally digitally encrypted, but only octal-binary encryption on channels 1,2,3a and C, 1111 is 7, with a maximum encryption of 7777. Of course Mode 4 is a bit different, running encryption from the KY encoder. . but that's all classified. I'm sure they have changed things a bit since I left the military 20 years ago. It's a lot more secure nowadays.

With a bandwidth of 1mhz, you should be able to easily isolate the individual pulses to look at them.

[JKos]

The data is not transmitted as discrete pulses. It is a modulated carrier. The carrier is only on when a packet of data is being sent. The carrier has constant power while on (typically). Each bit of the DSSS spreading code is called a chip. The signal is modulated by the chips XOR'd (typically) with the data.

So, the chain of events kinda looks like this... Data -> Encryption (if applicable) -> Error correction coding -> XOR with spreading code -> Modulate carrier.

- John

[KrisW]

	Quote: Originally Posted by JKos
	The data is not transmitted as discrete pulses. It is a modulated carrier. The carrier is only on when a packet of data is being sent. The carrier has constant power while on (typically). Each bit of the DSSS spreading code is called a chip. The signal is modulated by the chips XOR'd (typically) with the data. So, the chain of events kinda looks like this... Data -> Encryption (if applicable) -> Error correction coding -> XOR with spreading code -> Modulate carrier. - John

So it's pulse modulated FM. You can still see the synchronized signal on a scope, and tell what the amplitude is. The question remains . .what is the actual power. The only way to see is to isolate a block of the transmission and see what it's actual power is.

[JKos]

> So it's pulse modulated FM.

No, something like GFSK, BPSK, QPSK, etc.

Am I missing something because folks have already measured the power levels and they are available in the FCC test reports as well?

- John

[easytiger]

I tried to warn you guys, what, a year ago?
People started losing some serious aircraft, and a very distinct pattern emerged.
I see XPS as more than a ripoff, but rather a real threat. It's still being marketed as being perfect for large and potentially dangerous aircraft. It's just NOT suitable, any more than the Spektrum parkflyer receivers are. You go out and fly a 55 pound airplane on one single frequency on 2.4ghz? A frequency not reserved for models? That's insane.

Look at Peter Michel's twin turbine airliner that went out of control and crashed...that was BAD.


How long before someone really gets hurt?


Thanks to the guys who did all the testing, especially to Xjet, who has known what was what all along, and just got pilloried for it.

All the fanboys, well, there are Holocaust Deniers, too, who just will never be convinced, no matter what.

But all the guys who are actual shills for XPS, who got free gear in exchange for promoting the stuff, and there are many, shame on you. You know who you are. Some are still posting here.

[KrisW]

	Quote: Originally Posted by JKos
	> So it's pulse modulated FM. No, something like GFSK, BPSK, QPSK, etc. Am I missing something because folks have already measured the power levels and they are available in the FCC test reports as well? - John

Pulse modulated FM is FSK, ASK or PAK, depending on what aspect of the signal your are modulating to produce the encoding of the signal. . frequency, amplitude, or phase. All the other modulation styles, quadrature, gaussian, or binary are merely further steps to shift phase or modulation to encode in a different manner.

As for WHY we/I am asking this. . perhaps we are just overly curious. . perhaps we'd like to fully understand the strengths or weaknesses of each system, perhaps we'd like to delve into the theory behind the designs and RF links to ascertain WHY they work like they do, and wshat the shortcomings are of each type and design.

I want to see it on a Scope, with full disclosure of modulation and performance. Just a few traces on a scope, showing an entire channels pattern, aren;t good enough for me.

And, easytiger. I tend to agree with you. . just not in as forceful of language.

[easytiger]

All the guys who spent money and got dissapointed, well...that's between them and Jim. They will have to work that out.


But what is really on my mind is how long before someone gets really hurt? Before a field is lost? How many turbine jets, for example, have been lost due to XPS? A dozen? How many giant scale planes? Out of how many units out there? Not that many at all, in spite of Jim's claims.
If someone is flying foamies or whatever, it does not bother me. The chances of someone getting hurt are pretty slim. But this thing was and still is being heralded as being suitable for any kind of aircraft, and it just is not. It's pretty just a matter of time before they get hit and locked out.
That has been a consistent pattern....10 flights, 25 flights, whatever...people beleive the thing is fine, then pow.
These tests have proven that there are SERIOUS vulnerabilities in the system. You may get lucky, you may stay lucky for a long time. But sooner or later, you are going to find that there are very good reasons why none of the major manufacturers glommed onto such a simplistic solution using off-the-shelf components. Because they were simply unsuitable for the job. In a word...unsafe.

I had a freind who crashed recently...he had about a hundred flights, out to the limits of vision, using the AR6000 receiver. It's clearly labeled as a park flyer type, that you need to keep it close. After 99 flights, this guy was convinced the RX worked fine, right up to 1000 feet. He was telling me that I could use those receivers on my larger planes, instead of getting the 7100s with remotes. He was totally convinced. Right up to that hundreth flight where he got locked out.
So, for all the Dick Corbys or whoever, who say they have hundreds of flights, no problems, well...you're still playing with fire. Keep your XPS planes small, and keep them close, because what you have there is a very limited system, which will sooner or later reveal its weaknesses to you.

But I think people ought to take the same viral marketing approach that Jim Drew used to snowjob so many people, and use it to warn the RC world that they are taking a serious risk if they fly anything but a parkie with XPS.

I hope nobody gets hurt...so far, all that has been lost is tens of thousands of dollars worth of models.

[JKos]

Kris,
I understand what you are asking for. It is mostly already available via documents from Digi and the FCC test report. Check these documents: https://fjallfoss.fcc.gov/prod/oet/f...ive_or_pdf=pdf and http://www.digi.com/pdf/ds_xbeemodules.pdf and see if they answer your questions. Power level, modulation, data rate, etc., and even receiver sensitivity are provided. Even the antenna pattern of the receiver (at least in the plane of circuit board) is available from Digi.

There are still details such as processing gain and the error correction coding used left out but perhaps the receiver sensitivity being given at 1% PER helps fill in some of those details.

I can tell you from observing these systems (XPS, Spektrum, FASST, and Airtronics FHSS) that the tx power is constant for the duration of the transmission.

- John

[KrisW]

	Quote: Originally Posted by JKos
	Kris, I understand what you are asking for. It is mostly already available via documents from Digi and the FCC test report. Check these documents: https://fjallfoss.fcc.gov/prod/oet/f...ive_or_pdf=pdf and http://www.digi.com/pdf/ds_xbeemodules.pdf and see if they answer your questions. Power level, modulation, data rate, etc., and even receiver sensitivity are provided. Even the antenna pattern of the receiver (at least in the plane of circuit board) is available from Digi. There are still details such as processing gain and the error correction coding used left out but perhaps the receiver sensitivity being given at 1% PER helps fill in some of those details. I can tell you from observing these systems (XPS, Spektrum, FASST, and Airtronics FHSS) that the tx power is constant for the duration of the transmission. - John

Xbeepro modules. . geesh

[Four Stroker]

We know what the power levels are but some people dont' buy it. How do we know what the process gain for FASST is ? Paul Beard has claimed 18 dB which is just 10 log(64). From his (CYPRESS) chip spec sheet, he can use a 64 bit code so this is consistent. FASST does their own baseband and we will never get much about it.

[madmax]

Gents,

First off, I want to welcome all of the new members who have come out of the woodwork, and have brought value to this discussion. Welcome to FG! Also, one of our former lurkers hit the nail on the head. This thread is about non-biased testing, not all out bashing. Let's keep the topic exactally to that point, else staff will get on the nuke button. There is some awesome information here for all of us to learn from, let's keep this on point.

Glad to see you all coming in to contribute and thanks again fellas!

MD

[XJet]

	Quote: Originally Posted by easytiger
	All the fanboys, well, there are Holocaust Deniers, too, who just will never be convinced, no matter what.

As others have stated, I think it's important that this thread remains objective.

Those who have "faith" in XPS will never be swayed by any arguments put here so it's probably best we don't bother trying to mount cases for or against -- simply look at the data that's presented and draw your own conclusions.

I'll be posting a URL shortly for an article in which I hope to wrap up the XPS situation (again in an objective but "no punches pulled" manner) so that there will be a place on the web that those looking for informed information and opinion can go to find it -- without all the fanboys versus naysayers "noise" that has accompanied the presentation of such information to date.

In the meantime, we should wait for the rest of the test results to come in and avoid re-hashing long fought battles which really constitute little more than an irresistible force versus an immovable object.