Tek 2215: monitoring PMT pulse

[Dominic-Luc Webb]

I found an add for a Tek 2215 oscilloscope 60 Mhz. Anyone
know if this is suitable to monitor PMT pulses? I have heard
of inputting the output more or less directly to the screen,
not sure how, but maybe someone with experience can tell
me this will work or is hopeless. I do not want to count
pulses, etc, just to see the waveform. Any warnings/suggestions
about this particular model, in general?

Typically pulse from a PMT is 10 ns from baseline back to
baseline.

Thanks to any responder,

Dominic

 

[CBarn24050]

Typically pulse from a PMT is 10 ns from baseline back to
baseline.

You cant use a 60Mhz scope to measure a 10nS wide pulse.

 

[JeffM]

You can't use a 60MHz scope to measure a 10nS wide pulse.
CBarn24050

Yup.
http://groups.google.com/groups?q=0.35+bandwidth+risetime+convert&selm=vancleefDMG8K2.5x6%40netcom.com&rnum=1

 

[Dominic-Luc Webb]

On Mon, 29 Mar 2004, John Popelish wrote:

60 MHz is not fast enough to see single photon pulses, but you can
slow the waveform enough to blur several photon pulses together to get
a measure of the average flux. Supply the PMT with a high negative
voltage through a multi resistor divider (to produce the steps of
voltage for all the dynodes) to ground and ground the anode through a
low value resistor (50 to 1000 ohms). That resistor will drop voltage
in proportion to the PMT current. Parallel that resistor with a
capacitor to average the photon counts. That voltage is what you can
watch with the scope. Or you can amplify it with an opamp first.
Keep and eye on the supply current, because exposing the PMT to
significant light while powered can burn it.

--
John Popelish

Thanks a lot John! It was pretty clear 60 Mhz is too slow for
single pulses, but I specifically remember once seeing a trick
used to accomplish the same general task and I think this is
what you are describing. Expensive Ghz oscilloscopes are out
of the budget, unless something comes along I don't know about.
The trace I have seen was a number of dots on the screen from
multiples pulses. I think in principle if one were to draw a
line through the points (curvilinear regression) one would end
up with the mean waveform, which I think is good enough for my
needs.

Thanks on the warning on the light exposure. This much I knew
already....

Dominic

 

[CBarn24050]

Expensive Ghz oscilloscopes are out
of the budget, unless something comes along

Try ebay, you can get an old tek scope, at least 400MHz, for a few dollars.