Independent tests prove lack of frequency hopping with XPS

[Kiwi]

Gents.

I wrote to Maxstream asking some questions on the XBPRO module as I have a potential use for that type of device for data streaming information off mobile mining equipment. Unfortunately my mail server has quit tonight for some reason but as soon as it comes back on line I will copy and post the reply.

I asked the question from them as to what happens if the data stream is interfered with to the point that I start to get data errors. The application is dependent on a 100% secure data link. Currently done with cables but I want wireless.

I wont comment other than to say the information I received says the system cannot hop frequencies but the answer is over my head and I would like someone who does this for a living to decipher the reply.

As for my testing being non scientific I always and will continue to say I tested it in a modellers application. That is in a noisy environment with other systems operating in close vicinity. To me if the system has to go into failsafe to prompt a channel hop then how can the receiver tell the Tx where it is waiting if the two have no pre programmed idea of where in the heck to go. Fail safe has to be the last option, it cannot be a wait or hold alternative while the Rx and TX go there own seperate ways. The time it takes to relink even if it is only a second or two is way too long. To my mind any system that claims to be able to dodge noise and find a cleaner channel to maintain the link on must happen before the fail safe option comes into play.

I´m not bashing XPS, I just cannot make it hop by slowly raising the noise all around it or rasing the noise rapidly.

FUTABA FASST absolutely does what they claim and you can see that on the scanner.
SPEKTRUM does absolutely what they say it does. It runs on two spaced channels.
ASSAN does exactly what they say. It runs on one pre selected channel, it does not hop, they dont claim it can either.

But I got my Spektrum setup this week so I now have nearly all the available systems. They will all get the same treatment. Tested in a RC model friendly environment.

I have no pre conceived ideas or answers on how they will perform. Time will be the judge of that.

[Four Stroker]

Don't forget the Airtronics RDS8000 !

[XJet]

	Quote: Originally Posted by Kiwi
	Gents. I wont comment other than to say the information I received says the system cannot hop frequencies but the answer is over my head and I would like someone who does this for a living to decipher the reply.

I'm pretty sure their reply will be that the XBeePro modules will not hop by themselves but that the operating channel can be changed by the user's own software (firmware) if suitably written.

The problem with this is that we're trying to perform a transport-level reconfiguration at an application-level. In effect, it's a bit like trying to drive your car from the back seat -- the amount of control you've got is rather severely constrained.

	Quote:
	As for my testing being non scientific I always and will continue to say I tested it in a modellers application. That is in a noisy environment with other systems operating in close vicinity. To me if the system has to go into failsafe to prompt a channel hop then how can the receiver tell the Tx where it is waiting if the two have no pre programmed idea of where in the heck to go.

Yes, this is the biggest problem with hopping as implmented (or not) by XPS.

You can't negotiate where to hop if you can't talk to the other party.

JD bases his hopping algorithm on the assumption that an interfering signal will always appear at a low level and very gradually increase in intensity -- allowing plenty of time for his transmitter and receiver to detect the rising noise and renegotiate an alternative channel before all communications between them is lost.

It's been my experience however, that some of the worst interference on 2.4GHz simply appears out of nowhere and gives no warning of its arrival.

Remember that we're dealing with solid-state devices here. When someone turns on their video sender, leaky microwave or other 2.4GHz noise source then, just as when you turn on your LED flashlight -- the signal is there instantly and at full intensity. XPS can't cope with this kind of interference at all because there's no chance to renegotiate the channel in use -- it's a limitation of the XBee technology provided by MaxStream and there's nothing JD can do to change it.

JD says that the only time you get this "sudden" appearance of high noise levels is when the entire band is saturated -- which is patently untrue and I've got plenty of SA logs to show otherwise. Around where I do my UAV testing it's not at all uncommon for very high levels of noise to come and go in an instant on various segments of the 2.4GHz spectrum -- but I've *never* seen the entire band saturated.

I suspect the reason that JD has simple proffered an inadequate explanation and won't provide any format for a reproducable test scenario that proves XPS hopping is that he's hog-tied by the limitations of the RF modules he's using and can't.

It should be remembered that the XBeePro modules aren't designed for hard-realtime or RC model applications -- hence the fixed-frequency operation and absence of any diversity capabilities.

JD has taken some off-the-shelf modules, crafted up some relatively simple microcontroller code and grafted the two together into a module-based 2.4GHz system. It works for many people but is hardly the pinnacle of robustness from an RF or realtime perspective.

If it were, you'd have already seen a raft of other systems launched using the very same concept. In fact, I was doing some documentation for another company who was planning to do just this but they decided to cancel the project at a late stage when their testing showed issues created by the MaxStream XBeePro modules that could not be adequately addressed to their satisfaction.

[KrisW]

	Quote: Originally Posted by BZFrank
	A 'lockout' is a situtation when the RX does not respond to the TX anymore for a substantial time duration. There is no special mode for that in a XPS system (at least it would not make any sense to implement such), it is believed an error condition of the XPS due to various causes - low voltage, bad reception, channel swamping or triggered by other external events. 'Lockouts' in XPS are reported both be transitory when control could be regained after some seconds and permanent when control could not be regained until crash. Here is a video example of a transistory XPS lockout (German narration, however its easy to tell when its happens: 'kein Empfang' means 'No reception', 'Empfang' means 'reception', its was not a low voltage issue as a logger was running during flight): YouTube - rc jet Kangaroo 2007-12-23 XPS LockOut ? Frank

first off, those German guys are very hard core modelers. . cold. . windy .. yuck

Second, that is one short, and narrow runway, for a turbine job

Third, the guy flying did a great job getting the plane down in one piece.

And lastly. . the radio system LOCKED OUT, several times. they were very lucky to get the plane back and be able to fly it another day.

[KrisW]

double post, sorry

[Prof. Dr. YoMan]

	Quote: Originally Posted by Kiwi
	FUTABA FASST absolutely does what they claim and you can see that on the scanner.

Except that they switch frequency every 8ms only, but not as advertised every 2ms.

They send out 2ms long Bursts every 8ms.

[BaldEagel]

	Quote: Originally Posted by aviti
	My point is who cares if it CAN hop. If it doesn't hop at the point it is needed then the plane is in the ground. I don't know if it needs to hop or not but with all the controversy surrounding this deal, I'd rather not find out with my plane.

I have to admit, I also don't care if it hops or not, what I don't want to be doing is flying an aircraft that keeps locking out on me, which is what happened to me on my interferance prone site, just like flying PPM with lock out instead of glitches.

Mike

[xed]

	Quote: Originally Posted by BaldEagel
	I have to admit, I also don't care if it hops or not, what I don't want to be doing is flying an aircraft that keeps locking out on me, which is what happened to me on my interferance prone site, just like flying PPM with lock out instead of glitches. Mike

Well if you do not care if it hops than you should also not care if your plane goes nose first into the ground either because that is what will happen when XPS encounters interference.

XPS is equivalent to an expensive PCM receiver running on one channel that may co-exist with other 2.4GHz systems until it can no longer cope with the level of noise on the startup channel it settled upon when you first turned on your XPS system. So yes, it will be fine to fly on it until another system trounces all over your channel.

Would I care if it was a foamy? Nope.

Would I care if it was a giant scale flying hunk of a plane with a metal missile strapped to the nose? Hell Yes!

Buyer beware...

[Four Stroker]

I ordered a really cheap Spektrum Analyzer here:

http://www.dunehaven.com/minisa.htm

Although not as fancy as the WiSpy, it has about the same performance. I will report if it works as advertised. If not, it was $96 total shipped.

[klhoard]

	Quote: Originally Posted by xed
	. . .<snip>. . Buyer beware...

.
.
I don't really care about the (informed) buyers, its the innocent bystanders that I'm worried about!!
.
.

[Flyfast1]

Even if the XPS system does not change channels in response to interference on the channel currently being used, it's at least as safe as conventional 72MHz systems. I think it's actually much safer than these systems since the transmitter is bound to a particular receiver, so that data from other systems on the same channel can be ignored. It also retransmits information several times without having to wait for the next frame. It also has failsafe functionality that many PPM systems do not have. But, from my own experience, the receiver is more sensitive to installation location and orientation than convention 72MHz receivers, so I think more time needs to be spent on installation and range testing. I am also assuming that an appropriate on-board electrical power supply system is used.

I also had to make a comment on that video. Wow, those guys are hardcore! And I thought my runway was narrow! That pilot did a great job of bringing that jet down.

-Ed B.

[JimC-MD]

	Quote: Originally Posted by XJet
	Strangely enough, it didn't have an XPS module in it at the time -- but come to think of it, I don't think that would have helped -- it wouldn't have hopped out of the way of the oncoming vehicle anyway:-) It's not my only transmitter -- but I did have a few regularly-flown models programmed into it and it was half of the club's buddy-box setup so there won't be any training going on around here until it's replaced.

Xjet,

This thread has been a chore to digest at times, but that has to be the post of the month

Sorry to hear about the training being on hold. I hope it gets worked out soon.

[Wojcigitty]

	Quote: Originally Posted by Flyfast1
	Even if the XPS system does not change channels in response to interference on the channel currently being used, it's at least as safe as conventional 72MHz systems. I think it's actually much safer than these systems since the transmitter is bound to a particular receiver, so that data from other systems on the same channel can be ignored. It also retransmits information several times without having to wait for the next frame. It also has failsafe functionality that many PPM systems do not have. But, from my own experience, the receiver is more sensitive to installation location and orientation than convention 72MHz receivers, so I think more time needs to be spent on installation and range testing. I am also assuming that an appropriate on-board electrical power supply system is used. I also had to make a comment on that video. Wow, those guys are hardcore! And I thought my runway was narrow! That pilot did a great job of bringing that jet down. -Ed B.

Yes, it is able to coexist on the same frequency as another 2.4 ghz device, and it is a lot "smarter" in how it's bound receiver-to-transmitter.

However, it's to be expected that, out of nowhere, another device could start broadcasting on the same frequency as XPS and use up all available bandwidth. FASST and Spektrum have methods of dealing with this.

It's NOT expected that all of a sudden another device could suddenly take a single frequency on 72 MHz without warning. 72MHz systems are not designed to be frequency-agile, and when properly used (frequency control is implemented), they do not need to be. 2.4GHz systems ARE supposed to be frequency-agile, and even when used properly, they DO need to be.

[Wojcigitty]

	Quote: Originally Posted by BZFrank
	(German Kangaroo sub-zero death-defying flight of death) Frank

Hard man fly turbine in ice and snow.

Hard man easily deal with lockout and land jetplane on cratered minefield runway.

XPS for hard man. Yeeeeeaaaah.

[aviti]

Looks to me like any landing with a turbine on that runway would be considered a crash so who cares if you have lockout. You crash no matter what.