[time-nuts] Wavecrest vs. HP 5370 vs. SR620 Jitter measurements

SAIDJACK at aol.com SAIDJACK at aol.com
Sat Mar 20 04:21:22 UTC 2010


Hi Demian,
 
I can confirm that the Wavecrest achieves it's ~3ps noise floor (with some  
luck and getting one with a very good internal 100Mhz  reference).
 
For full calibration of the unit, one needs the VISI software and a GPIB  
connection.
 
It does this by having an internal resolution of 800 femtoseconds, and by  
averaging (say 1000 to 3000 readings). It can do this very fast.
 
On an 5370B I get 20ps rms with averaging only on a good day, and with  
lot's of tweaking of the front panel nobs.

BTW: the Wavecrest get's best results with very fast edges (such as  from 
LVDS drivers). Sine Waves are worst.
 
Bye,
Said
 
 
In a message dated 3/19/2010 21:03:36 Pacific Daylight Time,  
demianm_1 at yahoo.com writes:

I have  read through the data sheets on all of these and am very interested
in the  differences in reality. The Wavecrest claims 2 pS Jitter 
measurement.
The  HP has a single shot resolution of 20 pS, The SR620 is 25 pS and  the
Wavecrest specs say 20 pS. How does the Wavecrest get .1X minimum  Jitter
measurement of the others with the same resolution? Can that trick  be
applied to the others?

Demian



More information about the time-nuts mailing list