Help with simulcast design...

[flecom]

would it be possible to modify two transmitters with something like a reflock (http://gref.cfn.ist.utl.pt/cupido/reflock.html) gps refrence and use them in a two site simulcast (same freq) system?

obvioulsy i would need things like audio delay and a way of getting audio to both transmitters but all that asside, would it work?

[Bruce1807]

That would work fine but as you said audio delay and so on needs adressing.
Motorola also make GPS unit called the Oncore which is cheap and easy.
Off course if you want to do it the officail Motorola way you need a TRAK9100 or similar but any timing unit should work.

[flecom]

well i dont actually have any motorola site equipment i just dont really know that much about mamma /\/\'s site equipment and havent had much luck with it anyway... i just figured this board would be the best bet for finding smart people that would have a clue what i am talking about

right now i have a vhf repeater in downtown, its a pair of GE MDX's and i want to get more coverage in the south so i wanted to add another transmitter in the south end of the county for better handheld coverage... so i was thinking of just using another pair of MDX's... they use a 12.8MHz TCXO so if i remove the TCXO and get a reflock set to 12.8MHz with a gps reference it should be good to go right?

as for audio delay boards, i think the CAT guys make a nice little audio delay board, which would work nicely since i use their controllers anyway...

[RFdude]

I've been thinking about helping out the local hams to piece together a simulcast repeater. We have some Glenayre VHF and UHF high stab exciters and also audio delay lines made by Allen Avionics http://www.allenavionics.com/GPDL/RP.htm The RP1105 was used for digital paging and can delay up to 1105 microseconds. It is a passive device but has up to 10% distortion and I measured 8dB of ripple (not linear across the pass band). Does anyone have experience these these types of delay lines? Would this delay line work with voice audio or is it strictly for digital paging?

Thanks for the insight!

RF Dude

[chpalmer]

I've been thinking about helping out the local hams to piece together a simulcast repeater. We have some Glenayre VHF and UHF high stab exciters and also audio delay lines made by Allen Avionics http://www.allenavionics.com/GPDL/RP.htm

These are the same units used on an amateur system Im a tech for...

We have three micors in the field using hi-stab oscillators from 800mHz rigs. The crystals have been cut from the same stock.

The frequencies must be set but after a few years they seem to stabilize.

This system works pretty much flawlessly. Audio and frequency between the southern and central site is pretty much dead on.

The northern site is off a bit on the frequency due to an equipment change but a trip to the hill this summer should change that.. audio is right on though...

http://www.k7pp.com look for the articles by k7pp at the top and then look at simulcasting...

Ive considered using the same setup except with MTR's on some commercial stuff were doing...

[/\/\y 2 cents]

How are you transporting the audio between the satellite receiver and the comparitor?

Thanks,

Steve

[psapengineer]

I would take a web surf to:

http://www.simulcastsolutions.com

and check out their product line for Spectracom GPS Ageless Oscillators and "manual central delay" Convex delay lines. Spectracom's stuff can be optioned to to 12.8 and Convex can do the delay.


Bob

[flecom]

I would take a web surf to:

http://www.simulcastsolutions.com

and check out their product line for Spectracom GPS Ageless Oscillators and "manual central delay" Convex delay lines. Spectracom's stuff can be optioned to to 12.8 and Convex can do the delay.


Bob

that stuff looks incredibly expensive

this is for my ham repeater so im looking at things like the reflock and the CAT audio delay boards (things that wont require me selling my body )

[psapengineer]

Yes indeed. Simulcast and Expensive are, for the more part, synonyms.

[/\/\y 2 cents]

I still don't know how you are transporting the audio....I can help make you a simulcast on a budget, but what is going to kill you in the long run is the leased circuits / microwave backhaul. Let me know.

Steve

[xmo]

For an amateur project - you are free to experiment and learn. That's what it being a hobby is all about.

Just the same, you would be wise to study industry practice to understand the important aspects of the technology so that you can develop a system design that has a reasonable prospect of working well and not waste your time.

Although you could use IP transport for your voted receiver audio, using IP for the transmit audio presents great technical challenges. Time alignment of the outbound transmit audio needs to be controlled within a few microseconds. Variable latency in an IP connection could render that impossible without expensive time stamping at the source [voter or prime site] and synchronization the destinations [transmitter sites].

Probably not a problem anyway. Logically, as a ham with access to large chunks of spectrum [e.g. 440 and 902 bands] for linking, you would use RF to transport.

Your concept of the use of the CAT audio delay board is simply not practical. You need to precisely control the audio phase at each transmit site. This requires the ability to match the audio delay in very fine increments [e.g. microsecond level adjustment]. On the CAT delay board, a dip-switch selects delays of 62.5, 125, 250 or 500 milliseconds.

Milliseconds won't get the job done. Also, the audio distortion and frequency response characteristics of every component in the distribution portion of the simulcast transmit audio system are critical. You will probably want to use wide band flat audio link transceivers.

Your best bet for the rest of the necessary equipment would be the surplus commercial market. Many first generation public safety simulcast systems are being updated to P25. Lots of good equipment is showing up cheap including Rubidium and GPS system clocks and simulcast audio hardware such as Motorola's PON [Prime Optimization Node]

[flecom]

For an amateur project - you are free to experiment and learn. That's what it being a hobby is all about.

Just the same, you would be wise to study industry practice to understand the important aspects of the technology so that you can develop a system design that has a reasonable prospect of working well and not waste your time.

Although you could use IP transport for your voted receiver audio, using IP for the transmit audio presents great technical challenges. Time alignment of the outbound transmit audio needs to be controlled within a few microseconds. Variable latency in an IP connection could render that impossible without expensive time stamping at the source [voter or prime site] and synchronization the destinations [transmitter sites].

Probably not a problem anyway. Logically, as a ham with access to large chunks of spectrum [e.g. 440 and 902 bands] for linking, you would use RF to transport.

Your concept of the use of the CAT audio delay board is simply not practical. You need to precisely control the audio phase at each transmit site. This requires the ability to match the audio delay in very fine increments [e.g. microsecond level adjustment]. On the CAT delay board, a dip-switch selects delays of 62.5, 125, 250 or 500 milliseconds.

Milliseconds won't get the job done. Also, the audio distortion and frequency response characteristics of every component in the distribution portion of the simulcast transmit audio system are critical. You will probably want to use wide band flat audio link transceivers.

Your best bet for the rest of the necessary equipment would be the surplus commercial market. Many first generation public safety simulcast systems are being updated to P25. Lots of good equipment is showing up cheap including Rubidium and GPS system clocks and simulcast audio hardware such as Motorola's PON [Prime Optimization Node]

no no, i agree that commercial systems that use proven technology are definately the way to go as far as designing my system...

i have some people helping me out that have setup simulcast EDACS and motorola stuff so i have some knowledge i can tap into at least

i am planning on linking my audio back via 420~430mhz links

as for the audio delay, what am i looking at?

if i need something with more precise delays could i use something like this?

http://www.behringer.com/DSP2024P/index.cfm?lang=ENG

that will give me 2 channels of delay that i can specify in 1ms increments...

for refrence could i use a $20 gps module with its 1PPS output going to the reflock? that would still come out a lot cheaper than a Rubidium or similar reference...

[alex]

Your not carefully reading XMO's post.

You need to do some extensive reading on the theory of simulcast operations before you jump in and say "will a $20 part work instead of a $4000 part?"

The key with simulcast is timing and getting the audio correctly - one of the key words in the XMO's post is MICROSECONDS which is NOT milliseconds (ms).

It's a cool project to undertake, don't get me wrong. I don't think there are many ham systems running simulcast around, that's pretty cool. But make sure you throughly do your research.

Carefully reading other posts here and listening and asking questions is a great start. There's a lot of people on here who run simulcast systems and you can pick their brains.

The key to a good simulcast system is timing. You need to ensure that your GPS or whatever your using for time synchronization is on the money at the microsecond level before you can jump.

-Alex

[flecom]

ok well now sorry i missed MICRO vs MILLI

i have been reading several articles online about different systems from M/A-Com and Motorola along with various articles on just theory (not equipment specific)

well i was reading the linked article by K7PP

http://www.k7pp.com/art006.html

about their ham multicast system and they are just using high stability oscillators from 800mhz radios...

so hence why i was thinking that a gps with a reflock should work just fine... and from what i have read it should, in theory, be just fine as long as it is within a tolerance of a few hz

was reading that here

http://www.spectracomcorp.com/portals/0 ... ulcast.pdf

as for the audio delay, that seems to be a bit more complicated due to delays and phasing... most of the audio delays i have found are at least 10 microsecond steps which i think is too coarse for delaying the audio in the simulcast system?

i am aware that this stuff is expensive and very complicated but my hope was that as amateurs and expirimenters we could find some way of adapting existing equipment that isnt as expensive as the multi-thousand dollar commercial simulcast stuff...

maybe i am being unrealistic, but i think it can be done...

some other stuff i found in my travels

http://simulcastsolutions.com/PDF/glossary.PDF

[alex]

Don't take my post as being discouraging, it was meant more that when it comes to the science of doing it, you need to read very carefully, since simulcast is all about timing.

You will need high stability oscillators in order to make the system work, that I know, since timing is tight. There are a few websites out there you should be able to find 3rd party ones, or maybe someone around here has a surplus of 'em.

If your going to do it, I'd probably say finding equipment that comes out of the exact same "batch" and a couple spares probably isn't a horrible idea. That way, you have the exact same setup at each site, and is probably easily controlled/repaired. Sometimes as revisions change, slight specs may be just enough to offset the timing in the system.

-Alex
(FYI, I'm not claiming to be a simulcast expert by any stretch of the imagination, but I understand some of the theory).

[mr.syntrx]

For reference oscillators, there are GPS backed TCXO references oscillators with 1 Hz, 1 MHz and 10 MHz outputs in kit form for about $250 USD total parts cost.

There's one described in the first part of a multipart article in this month's issue of Silicon Chip, an Australian electronics magazine, described as being accurate to within according to the magazine, within 2x10^-8%. I don't think this particular one is available as a kit yet, but the magazine does have a parts list and schematic.

[flecom]

Don't take my post as being discouraging, it was meant more that when it comes to the science of doing it, you need to read very carefully, since simulcast is all about timing.

You will need high stability oscillators in order to make the system work, that I know, since timing is tight. There are a few websites out there you should be able to find 3rd party ones, or maybe someone around here has a surplus of 'em.

If your going to do it, I'd probably say finding equipment that comes out of the exact same "batch" and a couple spares probably isn't a horrible idea. That way, you have the exact same setup at each site, and is probably easily controlled/repaired. Sometimes as revisions change, slight specs may be just enough to offset the timing in the system.

-Alex
(FYI, I'm not claiming to be a simulcast expert by any stretch of the imagination, but I understand some of the theory).

I know its just a daunting task really and i know its going to be incredibly difficult to get working well...

as far as batches of stuff, i am planning on using the same kinds of radios for all the links and such so at least my audio shouldent be different from one reciever to the other

im planning on doing all this is phases, the first part is done (getting the actuall repeater on the air) the next part will be voting the repeater, then adding the second transmitter... if i want to get real crazy i might add a third, depending on the difficulty of adding the inital second one...

i am looking at mostly DIY stuff etc.... but I am keeping my eye out for decommissioned stuff and ebay of course... but i think it would be neat to do as much of it myself as possible

i also have some pretty decent test equipment (sig gen/function gen/oscilliscope/22gig spectrum analyzer etc) and access to service monitors and network analyzers etc

For reference oscillators, there are GPS backed TCXO references oscillators with 1 Hz, 1 MHz and 10 MHz outputs in kit form for about $250 USD total parts cost.

There's one described in the first part of a multipart article in this month's issue of Silicon Chip, an Australian electronics magazine, described as being accurate to within according to the magazine, within 2x10^-8%. I don't think this particular one is available as a kit yet, but the magazine does have a parts list and schematic.

ya i have made a few of the projects from Silicon Chip... i still get their e-mails...

now let me ask you, wouldent the ref-lock be the same thing? except the ref-lock lets me pick whatever frequency i want instead of 1hz/1mhz/10mhz

or is there something im missing about the reflock? its a gps diciplined oscillator no?

please bare with me guys, im sure im going to ask a lot of stupid questions but believe me i am trying to find information out online and in print but sometimes it makes more sence when someone explains it to you you know?

[chpalmer]

http://www.canalgeomatics.com/product_d ... duct_id=60

Another amateur system here on uhf using these.

He uses them on msf 5000's. I have no details on how he has them installed but the timing is within 5 to 10 Hz.

[mr.syntrx]

now let me ask you, wouldent the ref-lock be the same thing? except the ref-lock lets me pick whatever frequency i want instead of 1hz/1mhz/10mhz

To an extent, but changing the frequency involves farting about with the CPLD configuration files, and in any case, building the device requires the purchase of a $150 programming device ($300 if you want the USB model), assuming you find somewhere to get those CPLDs in small quantities. (I haven't checked the availability of that device.)