In the July 1998 QST magazine Brooks Shera, W5OJM, showed a GPS-based frequency standard. (http://www.rt66.com/~shera/index_fs.htm) With a small amount of parts one would be able to get a very stable frequency standard, and, as an added benefit, a very stable time reference, using the NTP code from Dave Mills. (http://www.ntp.org/)
I have decide to build this reference and you will on this page see a diary of how the project develops. Hopefully others will have some use of this.
A very nice building instruction can be found at http://home.teleport.com/~oldaker/10mhz_construction.htm
Another construction, using some parts of a HP-Z3801A, can be found on G3RUH's home page, http://www.jrmiller.demon.co.uk/projects/freqstd/frqstd.htm
An interesting distribution amplifier with an input of 10 MHz sine, and outputs of 1, 5, and 10 MHz sine, can be found at George Hrysanthopoulos, N2FGX, site at http://www.darksmile.net/ee/
I have now tested one of the 8-channel Oncore VP receivers in my HP-Z3801A #2, and it works splendid. The only drawback was that I had ordered receivers with straight, and not right angle, antenna connectors, so the receiver did not fit very well (the antenna connector clashed with an IC circuit, thankfully without short circuiting anything.) I have not yet seen any changes in the behavior of the HP-Z3801A, i.e. a speedier startup, etc. due to the new receiver.
The package from Synergy have arrived (was sent the 14th, so it was a quick delivery.) Will be fun to power it up and see the results from it.
I have sketched an interface for it, consisting of a voltage regulator, a MAX233, and two 74AC04. Nothing fancy, just to get the RS232, and 1PPS out, plus a LED for the 1PPS to see the action. This board will interface with an Oncore UT+ or the M12+ with an adapter board. At a later stage I will redo this for a M12+ only, without the adapter board. The problem with the M12+ is the small header contact with only 1.27 mm between the pins. These header contacts are very tiny and seem to come only in surface mount versions.
My second HP-Z3801A still have problems, PU of 20-30 us, wandering EFC, and jumps in TI. Probably some crystal problem. This unit also sings during the warm up of the outer oven. A hunch is that this is due to dried out electrolytes. I will build an ESR tester and take a look at the five electrolytes on the power supply board to see if they are OK or not, stay tune for further development.
I have now ordered a M12+ timing receiver, with an adapter board (to get the same footprint as for the old Oncore GPS receivers), a cable from MMCX to BNC, an EMI shield, and two old 8-channel Oncore VP (to be used as spares for my HP-Z3801A's) All ordered from Synergy Systems, LLC, http://www.synergy-gps.com/
My second HP-Z3801A has arrived! Have connected it to my laptop running GPSCon, but will later connect it to my Linux machine, and using the Linux tool for logging. The unit seem to behave good, it has a bit more wandering than my first HP-Z3801A. Hopefully it will improve as time goes by.
I have now added a Motorola Timing-2000 antenna to my HP-Z3801A GPS reference. The antenna seem to work better than the hockey puck like Magellan antenna I used before. I now also have no problem with interference from my 70 cm packet radio station, so the filters seem to do their work. Pictures etc. can be found at my HP-Z3801A page.
At the moment I am using a version of a Perl script written by John Ackermann, N8UR, for monitoring of the HP-Z3801A receiver. The original work can be found at http://www.febo.com/time-freq/gps/z3801a/tools.html. I will add more information about my version later on.
Well, I ordered a HP-Z3801A and it has now arrived. I also bought the GPSCon program (look at K8CU:s homepage mentioned above.)
The GPSCon program will produce a nice web page that gives you an idea on how the GPS receiver behaves. If you want to see the data on my receiver, take a look at my GPSCon-page. This page is manually updated, but could be automatically updated from the GPSCon program.
I will try to write a Linux based version of GPSCon, not with exactly the same layout of the diagrams and web page, but something with the same functionality.
On the Brooks Shera design I now more or less know what has to be bought to get a running system. So the next step is to do some more web surfing finding the cheapest deals on components.
I have now missed two HP-10811 and two HP-10544 crystal references on ebay (they went for $183.52, and $61.00 respectively for the HP-10811's, and $91.00, and $72.01 respectively for the HP-10544's) but won two Isotemp OCXO134-12 for $24.00 each. Perhaps not as good as the HP's, but they will do.
I am also thinking of buying the HP-Z3801A that goes for $250-$300 from different suppliers. These GPS-based references are quite new, late 1990, and comes with the HP-10811 oscillator in a double oven. I will probably buy one of these, but also build the Brooks Shera reference. Might be interesting to compare them.
Information about the HP-Z3801A, and also some similar products can be found at K8CU:s home page: http://www.realhamradio.com/GPS_Frequency_Standard.htm
I have ordered, and received, the circuit board from A&A Engineering (http://a-aengineering.com/), and the PIC and DAC from Brooks Shera. Next on the agenda is to do a bill of materials, and a little hunt for the cheapest source of this material.
I intend to use the Motorola Oncore M12+ timing GPS receiver (http://www.synergy-gps.com/M12_Oncore.html) This will create some problems with the interfacing, as this is a +3 Volt device, and all published interfaces have been for the +5 Volt devices. I will draw a schematic of my interface and put up here later on.
I also have to design a power supply of my own. This power supply will deliver +12V at 2A for all circuits, -12V for the A&A Engineering circuit board (used by the DAC), and +24V at 1A for the crystal oscillator oven. This is more than the circuits actually will use, but as this is an 24/7-device, it seems appropriate to design a heavier power supply than needed.
I have not yet been able to find a HP-10811 (or HP-10544) crystal oscillator, so this part is still needed. What I do have is a VCXO (voltage-controlled crystal oscillator) with unknown data, it is built into a reference manufactured by a Swedish company called Radiosystem Sweden AB. The unit is called RS920. Any information about this reference is greatly appreciated.