[time-nuts] Is Timelab with a Prologix-Eth and a PM6680 - working ?

John Miles jmiles at pop.net
Tue Jul 3 22:33:33 UTC 2012


> Ahhh ... Thanx now i know why my 1hr Trace Duration was so fast.
> I assumed that Timelab would set the sampling interval written in Timelab
> via GPIB. But i'll just select 1sec manually.

That can be a bit confusing.  The TimeLab drivers for GPIB counters don't
attempt to program the counter in any respect, beyond perhaps turning off
continuous acquisition so that the program can fetch readings synchronously.
The program may also disable various statistical features that are known to
interfere with reading a stream of simple TI or frequency measurements.  But
for the most part, TimeLab expects you to configure the counter manually,
using whatever settings and parameters you want.  Then you have to fill in
the acquisition dialog to tell the program what you're doing.  

This lazy approach violates the most commonly recommended GPIB programming
practice, where the program begins by resetting the instrument to its
default state and then programs all modes and settings of interest.  But
it's the only way I can afford the time to support lots of different counter
models, most of which I don't own. :)

For counters in talk-only mode, the acquisition driver can't program the
instrument at all, so the above is doubly true.

In the specific case of the Trace Duration field, TimeLab will estimate the
rate at which readings arrive from the counter if you press the "Monitor"
button and wait a bit for the estimate to settle down.  If it doesn't seem
to be converging on the right rate, try toggling the Monitor button off and
on.  Otherwise, if you don't use the Monitor feature, you will have to enter
the trigger rate manually.  
 
> I will try this now , and i assume i should then select Frequency in the
> Data Type - In TimeLab , as i don't use A->B measurements .. Or ?

Most TICs exhibit some dead time between GPIB readings in frequency-count
mode.  For instance, if you use an HP 5370B in FREQ mode, you'll see
something like 1.1 seconds between readings, instead of the 1-second gate
time you may have selected at the counter's front panel.  Not the end of the
world, but something to keep in mind when deciding whether you want to
capture TI or frequency data.

That being said, when I use my 5370B, I usually use frequency mode with a
1-second gate time.  It's quicker and much less error-prone to set up.  The
quantization/jitter floor isn't that much worse than what a TI measurement
can achieve, I don't have to worry about whether the START and STOP sources
are stable and close enough in frequency to avoid undersampling phase wraps,
and the resulting data quality is OK for many purposes.   This will be true
for most other counters as well.  When you are first getting started with
TimeLab, I'd recommend sticking with frequency readings until you're more
familiar with the overall process.  Once frequency readings are being
collected and displayed properly, you can consider switching to TI mode when
appropriate.

Finally, the Prologix adapters should work well with the various GPIB
counter options in TimeLab, but especially with the Ethernet adapter, you
may not want to use trigger rates faster than 1 Hz, at least not at first.
As you get closer to 10 readings per second, the likelihood increases that
the combination of the Ethernet adapter and the TimeLab driver won't keep
up.   The firmware in the Ethernet adapter doesn't handle these overrun
conditions very well, and may lock up hard enough to require a power-cycle
to recover.  When I use a GPIB-Ethernet dongle with my 5370B, I usually keep
the display rate control near 12 o'clock to limit the trigger rate to 4-5
per second.  That's very reliable in my experience, but 10+ readings per
second is not.

-- john, KE5FX
www.miles.io





More information about the time-nuts mailing list