Independent tests prove lack of frequency hopping with XPS

[XJet]

	Quote: Originally Posted by Wojcigitty
	Could JD have used TWO XBPro modules in concert on different channels, with some circuitry on the receiver side to decide which one to use?

There is a problem with that...

The transmitter in the XBeePro module is relatively powerful (100mW ERP) and having two modules that close to each other (even if on different channels) would likely cause issues related to one receiver affecting the other when it was transmitting its ACK packet back.

The situation would be even worse at the transmitter end where, in order to work properly, you'd also need two modules operating in *very* close proximity to each other.

This dual receiver model works okay with other 2.4GHz systems because the receivers are not also transmitters -- but imagine what happens when one receiver is busy transmitting an acknowledgment while the other is trying to receive a data packet. Even if they're on different frequencies, that powerful transmitter right next to a sensitive receiver has the potential to cause problems -- just like bringing a working GSM mobile phone near to an AM radio can produce a loud buzz -- even though they're hundreds of megahertz apart in frequency.

Once again, it's the reliable transport layer of the XBeePro module that becomes a handicap rather than an asset in this situation.

The larger the blocks you use to build something the less time it takes and the easier it is. However, the price you pay is a significant loss of control over the low-level details -- and that's the trade-off that XPS has made in opting for the XBeePro modules rather than the lower-level chipsets that other vendors are using.

[E_ferret]

	Quote: Originally Posted by Daemon
	If the Tx loses ACKs for some interval (Tx module isn't programmable, so who knows if it'll match the failsafe interval), it hops to backup. ian

Hi Daemon,
A hop based on tx/rx timeouts is feasible but like you describe, very time consuming.
The coms scenario seems to be very basic, transmit, wait for ack, if timeout then retransmit x number of times during frame period. If Nack retransmit.
I can see no reason why a retransmit reason code cannot be included in the data packet . That way the rx becomes aware that the tx did not get the ack/nak back and that the link is frail, before a link timeout period expires.
eg. reason code 0 = good packet, original transmit
reason code 1 = you send me an 'not acknowledge'
reason code 2 = ack/nak timeout occured for the first time
reason code 3 = ack/nak timeout occured for the second time
etc...
this should put a flea in the ear of the rx.

[zoomer260]

So many on this thread are so much more knowledgable about radio than others (ME). But it has been mostly kept in laymans terms that a simple electrician can grasp the "points".
I've really enjoyed this thread not because of XPS hopping or not, who cares at this point ? But because I've learned more about 2.4 than I ever would have had there not been XPS. Thanks XPS !

I'm a guy that likes to understand the "logic" used in creating a new system and why it came about and then can it be trusted.

Had this thread not been started I would know next to nothing about 2.4.
Thanks XJet !!

I wonder if most viewing this thread have come to believe (as I have) that a single channel system operating in the 2.4 range is not the way to go even if there were no alternatives in that range ? I mean, am I getting it ?

[KrisW]

	Quote: Originally Posted by zoomer260
	So many on this thread are so much more knowledgable about radio than others (ME). But it has been mostly kept in laymans terms that a simple electrician can grasp the "points". I've really enjoyed this thread not because of XPS hopping or not, who cares at this point ? But because I've learned more about 2.4 than I ever would have had there not been XPS. Thanks XPS ! I'm a guy that likes to understand the "logic" used in creating a new system and why it came about and then can it be trusted. Had this thread not been started I would know next to nothing about 2.4. Thanks XJet !! I wonder if most viewing this thread have come to believe (as I have) that a single channel system operating in the 2.4 range is not the way to go even if there were no alternatives in that range ? I mean, am I getting it ?

The more channels the merrier, IMHO. I wish Spektrum had a firmware update that allowed the user to manually switch channels in mid flight. It would be simple to initiate this using the Bind button and a double push for the function.

[zoomer260]

	Quote: Originally Posted by KrisW
	The more channels the merrier, IMHO. I wish Spektrum had a firmware update that allowed the user to manually switch channels in mid flight. It would be simple to initiate this using the Bind button and a double push for the function.

You know when I got into computers the guy that set me up, advised me etc. was doing computers since 'Nam and loads up THREE credit cards at the "show" each year with stuff that isn't even in production yet. He is in school forever (literally) on CAD and other programs and he is one of the guys the big boys send BETA copies to find the problems with them. He does not write programming just finds what can make it quit working and suggests features that should be intergrated and those that are crap.

Are there any guys out there that have this kind of relationship with the software design on RC radio systems ?
Seems to me that they would at least ask their sponsored pilots for imput such as your suggestion to change channels on the fly.

[dick hanson]

	Quote: Originally Posted by zoomer260
	You know when I got into computers the guy that set me up, advised me etc. was doing computers since 'Nam and loads up THREE credit cards at the "show" each year with stuff that isn't even in production yet. He is in school forever (literally) on CAD and other programs and he is one of the guys the big boys send BETA copies to find the problems with them. He does not write programming just finds what can make it quit working and suggests features that should be intergrated and those that are crap. Are there any guys out there that have this kind of relationship with the software design on RC radio systems ? Seems to me that they would at least ask their sponsored pilots for imput such as your suggestion to change channels on the fly.

The theory on how 2.4 works --for our use -certainly runs the gamut on how it might work and how it really works -- all part of thr fun of discussing it . In suggesting that the Spektrum (DSM2) might be able to change channels whilst flying - suggests a little more looking into the DSM2 is in order:
basically you can't do a "new set of channels unles the system is at low power - and the search for new channels shows a solid set of available ones -also the code for Model Match would be gone - maybe I Missed the objective of th conversation -in that case - please ignore my comments

[Flyfast1]

	Quote: Originally Posted by zoomer260
	I wonder if most viewing this thread have come to believe (as I have) that a single channel system operating in the 2.4 range is not the way to go even if there were no alternatives in that range ? I mean, am I getting it ?

I believe that the XPS system is superior to conventional 72MHz systems for several reasons. First, upon startup the system does scan the available channels and pick a clear one. With 72MHz, one has no way of knowing how noisy the channel is, other than doing a range check. Second, the transmitter is bound to the receiver, which eliminates the problems associated with someone else turning on a transmitter on the same channel. This second reason is important to me. I fly quarter scale warbirds that are not only expensive, but take forever to build. Even if money is no issue, you just can't go buy a replacement. Many of them truly are one of a kind planes. This has always concerned me at large warbird fly ins. Most of these events have excellent transmitter impound procedures and the pilots are very careful and respectful of others. It's when I walk by a trailer full of planes and see five transmitters sitting there that I get nervous. Third, the XPS system does have the ability to retransmit frame data several times before the next frame, which should, at least in theory, reduce the number of bad frames. Fourth, the system does provide two-way communications, which at some point should allow the transmission of telemetry information back to the transmitter. This has yet to be implemented, but I really hope it does.

So, as we have seen, the XPS system has its limitations and I really hope they find a way to address the channel change situation. It would also be great to have a model match feature. But, I think it's a lot better than conventional 72MHz systems and if I could only have this 2.4 GHz system, I would take it any day. I am a current XPS user and have been very satisfied with the system. I was an early adopter and have made a significant investment, so I will continue to use the system, but I am keeping my eye on the other offerings.

Yes, this thread has been excellent and I am thankful for the contributions of XJet and others.

-Ed B.

[Daemon]

	Quote: Originally Posted by dick hanson
	The theory on how 2.4 works --for our use -certainly runs the gamut on how it might work and how it really works -- all part of thr fun of discussing it . In suggesting that the Spektrum (DSM2) might be able to change channels whilst flying - suggests a little more looking into the DSM2 is in order: basically you can't do a "new set of channels unles the system is at low power - and the search for new channels shows a solid set of available ones -also the code for Model Match would be gone - maybe I Missed the objective of th conversation -in that case - please ignore my comments

For whatever it's worth, Spektrum *can* change channels in flight. It's easy. Switch
the Tx off and back on. It'll scan for new channels every time. The Rx will lose the link for
about 1 second, and then start scanning for whatever new channels the Tx is on
and re-link, which usually takes less than 2 seconds on anything but the AR6100 Rx
(which based on my testing may simply fall from the sky, before link is re-established )

ian

[erichevy]

	Quote: Originally Posted by Daemon
	For whatever it's worth, Spektrum *can* change channels in flight. It's easy. Switch the Tx off and back on. It'll scan for new channels every time. The Rx will lose the link for about 1 second, and then start scanning for whatever new channels the Tx is on and re-link, which usually takes less than 2 seconds on anything but the AR6100 Rx (which based on my testing may simply fall from the sky, before link is re-established ) ian

This is not my expierience!!!!
If I turn my radio off/on too quick it can take up to 10 15 minutes to re-bind!!
That's with my Futaba 9C with spektrum 8 channel module.
The led in the module will keep blinking when switched too quick! I've try'd that many times.
It will only (most of the time)succeed if you turn the radio off for around five seconds before turning on again. Then it will rebind as normal.
And sometimes 5 seconds is too quick to turn on, again.... it CAN take a very long time then to re-bind.
Be very carefull with this, don't try this while being airborn!!!

Eric.

[dick hanson]

	Quote: Originally Posted by Daemon
	For whatever it's worth, Spektrum *can* change channels in flight. It's easy. Switch the Tx off and back on. It'll scan for new channels every time. The Rx will lose the link for about 1 second, and then start scanning for whatever new channels the Tx is on and re-link, which usually takes less than 2 seconds on anything but the AR6100 Rx (which based on my testing may simply fall from the sky, before link is re-established ) ian

I misunderstood the original comments - yeh -I have also done the power off/on in flight - an absolutely stupid thing to do as it may and like will crash a model.
I was thinking of trying to switch to a new "Model code" in flight - Don't be so hard on the 6100! get a new one -try your tests again.

[Costas]

	Quote: Originally Posted by erichevy
	If I turn my radio off/on too quick it can take up to 10 15 minutes to re-bind!!

Turning the Tx On/Off is not really a good test as the Tx module actually needs to boot up before it scans and transmits. Firmware could be changed so that the module could rescan and retransmit on a new pair of channels and that would be a lot quicker.

Mind you 10 to 15 minutes sounds odd. My Spektrum module installed in my 10X only takes less than about 5 seconds to rebind when I switch it on then off at the Tx.

Is it possible that you have some other 2.4GHz radio sources (non RC) in the close vicinty?

[Daemon]

	Quote: Originally Posted by dick hanson
	Don't be so hard on the 6100! get a new one -try your tests again.

Someone (Rob Hair) over on RCG confirmed that the low-power reboot/re-link is fast with
the newest firmware AR6100, but that the "long time" (about 1 second) loss of signal re-link time
is unchanged, and is still very slow compared to any of the other Spektrum Rxs.
Pretty sure the difference is that it's the only with only one Rf receiver, AR6000 (DSM1), and AR7000 (DSM2)
and others all have more than one. I think that when they scan, they keep one
receiver on the old channel pair, while the other receiver scans, so if the signal returns
on the old channels, it can hot link immediately. The AR6100 only has one Rx
and it can either listen on one pair of channels, or scan for new ones but not both,
thus the slow re-link time.

ian

[Daemon]

	Quote: Originally Posted by Costas
	Mind you 10 to 15 minutes sounds odd. My Spektrum module installed in my 10X only takes less than about 5 seconds to rebind when I switch it on then off at the Tx.

I'm guessing he meant 10-15 seconds, and if he happens to be using the AR6100 Rx
with his module then 10-15 seconds is very possible.

[edit]See latest post here http://www.rcgroups.com/forums/showt...=8#post9395025
New firmware, same results.

ian

[erichevy]

No,
I really meant 10 to 15 minutes.
I did this test when I bought my spektrum equipment new, a tm8 module for my 9c radio and 5 AR9000 recievers.
You realy can't tell how long every time it's going to be when switched off/on quick because it differs, between a minute and sometimes as long as 15 minutes.
Meanwhile the led in the module is blinking.
Btw, this has nothing to do with quick connect, which is tested by switching the reciever power off/on and not the radio.
It is no problem to me anymore, because I know now this "side" effect will happen, and will avoid a quick off/on switching because of that.
The quickly swichting off/on is never needed, and must be avoided (NEVER DO THIS WITH FUTABA fasst) but I know now what it will do with my equipment, and because of this, I thought I mention/warn you not to do this while flying, hoping to find a new channel in flight!

But now I am curious what you're Spektrum equipment will do when you switch you're radio off/on quick?? (with reciever turned on )
I do not think this is only happening to me!

Eric.

[erichevy]

I did the radio off/on test again..
Hmmm. now it "only takes 5 to 10 seconds everytime again.

What had happened was that one AR7000 reciever at the time was acting up.
And the binding was doing weird things at that time.
Maybe this was the result of it taking a very long time to re-bind after switching off and on of the radio?
As if the bad reciever was influencing the sending module??, it was acting up (the long binding time) on all 5 of my AR9000 recievers!
When I send that AR7000 in for garantee and asked for a AR9000 instead this long binding effect has not been there anymore.
I just did find this out by testing it again now.
My "input" was to warn you not to do the switching while flying because of this.
But I'am still very curious wether it can happen with someone else?

Eric.