[hpsdr] Frequency Calibration in PowerSDR?

[Graham / KE9H]

Pete:

You should not have to use the PowerSDR frequency calibration with 
Excalibur.

There is a danger using the Phase window to set Excalibur, if you don't 
already know
that you are real close to frequency to start with, in that you can get 
a false stable image
on the phase display with a large error, that is really an aliasing or 
stroboscopic image.

In other words, when you get an error that is close to your image 
refresh rate,
things that are moving fast can appear to be still. 

In my experience, the Phase display watching WWV is useful if you are within
about 3 Hz of perfect to start with.  Beyond that error, you are as 
likely to lock
to a false stable image as you are the real one.

It sounds like you need to compare your Excalibur with a known accurate
frequency source.

Do you have access to a friend with a secondary standard? GPSDO?
recently calibrated counter?  One of those eBay rubidium oscillators?

--- Graham

==






n3evl wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> I noticed the other day (when setting up PowerSDR 1.19.3 for the first 
> time) that the frequency calibration was a little off (about a couple 
> of hundred hertz: enough to make SSB signals sound off-frequency, for 
> example when tuning to the nearest 1KHz or 500Hz).
>
> I did some comparisons against my SDR-1000 using on-air signals, 
> including WWV and also using my Elecraft XG1 at a nominal 7040 KHz.  
> In all cases the SDR-1000 seemed to be spot-on whereas Mercury seemed 
> to be off as described above.
>
> I am using Excalibur and have performed the slight adjustment (using 
> WWV or the XG1) to the on-board Osc using the Phase display technique 
> to get minimum rotation of the pattern so I think Excalibur is ok.  I 
> don't have any external input to Excalibur.
>
> I  was able to correct the apparent error by, for example, tuning to 
> WWV at 10 or 15MHz and running the PowerSDR Freq. Calibration.  I did 
> this by tuning the signal to coincide with the red line in the center 
> of the panadapter display and running the calibration.  The peak of 
> the as-tuned signal now appears to line up correctly with the nominal 
> frequency red line.
>
> I guess my question is: is this PowerSDR correction normally necessary 
> with HPSDR Mercury/Excalibur as described?  If not, is there something 
> amiss with my HPSDR hardware to cause this frequency error?
>
> Pete, N3EVL
>

 1271987258.0