[time-nuts] WG: [tacgps] Strange oscillations

Hubert v. Bonhorst hbonhorst at aol.com
Fri Sep 10 15:19:41 EDT 2004




Von: Hubert v. Bonhorst [mailto:hbonhorst at aol.com] 
Gesendet: Donnerstag, 2. September 2004 21:30
An: 'TAPR GPS and Timekeeping Mailing List'
Betreff: AW: [tacgps] Strange oscillations


Alberto, you are right. This cycling is caused by the oven itself. If you
are not happy with the output accuracy of your  frequency standard you can
try to open the box and play around with the integrating constants of the
internal temperature levelling loop. But before make sure that the crystal
temperature is correct. To be honest, I personally would not change
anything. It is extremely difficult to improve a good OCXO.  Finally I' m
interested how you realised your GPSDO.

73, Hubert DB7ME



Von: tacgps-bounces at lists.tapr.org [mailto:tacgps-bounces at lists.tapr.org] Im
Auftrag von Alberto di Bene
Gesendet: Donnerstag, 2. September 2004 14:54
An: time-nuts at febo.com; tacgps at lists.tapr.org
Betreff: [tacgps] Strange oscillations


I would like to ask your opinion about some strange oscillations.
I built a GPSDO, not a replica of the Shera's project, a new one,
and monitoring the EFC to check its behavior, I noted that, after
it stabilizes, there are some pseudo periodic oscillations, which
I don't think are caused by an instability of the PI algorithm used 
for controlling the EFC. 

I have graphed them below, capturing via a serial port the EFC values,
and converting them to frequency deltas, knowing the sensitivity of
the EFC input of the OCXO used, an Isotemp OCXO-134, using Matlab
for the conversion and the graph.
My suspect is that they are caused by a thermal cycling of the oven,
given the period (about 5 minutes), albeit I am pretty sure this
oven has a proportional control system, not bang-bang.
This suspect is also confirmed by the fact that blowing energically 
on the oven causes a major disruption of that cycling.

I would know your opinion, especially from those who have a greater 
experience than me on this stuff,  thanks.

73  Alberto  I2PHD

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 8588 bytes
Desc: not available
Url : http://www.febo.com/pipermail/time-nuts/attachments/20040910/55173787/attachment-0006.gif 

More information about the time-nuts mailing list