[App_rpt-users] GPSDO comparison

[Bryan Fields]

On 8/3/15 2:29 PM, Tom Hayward wrote:
> Since asking about transmitter steering, I've been looking into the
> feasibility of simulcast. The marginal difference between transmitter
> steering and simulcast (in terms of equipment) is a GPSDO and a
> transmitter that accepts a 10 MHz reference input. I have some leads
> on transmitters--Quantars or some Daniels MT-4D. Other transmitter
> recommendations are welcome, but what I'm really curious about is the
> GPSDO.
> 
> It looks like the Trimble Thunderbolt is the gold standard.

It is, and the RTCM interfaces with it.  Spend the money if your time is worth
it, or unless you want the enjoyment of re-engineering stuff.

> At less than half the going price of Thunderbolt, you can find these
> BG7TBL GPSDOs:
> http://www.ebay.com/itm/121530825744
> 
> Some are labeled Trimble and Huawei, so I think it's safe to say they
> are neither. The question is: are they any good? It looks like KE5FX
> has done some testing:
> http://www.ke5fx.com/gpscomp.htm
> 
> I don't really know what to look for in the graphs, but he says in his
> conclusion "the BG7TBL unit is actually running at about 9.999 999 999
> 800 Hz". Is this a deal-breaker for simulcast? If all transmitters
> have a BG7TBL GPSDO, does it matter that they aren't exactly 10 MHz?
> Is anyone running these GPSDOs with RTCM simulcast that can comment on
> their performance?

I doubt it as the PLL is going to divide the signal down to 5 or 6.25 KHz, and
then use it as a reference to compare it to the VCO. (in a quantar for
example).  This is an error of 200 uHz, but it looks like it's a consistent
error (it's precision is excellent but accuracy is meh).

This is an error of 0.000020 PPM or .020 PPB.  This is amazingly good.  For
example this is an error of  0.0185 Hz at 927 MHz.  It's only going to be
better lower in frequency.

This would be the difference between a true 10.000000000000000 Clock and this
clock.  Assuming all the clocks are the same amount off, your sites will be
off a few ten thousandths of a Hz from true, but all will be the same amount
off so it's not going to matter in any case.

Keep in mind the Quantar with the UHSO option is like 10 PPB  and still used
for simulcast in some applications at VHF/UHF since it's about 1-2 hz of error.

I've got a 220 hamtronics simulcast setup working, not really deployed it long
term yet, but I'd like to.  I have some Pic code posted here about my PLL
programing.

73's

-- 
Bryan Fields

727-409-1194 - Voice
727-214-2508 - Fax
http://bryanfields.net