Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Dec 09, 2011, 01:55 PM
Registered User
Joined Apr 2004
46 Posts
Discussion
Exactly How Does DSM Switch Channels

I am looking for an expert answer or an link to an advanced discussion that might answer the following. I can only see one obvious answer but I can't find it in writing anywhere.

Its my understanding that
- DSM is a one way protocol
- a link requires 2 good channels of the 80 available

I am guessing when transmitter turns on, it listens for open (no traffic) channels and starts transmitting on the first 2 free channels *. The receiver will listen on all channels looking for traffic with its matching guid. If it finds 2 of them, the servo ports become active.

* http://www.rcmodelreviews.com/dsm2flaw.shtml showed DSM2 uses the first 2 free channels even if they are very very close to each other and it is possible a 2.4ghz camera (perhaps of questionable quality or legality) could sit right on top of both of those channels and swamp those two channels.

Does that imply the transmitter is always transmitting on all 80 channels (obviously not using all the bandwidth, ie using packets delivered at some regular frequency or slot) ?
Does that mean all transmitters are always transmitting packets and there are random collisions all the time. A wireless version of eithernet. Or is there some sort of syncronising effort of the last transmitter turned on to transmit specifically in the "empty slots" of the transmitters already on.

Also, is it impossible for a 2.4ghz camera say with a 2 channels wide bandwidth to shoot down a DSM plane. The only scenario I can think of is there are 40 planes in the air (all 80 channels in use) and the camera is turned on AND a specific plane is using BOTH of the camera channels (as opposed to just one of the camera channels) ?
DEdwards is offline Find More Posts by DEdwards
Last edited by DEdwards; Dec 10, 2011 at 10:20 AM. Reason: correction
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 09, 2011, 02:15 PM
Play loud, Fly high
Gil Gauger's Avatar
Vincennes,IN USA
Joined Apr 2007
649 Posts
2.4 technologies

Check out this site..

http://rcmodelreviews.com/
Gil Gauger is offline Find More Posts by Gil Gauger
Reply With Quote
Old Dec 09, 2011, 02:27 PM
Registered User
Joined Feb 2008
2,661 Posts
Quote:
Originally Posted by DEdwards View Post
* there is a link elsewhere that showed DSMX uses the first 2 free channels even if they are adjacent to each other and it is possible a 2.4ghz camera of questionable quality or legality could sit right on top of both of those channels and swap those two channels.
I'm pretty sure the possible two adjacent channels issue was with respect to DSM2, not DSMX.
Dave H. is offline Find More Posts by Dave H.
Reply With Quote
Old Dec 09, 2011, 02:33 PM
AndyKunz's Avatar
Illinois
Joined Sep 2001
22,477 Posts
And they weren't adjacent - they were 8 MHz apart.

Andy
AndyKunz is online now Find More Posts by AndyKunz
Site Sponsor
Reply With Quote
Old Dec 09, 2011, 03:06 PM
59 years of RC flying
Daedalus66's Avatar
Canada, ON, Ottawa
Joined Feb 2006
16,032 Posts
You need to be clear about differences between DSM2 and DSMX and not use the generic DSM. You've got them switched in your post.
Daedalus66 is online now Find More Posts by Daedalus66
Reply With Quote
Old Dec 09, 2011, 03:06 PM
Registered User
Joined Apr 2004
46 Posts
Quote:
Originally Posted by Dave H. View Post
I'm pretty sure the possible two adjacent channels issue was with respect to DSM2, not DSMX.
oops, corrected the mistake!
DEdwards is offline Find More Posts by DEdwards
Reply With Quote
Old Dec 09, 2011, 03:16 PM
Registered User
Joined Apr 2004
46 Posts
Quote:
Originally Posted by AndyKunz View Post
And they weren't adjacent - they were 8 MHz apart.

Andy
Thanks for the correction. I reread the article.
http://www.rcmodelreviews.com/dsm2flaw.shtml
DEdwards is offline Find More Posts by DEdwards
Reply With Quote
Old Dec 09, 2011, 03:37 PM
Registered User
Joined Apr 2004
46 Posts
Quote:
Originally Posted by Gil Gauger View Post
Check out this site..

http://rcmodelreviews.com/
Thanks, I reread the "how things work" section. There is no talk of DSM2 receivers switching frequencies on the fly. It sure looks like on power up, the transmitter finds 2 quiet frequencies, the receiver finds them and that is it.

I want to make it absolutely clear that my only agenda here is (after 2 years of trying), be absolutely clear how it works and what damage a low end (perhaps illegal perhaps out of bandwidth) 2.4ghz camera could do on our field.

It sure looks like DSM is a packet based system. Anyone know how many (micro)seconds a packet is and how often it is transmitted per second for DSM2 and FSSH ?
DEdwards is offline Find More Posts by DEdwards
Last edited by DEdwards; Dec 09, 2011 at 05:31 PM. Reason: correction
Reply With Quote
Old Dec 09, 2011, 04:12 PM
Registered User
Daemon's Avatar
Lakewood, Colorado
Joined Aug 2002
28,251 Posts
As far as I know, DSM2 sends each packet twice per control frame, once on each
channel. Tx and Rx are synchronized in time, so Rx has an opportunity to receive
the same information twice, once on each channel. Nothing more complicated than that.

If the band is really quiet when a DSM2 Tx switches on, it may choose two channels
fairly close to each other. A single analog 2.4Ghz video transmitter switched on
afterward may manage to step on both channels, but is generally only going to
cause issues if it's *much* closer to the RC Rx than to the RC Tx.

If the analog video Tx is on and relatively close before DSM2 Tx switches on, then
it'll choose non-overlapping channels and there's no issue.

All of this is moot though.
The practical reality is, someone using 2.4Ghz video for FPV is much more likely to have
their video signal visibly thrashed by various 2.4Ghz RC transmissions, than
the other way around. I stopped using 2.4Ghz video years ago for that reason.
If I was flying far away, and someone switched on their 2.4ghz RC Tx anywhere within
50 feet of my ground station, the video became unacceptably noisy, almost unusable at times.
Noise from DSM2 gear was actually the worst because if there is overlap, the noise is nearly
continuous.. Hopping 2.4ghz systems manage to step on the video signal a lot, but
not continuously like DSM2.

ian
Daemon is online now Find More Posts by Daemon
RCG Plus Member
Old Dec 09, 2011, 04:13 PM
59 years of RC flying
Daedalus66's Avatar
Canada, ON, Ottawa
Joined Feb 2006
16,032 Posts
Quote:
Originally Posted by DEdwards View Post
Thanks, read the "how things work" section. There is no talk of DSM2 receivers switching frequencies on the fly. It sure looks like on power up, the transmitter finds 2 quiet frequencies, the receiver finds them and that is it.
That's my understanding. Most of the time it works very well indeed, but it isn't totally bulletproof, especially in the presence of noise generators that could just pop up after the frequency selection process is done. The argument was that sometimes the Tx picks two frequencies close enough together that both can get zapped. Or than under rare circumstances two sources of interference pop up pretty close to the channels you're on. Countering this is the argument that each channel is a broad signal that's quite resistant to interference.

I've had very good results for nearly five years on DSM2, but I generally have only moderate noise locations. I've had only one incident that I would attribute to interference and that was in a university field house, probably a strong WiFi zone, where the ultra-micro models sometimes had a hard time linking. But once linked they never lost signal.
Daedalus66 is online now Find More Posts by Daedalus66
Reply With Quote
Old Dec 09, 2011, 04:21 PM
Registered User
Daemon's Avatar
Lakewood, Colorado
Joined Aug 2002
28,251 Posts
Quote:
Originally Posted by Daedalus66 View Post
I've had very good results for nearly five years on DSM2, but I generally have only moderate noise locations. I've had only one incident that I would attribute to interference and that was in a university field house, probably a strong WiFi zone, where the ultra-micro models sometimes had a hard time linking. But once linked they never lost signal.
Yes, the link problem is that during initialization it scans for open channels
and finds none, as Wifi traffic covering all of them. Might actually help to
enclose the RC Tx antenna with your hand, to attenuate the external Wifi noise so
it has a better chance of picking channels.
The interesting thing here is Wifi protocols pretty much always plays nice
with the spectrum (they always listen before they transmit), so if DSM2 chooses
an overlapping channel or two, the Wifi routers will choose to not transmit at all
while Spectrum packets are being transmitted. Luckily DSM2 doesn't occupy
the channels 100% of the time, or it would effectively shut all Wifi traffic down.

ian
Daemon is online now Find More Posts by Daemon
RCG Plus Member
Old Dec 09, 2011, 05:08 PM
Registered User
United States, FL, Miami
Joined Aug 2009
955 Posts
I thought the biggest issue with DSM2 was when enough multiple frequency shifting/hoping transmitters were on. It would effectively jam the DSM2 transmitter.

Which is probably the main reason for DSMX.
pach is offline Find More Posts by pach
Reply With Quote
Old Dec 09, 2011, 05:30 PM
Registered User
Joined Apr 2004
46 Posts
Quote:
Originally Posted by Daemon View Post
All of this is moot though.
The practical reality is, someone using 2.4Ghz video for FPV is much more likely to have their video signal visibly thrashed by various 2.4Ghz RC transmissions, than the other way around. I stopped using 2.4Ghz video years ago for that reason....
ian
Thanks Ian. Yes that has been other peoples experience at our field. I'm trying to get consensus if there is justification that our club has a specific recommendation if not rule to not use 2.4ghz camera's. Clearly users won't be happy with the results but could they accidently shoot down an unsuspecting $3000 dsm2 plane too and under what scenario and is the probability more than microscopic.
DEdwards is offline Find More Posts by DEdwards
Reply With Quote
Old Dec 09, 2011, 06:40 PM
Registered User
grumpy42's Avatar
Canada
Joined Aug 2004
848 Posts
Quote:
Originally Posted by DEdwards View Post
could they accidently shoot down an unsuspecting $3000 dsm2 plane too
Most definitely. This is why it is ill advised to use dsm2 in a high value model. There is just no reason to use it with the multitude of FHSS alternatives.

For high value models - just say no to DSM2!

Quote:
Originally Posted by DEdwards View Post
and under what scenario and is the probability more than microscopic.
I believe that anytime you mix 2.4Ghz VTXs and DSM2, the probability of sadness is not insignificant. If the tx chooses two adjacent channels, then just one VTX could shoot down the aircraft. Even if the tx chooses two non-adjacent channels, two VTXs on (or near) those channels could shoot down the aircraft. The probability of these scenarios may be small, but hardly insignificant. It is just not worth the risk.
grumpy42 is offline Find More Posts by grumpy42
Reply With Quote
Old Dec 10, 2011, 04:01 AM
Registered User
UK
Joined Aug 2000
1,082 Posts
DEdwards, I can answer some of your questions. DSM2 hops between two frequencies. The DSM2 frequencies are chosen at random so can unfortunatly be adjascent. The Tx checks how busy they are before commiting to them. Once it starts transmitting it never stops or changes them or the timing. The same data is sent on both so only one needs to get through. Transmissions take about 1.4ms. There are two in every 22ms, one on each frequency.

DSMX is broadly the same but each Tx uses a predetermined set of 23 frequencies spread across the 80 channel band.

The DSM2/X protocol converts random stick positions into pre-determined encoded patterns which are what are actually transmitted. So the receiver only has to recognise patterns. It's like two people talking in a noisy bar. If you understand the language and context of the conversation, you can often makes out the words surprising well. Clearly if there is too much noise/channel too busy, then the signals will not be recognised.

I think this technique can be successful with other RC radios that all transmit in short bursts and then stay silent for longer periods. Video/audio tends to be a continuous stream and usually covers many channels that out radios use. So DSM2 can work mixed in with video but is a high risk which I would recommend against.

I have an 'FPV' section on this page which allows you to see the width of videao/audio compared to normal RC http://www.flyelectric.org.uk/24scanner.htm
Regards, David.
David T is offline Find More Posts by David T
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question How exactly does a UBEC power output work? Multicopter Power Systems 3 Aug 27, 2011 06:58 PM
Discussion How does BID exactly work? pupeno Batteries and Chargers 9 Jul 27, 2010 03:17 PM
how does a boost bottle work exactly ronmeister Nitro and Glow Power Cars 6 Nov 12, 2005 12:48 AM
How exactly does the ESC regulate voltage ? dtrip Beginner Training Area (Aircraft-Electric) 4 Feb 04, 2005 09:09 AM
E-Sky Honey Bee CP 3-in-1. How does it work exactly? Lynxman Micro Helis 5 Oct 12, 2004 06:06 AM