[time-nuts] Fluke PM6681 triggering

Rex rexa at sonic.net
Thu Jun 9 21:32:38 UTC 2011


Magnus,

Thanks for the detailed reply and history.

The counter (just got it recently) came with the Fluke CD which had an 
op manual on it. I also found the CNT-81 manual on a Pendulum site. They 
seem to have identical content except for name changes. One note, if 
anyone else has one of these counters, one difference is that the CNT-81 
pdf is searchable where the PM6681 pdf is image based, so isn't searchable.

Reading the manual is where I got the impression that hold off could 
solve my triggering issue. I just looked at the manual again and found 
there is a performance test setup described for verifying the hold off 
function (pge 9-16 in the CNT-81 manual). Using that with a 3314A 
generator I verified hold off working as expected. One difference: the 
3314A couldn't make a two-cycle pulse as fast as the test described, but 
allowing for that with slightly longer hold off period, it worked as 
expected seeing freq measured at 20K or 10K with hold off.

That was just as I expected it to work and seemed I should have been 
able to use hold off in my problematic measurement to mask out a glitch 
from causing a second trigger on the leading edge of the signal. So, 
today I used that 3314A test setup to compare in the mode of the counter 
I was using. I was using the counter to total the number of random 
pulses over a preset interval of several minutes. I set the counter into 
TOT A-B mode with the measuring interval set by Aux configuration of 
21.12 as described on page 4-18. Today I put the counter in this mode 
and fed it the same 3314A 2-cycle bursts with 100 uS interval. The count 
was what I should expect for the measurement interval I used. I turned 
on hold off with the delay time as set in the verification test above. 
The expected result would be the count dropping by 1/2. It did not change.

So my idea the hold off could solve my measurement issue, bypassing the 
glitch, was ok in principle, but for some reason the hold off function 
doesn't seem to work in this Totalize mode.

Just closing out with a verification of why it didn't seem to work as I 
thought it should. Thanks for getting me to think about this again and 
finding a verification of my results with less random test equipment.



On 6/9/2011 11:43 AM, Magnus Danielson wrote:
> On 06/09/2011 07:30 PM, Peter Vince wrote:
>> Hi Rex,
>>
>>       You could try asking the guys at Pendulum.  A couple of them came
>> down a gave a talk at the UK's National Physical Laboratory "Time&
>> Frequency Club" meeting a few years ago and were very interesting and
>> approachable.  It looks like they might have been swallowed by
>> Spectracom now - see:
>>
>> <http://www.spectracomcorp.com/ProductsServices/TestandMeasurement/tabid/1244/Default.aspx> 
>>
>
> They are part of Spectracom now, but the people is still there. The 
> CNT-81 production stopped only due to lack of key components. The 
> CNT-81 has a full custom chip in it for the time counter stuff.
>
> I had one for a while but swapped gear with a fellow time-nut.
>
> The manuals is downloadable from Fluke if you follow the PM-6681 
> search. The Fluke connection goes back to they days when Philip 
> Industries in Järfälla designed and manufactured a lot of the Philips 
> stuff. Then Fluke came into the picture and after number of years 
> Philips killed of that branch of the buissness, but Harald took over 
> the remains and formed Pendulum. Pendulum itself has swallowed XL in 
> the US and a result of that was the development of the CNT-90XL which 
> brings the CNT-90 base into higher frequencies, using the base 
> functionality of the CNT-90 with the key technology of the XL stuff.
>
> The CNT-81 is a nice box, but the CNT-90 is much more versatile. 
> Still, there is room for improvements!
>
> If there is any specifics, let me know and I'll try to find out.
>
> To the best of my knowledge regarding this thread, the most specific 
> issue related to triggering on that specific signal. I've often found 
> good use for a good scope along-side counters, because you need to do 
> the reality check to know what your actual signal is, and then be able 
> to trigger properly on it. For more expensive counters, I also want to 
> check the signal range. That way I can see if I have any runt pulses 
> to avoid or possibly filter out....
>
> Cheers,
> Magnus
>
>





More information about the Time-nuts_lists.febo.com mailing list