[time-nuts] Question for any HP105 owners HP5370 Settings

John Miles jmiles at pop.net
Tue Aug 30 15:00:35 EDT 2011

> 1) Under    Function
>  The counter is in the "TI" mode
> 2) Under statistics  mode
> The counter is in the "Mean" mode
> 3) Under Sample Size
>  Number "1" is selected
> 4) Under arming
> The "+TI" button is selected.
> 5) Under display rate.
> Fully clockwise  the display nor the GPIB interface will update if this is
> in the "hold" position
> 6) Inputs section
> All three pots are in the "preset" position.
>  10 MHz Z3801A Reference is connected to "Start"    DUT is connected to
> "Stop
> Start Stop inputs are "positive slope"  and each is "50 Ohms" Termination
> Both input switches are in the "divide by 1 " position.
>  External BNC has no input.
> External reference drive, in the back = the Z3801A

It's probably better not to use one of your test sources as the counter's
external reference.  That's asking for systematic-bias trouble.  The
internal 10811 reference is quieter than the rest of the counter for
measurements less than ~10 seconds apart, so there is rarely a good reason
to use an external reference at all, unless you are looking to measure
absolute frequency.

> LED displays " 53.20  with the. 20 changing  N secs"
> Remote led lit
> "Talk"  in the GPIB area is lit.

What limits the rate at which readings are coming out of the counter?   You
need to either back off the rate control to yield about 10 readings/second
or fewer (and let TimeLab measure the interval), or use a 1-pps or 10-pps
source on the START or external-trigger jacks.   Otherwise I'm surprised the
GPIB-Ethernet adapter is keeping up... and at best you are likely to get
irregular tau intervals.

If/when you go to a pulsed source on the START jack, don't leave the level
control in its preset position, or you'll probably get some extra noise near
0V.  Use DC triggering and set the level about halfway up the edge (e.g.,
with 5V pulses in from a CMOS divider,  you could set ~2V trigger level in
x1 or 0.25V in x2, whichever gives you the lowest residual noise.) 

-- john, KE5FX

More information about the time-nuts mailing list