From what I read in the manual, that filter sounded a bit extreme for
this signal. With the external filter I can see what the resulting
signal looks like on a scope. With the internal filter it would require
some analysis and faith as to what I was measuring.
If the hold-off worked the way i thought it did, it would have been a
clean predictable solution. But it seems it doesn't do what I thought it
would.
Thanks for all the comments and suggestions, though. My question was as
much about understanding the counter features as it was about making
this particular measurement. I expected that I could clean the signal to
remove the glitch if I needed to.
On 6/7/2011 5:45 AM, Heinzmann, Stefan (ALC NetworX GmbH) wrote:
The PM6681 has an analog filter that you can engage with a button on the front
panel. Have you tried using that?
Stefan
-----Ursprüngliche Nachricht-----
Von: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] Im Auftrag
von Rex
Gesendet: Montag, 6. Juni 2011 00:42
An: Discussion of precise time and frequency measurement
Betreff: Re: [time-nuts] Fluke PM6681 triggering
I forgot that I had a 50 ohm load on the scope end of my cable for that
picture. The audio output is cap coupled. That's why the short pulse
with long exponential tail. If I use 1M input at the scope it is a
square pulse>200 uS but still has that glitch on the leading edge.
I put a simple RC filter on the output; the rise time gets stretched but
the negative glitch becomes a step. Good enough for the counting.
On 6/5/2011 3:53 AM, Bruce Griffiths wrote:
Maybe, but the circuit diagram indicates that the pulse that you see
is nothing like what it should be.
There may well be a circuit fault.
The circuit already includes a monostable.
Bruce
Rex wrote:
We are getting pretty far afield of my original counter triggering
question.
As far as I know, any specific quenching is only necessary for
achieving the highest counting rates which isn't involved in my
measurements, so far as I know.
Like I said, this is a 1960's CD counter. Pretty impressive that it
still works. They did a pretty nice job with minimal components, I
think. If you really want to see the trivial details, here is the
circuit (updated by someone to make it more logical from it's
original drafted version).
http://www.cs.utah.edu/~hatch/images/lionel.gif
I am counting out of the suboptimal Audio output.
I did change the original "hotdog" tube to a pancake style detector
for these measurements.
On 6/5/2011 2:12 AM, Bruce Griffiths wrote:
Is the GM tube internally or externally quenched?
Its necessary to quench each avalanche discharge either by using an
internal quenching gas (eg a halogen) or to use suitable circuitry
to ensure the discharge terminates.
Bruce
Rex wrote:
Bill and Bruce,
Clearly, fixing the messed up signal is the proper approach. What
you are missing is that I got a shiney new (for me) expensive
hammer and I thought that it should be able to drive defective
nails. :)
I got an off-list reply that suggested that hold-off affects the
counter gating -- which either doesn't matter in this totaling app
or complicates it. He also suggested using the negative slope of
the pulse to trigger. Doh! The negative slope is more gradual and
would affect timing accuracy, but that doesn't matter in my
counting situation.
Oh, and as reply to the question of more detail on where the signal
comes from, this is a 1960's CD-700 (civil defense, yellow) gieger
counter. The signal is the earphone output. In the future I think
I'm going to make my own circuits to connect to a geiger tube or a
scintillator/PMT MCA application, but that is even further from
playing with the nice new counter.
Thanks for the feedback -- any more welcomed.
-Rex
On 6/5/2011 12:42 AM, Bruce Griffiths wrote:
A Geiger Muller (GM) tube produces an output pulse as a result of
an avalanche discharge in the gas filled tube initiated by the
passage of ionising radiation through the tube.
A high voltage is initially maintained between an outer usually
cylindrical electrode and an inner small diameter wire electrode.
The discharge current develops a voltage across a resistor in
series with the inner electrode. The pulse amplitude is relatively
large and little gain is required to drive a speaker.
Pulse shaping using a suitable differentiating and integrating RC
time constants is typically used to shape the pulses and maximise
the SNR of signals from scintillators and proportional counters.
For Geiger counters the signal is so large that such shaping to
maximise SNR isnt usually required.
Using a non retriggerable monostable to define the deadtime in
nuclear counters is relatively common.
The pulse risetime for a GM tube is relatively slow so that
something like a 74HC series monostable should suffice.
An HCMOS monostable also has the advantage of a high input
impedance so that little or no amplification should be necessary,
Bruce
WB6BNQ wrote:
Hi again Rex,
I should have asked these questions in the first place.
How are you connecting the Fluke to the geiger counter ?
Is this a signal that drives a speaker or some other kind of
noise maker ?
What happens if you load that line with some capacitance like 1
uf or more ?
If the capacitance helps you will have to experiment with the
value so as to not
completely destroy the pulse shape. Never played with a geiger
counter so have
no real idea how they do the noise making.
Bill....WB6BNQ
Rex wrote:
I recently picked up a Fluke PM6681 counter (same as a Pendulum
CNT-81).
Looks like a sweet device.
I was just trying to use it for a not-so-much-timing purpose and
was
hoping to find an expert here who might help me with a
triggering question.
I just set it up to count total pulses, over a 5 min interval,
coming
randomly out of a geiger counter. Basically I set it up and it
works
except for a subtlety. The pulses out of the geiger counter are not
clean. At a low count rate they have a big glitch on the leading
edge.
Here is a picture of the pulse:
http://www.xertech.net/geiger/single.jpg
The glitch causes the count to increment by two on each event
except
that when the pulse rate gets high the pulse shape changes
causing the
the glitch to smooth out and the peak amplitude to drop, like this:
http://www.xertech.net/geiger/multiple.jpg
If I set the trigger voltage on the counter to just above the
glitch
peak I can get proper counts, but finding a sweet spot on the
changing
wave shape is not ideal.
I thought I could use the counter's Hold Off feature to get a clean
solution but it isn't working as I expected. Reading the Operator's
Manual I thought that the Hold Off period started at a trigger
event and
would prevent another trigger event until after the hold-off
period. I
thought I could set the trigger level to occur around the middle
of the
glitch rise (about 3 volts) and set the hold-off time for 1 uS
or more
to prevent a 2nd trigger on the big rise just after the glitch.
I tried
hold-off values of 250 nS through 20 uS, but I still see the count
incrementing by two on the glitchy pulses.
I'm wondering if anyone has experience with this counter and can
tell me
if I have mis-understood the Hold-Off function. Or maybe it has
something to do with me using Total A-B mode. The Op Manual
covers a lot
of ground, but it isn't the easiest to follow the finesse stuff
unless
you happen to need to do exactly what they are showing in an
example.
_______________________________________________
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
