[time-nuts] HP Z3801A Question
SAIDJACK at aol.com
SAIDJACK at aol.com
Thu Nov 13 18:10:33 EST 2008
Glad your HP is still working well.
Well it seems you found out how well the Fury is working with an excellent
The unit running in Mexico achieves similar performance with our standard
double oven OCXO, see for example:
It has not jumped in several days, and the standard deviation is between
1.7ns to 1.8ns when it doesen't jump during the last 24 hours.
Yes, the TI as displayed by GPSCon is much better than the HP unit. I think
this is mostly due to the much better GPS receiver in the Fury, since the
OCXO of the HP unit is extremely good.
Tom and I had discussions about this being caused by the small sawtooth
low-pass filtering we do on the TI intervall (Z3801A does the same thing,
otherwise you would see the full GPS Sawtooth). The theory was that this was what
was causing the Fury to be so much better on GPSCon.
To check this, we modified the firmware to completely remove this sawtooth
filtering, and output the raw data instead, with all the nasty sawtooth
artifacts on it.
This did not change the result much, I think the max range went from +/-10ns
to +/-15ns if I remember correctly.
So even if we display true raw capture data, we are still significantly
better than the Z3801A which displays low-pass sawtooth filtered data.
I think this can mostly be attributed to the extremely good performance of
the M12M GPS compared to the Oncore GPS in the HP unit.
Also, I think you will see that your HP unit will get better over time, as
the Crystal retraces and ages over the next weeks, and the control voltage
changes will hopefully get to be smaller and smaller on your unit.
In a message dated 11/13/2008 14:38:57 Pacific Standard Time,
ernieperes at aol.com writes:
But both unit , the HP and the FURY board running under the same
software simultaneously / GPS Con / and I was mentioning the same small
"windows" TI to GPS, and the FURY displays much better figures then
HP.....but probably each units gives out / display / different
Also the graphic display / BLUE / on the FURY board the max ampl
between +2,2nsec and -2,0nsec sometimes +0,7nsec ,
the HP unit never goes under +15nsec and -20nsec...... almost 10 times
Anyhow I am happy because it seems to me that the HP is still OK after
3 years of sleep.....
More information about the time-nuts