The MACT button is momentary. As I recall, when I pressed the MACT button before I had the oscillator locked to GPS, I got output. It was not exactly 10MHz, a bit off, and remained off even after GPS lock. OCXO was warmed up but frequency stayed the same. A SYST:PON reset the unit, and since everything was warm it synced up quickly and then the 10MHz was on frequency. It's almost like it locked the DAC voltage (but I didn't check).
The 4 pin serial port is RS232 with RS232 like levels. There is a level translator on the board. So you must use a real serial device - a TTL level adapter will not work on that port. I have not looked at the ports on the high density connector.
A couple tips for improving the operation of these units (at least the Symmetricom ones). Power up the unit and let it do it's survey and get all nice and stable. (Make sure SYST:STAT? shows Position MODE Hold). Then issue a GPS:POS:HOLD:LAST? command. The output will be your position:
UCCM-P > GPS:POSition:HOLD:LAST?
N,+42,+34,+3.40880E+01,W,+74,+39,+2.72220E+01,+1.10340E+02
Command Complete
UCCM-P >
The output is conveniently in the form used to manually set the position. Use these values in the future to skip the survey. We will also want to set some other parameters to optimize the operation. What you do is send the UCCM a series of initialization strings. I've posted the ones I use below. Send them with a terminal program, a micro-controller, whatever.
SYNC:REF:DISABLE LINK
SYNC:REF:ENABLE GPS
REF:TYPE GPS
OUTP:TP:SEL PP1S
GPS:SAT:TRAC:EMAN 20
GPS:POS N,+42,+34,+3.40880E+01,W,+74,+39,+2.72220E+01,+1.10340E+02
What all this does is turns off the other timing sources (which are unconnected but could still have influence) and using only GPS. It also makes sure the PPS output is 1 PPS and not 1 PP2S (1 per 2 seconds). The satellite elevation mask is change to 20 degrees to avoid multipath and other anomalies from satellites near the horizon. This improves the timing and frequency stability. Finally we set our position, which will skip the auto survey. Note that it doesn't seem to really like slamming the commands in. So wait for the UCCM prompt to come back before sending the next command.
I would bet there's a way to apply battery to the board and avoid all this, but I don't know what it is yet. It's easy enough to do this with a micro or just a script, but still slightly annoying.
Finally, here's a plot of the stability. Don't take it as gospel, it's just one sample.
/?action=dlattach;attach=206950;image)
Home
Help
Search
Login
Register
