Independent tests prove lack of frequency hopping with XPS

[jonkoppisch]

	Quote: Originally Posted by tadawson
	... So, for me, it comes down to this: ONLY TESTS DONE AT IN-FLIGHT SIGNAL LEVELS HAVE ANY CREDIBILITY. THE REST ARE FAR TOO PRONE TO OTHER ISSUES TO BE DEFINITIVE IN THEIR CONCLUSIONS.... - Tim

Hey Gordito Volador,
Get ahold of the guy that just lost his plane and tell him that the tests have no credibility...


http://www.rcgroups.com/forums/showt...18#post8770718

	Quote: Originally Posted by JimDrew
	We have the only system that communicates noise in the environment and in fact WILL hop to a cleaner frequency if the noise floor rises. That hop may not even be due to the aircraft experiencing a problem, it could also be from the transmitter side (like other 2.4GHz systems on the flight line). ......

[Toumal]

	Quote: Originally Posted by tadawson
	That has not, and cannot be proven with any of the test gear in that video. The saturation being discussed here is RX front end - not band, which cannot be seen on any external test equipment - you have to monitor the internals of the RX to see it.

That is incorrect. If the transmitter's power levels were high enough to saturate the receivers front-end, it would not matter whether it was transmitting on the right channel or an adjacent one. Since the video clearly shows XPS stopping to work the moment the interfering TX occupies the exact same channel, saturation is highly unlikely.
EDIT: Unless the receiver has a super-selective front-end... Like, oh, some military gear.

And again, for the record, you don't "saturate the band", saturation is a property of the receiver.

Besides, FASST and Spektrum were also tested and were not "saturated" either. They kept working, and that despite JD claiming that XPS is less susceptible to saturation (-well, due to the fact that he held the receiver further away from the TX...) in his own video. So regardless of whether you believe me that saturation does its fatal work even on adjacent channels or not, there's still the fact that the other two systems were not experiencing any difficulty whatsoever.



Furthermore, I agree with Kiwi that the discussions here have become an exercise in futility, and apparently that's the plan. The amount of half-knowledge is astonishing, and it's disturbing to such statements to be put out into the wild, contradicting the laws of physics, contradicting logic, and contradicting facts upheld by video evidence and controlled tests. I know there's room for debate on a variety of things, but when people make unfounded statements as to whether or not a groundplane monopole antenna has a significant blind spot or not, talk about saturating a frequency band, and make illogical statements as to whether or not hopping on interference happens under super-special circumstances only... that's not a debate, that's a waste of time.

Here's the deal: I don't tell you how to set up a gasoline or glow engine, because that's something I have little clue about, and you don't tell me that the stuff I studied for years and the things I work with all the time are wrong and all my experience with RF tech just an illusion. If you're telling someone that flying with a single 2.4GHz groundplane monopole antenna in a model with a big lump of an engine block is no problem and that the antenna's "dimple" is insignificant, you're potentially endangering the lives of other people. If you're telling someone that frequency hopping under "special circumstances" is enough while confusing interference with saturation, that's irresponsible. This is not a "no, you're wrong"-contest. A lack of reliability, especially when the manufacturer is in the phase of denial, can cost lives.

I hope the moderators do the right thing and prevent the "everything's fine with XPS"-messiahs from turning this forum into an Irrlicht for people who seek factual information. I don't want this thread closed either, but I think an IP check would be a good thing to do.

[tadawson]

	Quote: Originally Posted by Toumal
	Here's the deal: I don't tell you how to set up a gasoline or glow engine, because that's something I have little clue about, and you don't tell me that the stuff I studied for years and the things I work with all the time are wrong and all my experience with RF tech just an illusion.

Well, since you brought it up . . . My background is as a degreed engineer in communications and solid state electronics. Not to be a prick, but everywhere I have been, engineer > technician when it comes to understanding this stuff. So, I think I have a pretty darn good grip, and no amount of waffling that I have yet heard has in any way invalidated my concerns with the shortcomings of these tests . . .

- Tim

[XJet]

	Quote: Originally Posted by tadawson
	Well, since you brought it up . . . My background is as a degreed engineer in communications and solid state electronics. Not to be a prick, but everywhere I have been, engineer > technician when it comes to understanding this stuff. So, I think I have a pretty darn good grip, and no amount of waffling that I have yet heard has in any way invalidated my concerns with the shortcomings of these tests . . . Tim

Like yourself Tim, I'm a EE who's been in this game (design, development, manufacturing, maintenance) for far more years than I care to admit. I started my training in RF back in the mid 1970s and just a few years later became strongly involved in digital control/telemetry systems.

Even "experts" can disagree occasionally :-)

In Kiwi's tests, the fact that the system continue to operate when the interfering signal was outside the receiver's operating frequencies shows that the receiver's front end was not overloaded to the point where it could not have performed a scan.

Typical front-end overload causes receiver sensitivity to plumet *and* often results in intermod products that cause an off-frequency signal to block an on-frequency signal due to those IM artifacts. If there's AGC involved, it can also kick in and simply wind down the entire sensitivity of the RF chain.

You know that but others might not.

The fact that XPS continued to operate *until* the interfering signal transgressed onto the portion of the band that the receiver was using indicates that there was no significant front-end saturation.

Remember also that the front end of these 2.4GHz systems is tuned *very* broadly (80Mhz) so there's no in-band, off-frequency rejection there.

If the XPS receiver was still able to "hear" the desired signal when subjected to an in-band/off-channel interfering signal then the front end wasn't saturated or significantly desensitized by that interfering signal. This means that *if* JD's hopping worked, he could have done a band-scan, located a free portion and jumped to it.

The reality is that without the ability to negotiate new frequency, it's just not possible to reliably hop. Until the Tx and Rx can agree where to hop, there'd be no point in trying. This is an intrinsic limitation of any system built around a single frequency RF link with a single receiver.

That's why I have always disputed JD's claims right from the start -- it simply CAN NOT be done with the XBeePro modules.

Spectrum has already negotiated a fall-back frequency and FASST has an already negotiated hopping schedule so they can handle interference.

By the time XPS realizes that it's been hit by interference - it's already too late to hop because neither end can inform the other where to go. Any system that relies on negotiating a hop before the link is lost on the current frequency runs the risk of losing sync and thus any hope of restoring control in less than several seconds (ie: crash).

Note also that when this whole issue blew up, JD promised he'd upload a video proving that XPS would hop.

What happened to that?

Instead, he tried to change the subject by uploading a clearly "rigged" video showing how the competition would stop working due to saturation when held right beside the antenna of an interfering source -- while XPS continued to work (when it was held at a distance that would have produced a field around ten times *LOWER* than the others were subjected to.

If it really did hop, why didn't JD just spend his time uploading the video evidence?

The answer: because the damned thing will *only* hop when his very stringent (and totally unrealistic) scenario of a slowly rising noise floor over 15 seconds is experienced -- a scenario that even he was clearly unable to duplicate on the bench.

Now as I said before, if JD has gotten DigiChip to rewrite the XBeePro firmware to support true satellite receivers and perhaps he opts for a dual-frequency multiplexed transmission then XPS may turn into a very trustworthy and useful system.

But that won't recover the huge loss of trust that has been produced by his simple refusal to admit that he's grossly overstated the frequency agility of this system.

[tadawson]

Not much to argue with there, but what has not really been considered, is what is the effect of the interference signal on the receive side of the RX module? Once again, with everything in such close proximity, I don't think that is something that can be ignored. Also, the intermod products will be different being right on top vs. being on an adjacent channel, and not knowing the details of this receivers internals, do we know where those will fall, since we are dealing with pretty wideband signals, but close in freq? Would that also not tend to make the IM products fall much closer to the main signal than with a different channel of the interference signal? Nothing certain here, but I still have to wonder . . . Do you not agree, at the very least, that the test would have been better done at closer to in-flight RF field strengths? If nothing else, it would remove what (at least in my mind) I see as a potential variable to the result . . . it may not change anything, but then again . . . we can be sure unless that is tried as well, no?

- Tim

[Toumal]

	Quote: Originally Posted by tadawson
	Well, since you brought it up . . . My background is as a degreed engineer in communications and solid state electronics. Not to be a prick, but everywhere I have been, engineer > technician when it comes to understanding this stuff. So, I think I have a pretty darn good grip,

Tell you the truth, I'm an engineer in communication ("Nachrichtentechnik") but that doesn't really make me an expert in these things. I'd say that building your own 2.4GHz "Pringles"-antenna can be a very helpful learning experience, but then again they also forced me to do field calculations and stuff. Stuff like "How much dB loss do you get if an antenna with so and so many inches emitting a x MHz frequency points straight at you, compared to a 90 degree elevation angle" was actually a typical test question, and I've experienced the effects of such things first-hand. But hey, don't listen to me, listen to XJet and Kiwi, who probably built their own 800W CB amps when I was still playing with Legos.

	Quote: Originally Posted by tadawson
	and no amount of waffling that I have yet heard has in any way invalidated my concerns with the shortcomings of these tests . . . - Tim

I think we can agree on two things:

1) It would be great if a future test were held at larger (and potentially more realistic) distances - i.e. 50 meters between RX and TX, interfering TX also 50 meters away. Alternative test setups should be explored also, to simulate the aircraft being far away from the pilot but relatively close to an interfering transmitter, for example.

2) The distances as well as a top-down layout should be recorded so that the test is repeatable by others under the same circumstances. Transmitter power should be measured as well.

	Quote: Originally Posted by XJet
	That's why I have always disputed JD's claims right from the start -- it simply CAN NOT be done with the XBeePro modules.

I don't know if that's really the case! Think about it: TX and RX could negotiate a hopping scheme right after powerup. That scheme could cover all channels and could then just be run through in a sequence that's also pre-negotiated. XPS could stay on a channel until there's no ACK from the receiver (TX side) or no more new frames from the transmitter (RX side). In either of these cases, the transmitter and the receiver could proceed to the next frequence in a fixed time window fashion - that way you would avoid having the RX start the hop sooner than the TX, and so forth. Or, the system could hop all the time! That would make things a lot easier too.

...or is there some limitation in XBee that I'm not aware of?

[ss40]

I am not going to deny anything. I am just going to restate. XPS works for me ...for close to a year now.


And 5-7 lines of code is definitely Gartner type crap. Almost everyone I worked with and for produced more, way more and that is a fact. I would not have been as successful as I was over 30 years in software development if I had not. Even in the early 70's I maintained a system, developed by a half dozen people in a couple of years and it had more than a million lines of code, and that was high level coding.

Go ahead and spread your doom and gloom and say the sky is falling. If I listened to all the experts I ran into in my life I would be either very poor and homeless or dead. Having my own mind and investigating stuff on my own has paid off in spades for me. So the experts can just keep singing the same song. I am not a fanboy. I am an XPS user that finds that it works for me and no one can refute that.

By the way I watched people code report programs over weeks with debugging...I worked with consultants that spent a couple of weeks writing code, in the same language, that generated report programs in about an hour's effort. We sold the resulting programs for the same price as the other guys. We made a lot of money and they thought we were going broke. That code was being used 20 years ago and it is still being used to advantage today.

PS: Someone I know was flying with Spektrum only a couple of times. He had some"hits" and you could see his confidence was disappearing quickly. In the end a good look at the model showed a bad servo on the elevator and Spektrum proved completely solid, but it would have been all over locally here how that brand was crap if it hadn't have been investigated. And if there had been a crash instead of saves on a couple of flights there would have been no way to investigate. That's how easy it is to blame the wrong thing!

[Wojcigitty]

The issue here isn't XPS's track record. The issue here is not "XPS will not work." If you use XPS, fine. Have fun.

The issue here is that XPS DOES NOT HOP IN A USEFUL FASHION, despite JD's claims. Glad to hear that it's working for you, but that's not what we're trying to discuss.

[JEFFRO503]

	Quote: Originally Posted by ss40
	I am not going to deny anything. I am just going to restate. XPS works for me ...for close to a year now. And 5-7 lines of code is definitely Gartner type crap. Almost everyone I worked with and for produced more, way more and that is a fact. I would not have been as successful as I was over 30 years in software development if I had not. Even in the early 70's I maintained a system, developed by a half dozen people in a couple of years and it had more than a million lines of code, and that was high level coding. Go ahead and spread your doom and gloom and say the sky is falling. If I listened to all the experts I ran into in my life I would be either very poor and homeless or dead. Having my own mind and investigating stuff on my own has paid off in spades for me. So the experts can just keep singing the same song. I am not a fanboy. I am an XPS user that finds that it works for me and no one can refute that. By the way I watched people code report programs over weeks with debugging...I worked with consultants that spent a couple of weeks writing code, in the same language, that generated report programs in about an hour's effort. We sold the resulting programs for the same price as the other guys. We made a lot of money and they thought we were going broke. That code was being used 20 years ago and it is still being used to advantage today. PS: Someone I know was flying with Spektrum only a couple of times. He had some"hits" and you could see his confidence was disappearing quickly. In the end a good look at the model showed a bad servo on the elevator and Spektrum proved completely solid, but it would have been all over locally here how that brand was crap if it hadn't have been investigated. And if there had been a crash instead of saves on a couple of flights there would have been no way to investigate. That's how easy it is to blame the wrong thing!

This is a good thing you haven't lost an aircraft to XPS! The place where you fly might be at a very low noise area , which hopefully gives you years of trouble free flying. Some of us aren't so lucky. If you we're to take a bunch of time , you could read through this entire thread and see that there have been a bunch of reports of people losing their airplanes , and we're ALL having probelms with the XPS system ( Except Sweatpea.....who mentioned he never really could prove it.)


As big as this thread has gotten , it would probably take a couple days to get through it , but it is a great read for sure.

I'm not an expert of FHSS , and in fact i know very little about it. Just from everything i have read here on this thread and some other threads on other sites , i'm starting to get a grasp on how this stuff actually works. I probably have about 1% of the knowledge that some of these guys have when it comes to this stuff. BUT.....I DON'T need to be an expert to see that my aircraft is having probelms. I consider myself a pretty good pilot , and can see when things are not working correctly.

The only reason i ever got involved in this thread was because i wanted to see if "ANYONE" could make XPS hop......so far at 103 pages of debating.......NOBODY has ever made the XPS system hop....EVER! Only a claim by JD , that he had gotten it to hop , and AGAIN........no proof what so ever.


Now.....IF you we're to read through this entire thread , you would read about more aircraft meeting their doom to an XPS failure then any other system out there. In fact , there are more completely destroyed aircraft in this thread alone , than all the other systems combined , in any other threads i could possibly find.


There are some guys whom jumped ship over here to FG , from RCG , because they we're getting flamed for stating problems , or just asking a question.......Hell , they we're lucky to even be able to keep their post up..........JD would just delete it!

OF COURSE you read about more XPS failure here...........JD isn't here to delete the post.

God....what a horror story!

[wrightme]

	Quote: Originally Posted by Wojcigitty
	The issue here isn't XPS's track record. The issue here is not "XPS will not work." If you use XPS, fine. Have fun. The issue here is that XPS DOES NOT HOP IN A USEFUL FASHION, despite JD's claims. Glad to hear that it's working for you, but that's not what we're trying to discuss.

Maybe that is not what YOU are trying to discuss, but many people here ARE discussing how "wrong it is to even THINK about using XPS...." Not everyone here holds your relatively benign view. Xjet himself spoke in the XPS forum noting the "less than warm welcome" (to paraphrase) that those who use XPS receive here.

According to several here, reports of XPS can't be anything except a post from Paid Shills! Plenty of those here have done nothing to engender any feeling of attempting to FIND an actual problem with XPS. Instead it almost always degenerates to "Hops/doesn't hop," and real world troubleshooting of a system moves offline where no one benefits from cooperation. Wocketman just did it again in the XPS forum, adding a claim of "no hop" to a thread about telemetry.

Those of you who hammer on about how bad XPS is are really transparent in your vendetta. You all should really consider giving it a rest. Frankly, you give FG a bad name.

[Wojcigitty]

	Quote: Originally Posted by wrightme
	Those of you who hammer on about how bad XPS is are really transparent in your vendetta. You all should really consider giving it a rest. Frankly, you give FG a bad name.

To be fair, if I bought XPS and lost an airplane to it, I'd probably have a vendetta as well.

These threads have been very enlightening, but at a hundred pages, it's getting redundant (except for Toumal's "In first approximation, a cow is a spherical object that emits equal amounts of milk in all directions.")

[Woketman]

Or maybe those who bought XPS just are p_ssed about being lied to by the manufacturer.