[hpsdr] Status of THOR - Gibralter?

[michael taylor]

On 6/12/07, Chris Albertson <chrisalbertson90278 at yahoo.com> wrote:
>
> The problem I've seen with the low cost consumer GPS units is that
> the output is serial NMEA.  The internal clock might be very good but
...
> There are some GPSes that provide a TTL level pulse per second
> This is what yo want.  But I don't think these sell for under $100.
> (But I hope some one will correct me here.)

These are not SiRF chipsets, but they are $50-$100 and provide a PPS
(pulse per second) to align the NMEA sentence. I forget which one, but
at least one claims within +- 50ns of UTC.

<http://www.sparkfun.com/commerce/product_info.php?products_id=7951>
<http://www.sparkfun.com/commerce/product_info.php?products_id=163>
<http://www.sparkfun.com/commerce/product_info.php?products_id=8291>
<http://www.sparkfun.com/commerce/product_info.php?products_id=8266>

I believe the Trimble Resolution T (within 15 ns (one sigma)) is
between $75-100 in quantities of 1's.

> But do we care?  What really matters is how long one must integrate
> the GPS time to get the desired accuracy goal of what? 10E-11 seconds?
> The GPS signal is only used to control long term drift in the ovenized
> crystal.

Actually some look at it the other way around. The ovenized crystal
oscillator is there to reduce the "phase noise" of the GPS "clock"
output.

> of a small device and hold it very stable.  The solution used here would
> be a double oven with resistive heating.  But they had another design.
> The part was clamped to a small block of silver, a Peltier device (aka
> thermo-electric cooler) was clamped to the other side of the block and

I've never priced any, but I had the impression that Peltier devices
were fairly expensive when you're not talking surplus parts. I don't
think I've ever seen an oscillator built with a Peltier device.

-Michael, VE3TIX

 1181711145.0