[time-nuts] HP 5328A Divider / Timebase Output performance

Magnus Danielson magnus at rubidium.dyndns.org
Wed Mar 17 22:52:58 UTC 2010


Hi Bert,

Reverting back to the list for more people to see.

Bert, VE2ZAZ wrote:
> Hi Magnus,
> 
> No, I don't see why there would be any cycle slips since the readings are around 65 ns and both Start and Stop come from the same reference OCXO. Quickly browsing through the data does not show any.

You are correct, there isn't any cycle-slips. I just wanted to rule that 
out.

> I have attached the .TIM file in which you will find the raw data. If you can, please check for frequency drift as well.

A quick look at the data (using fityk - a nice handy tool) I do see some 
form of slow sliding effect, with quick retraces. These span about 15 ns 
of range! Those effects does not look like coming out of the normal 
noise but rather from some form of systematic property. It also is 
interesting to note that the jitter seems to be lower for shorter 
delays. The "quick" restoration slopes has linear shift while the 
release is a much slower exponential response.

The noise is still substantial.

It would be interesting to see similar plot for a higher rate, such as 1 
kHz.

Cheers,
Magnus



More information about the time-nuts mailing list