Th output from a photo multiplier tube (PMT) is basically a current signal. It may be converted to voltage by passing it through a low noise 50 ohm resistor. The voltage is typically of negative amplitude because of use of negative (~ -1kV) dc bias to PMT. Once the voltage is available, it may be amplified using a voltage amplifier if it output is too low else voltage can directly be used.

Instead of using 50 ohm resistor and amplifier one can also choose a trans-impedance amplifier available from Hamamatsu, Japan.

This was how to obtain analogue voltage from a PMT. Now, this voltage  can be obtained on PC in several ways as suggested below:

1) First method: Feed the PMT output voltage to a digital storage oscilloscope (e.g TDS 2024 from Tektronix). See the waveform on one of its channel (Ch1, say). Connect DSO to PC via RS232 link/GPIB port . Use a DSO driver on PC to view the DSO screen on PC.

2)Feed the PMT output voltage to a digital storage oscilloscope (DSO) (e.g TDS 2024 from Tektronix). Write your own simple driver using RS232 commands to acquire sampled Ch1 (where PMT output is connected) on PC.

3) Third Method: Use a data acquisition card (DAQ) from NI or Dynamic Signal, USA or from any brand with sufficient sampling rate, band width and vertical resolution (i.e. number of bits. e.g. 8, 12 or 16). Connect the output of PMT to DAQ's analogue channel input. Use its DAQ's driver (Lab VIEW or Matlab based) to acquire the sample in PC.

Once the samples are available , one can easily play with these for further application development.

Hope these suggestion help you.

Manoj Kumar Saxena

Scientist.

Raja Ramanna Center for Advanced Technology. Dept. of Atomic Energy,

Govt. of India

Fiber Sensors Lab. Indore, India.

The following URL is a link to an extensive but easy-to-read documentation on PMTs and their applications.

https://www.hamamatsu.com/resources/pdf/etd/PMT_handbook_v3aE.pdf

Chapter 5.3 (“Connection to an External Circuit”) describes several possibilities depending on the amplifier you want to choose, including the required math.

If you want to observe slowly changing signals (sampling in the order of several hundreds of ms or slower), connecting the output amplifier (low output impedance) to a data logger might be the easiest way to obtain data you can directly transfer to your computer and process in Excel. If you want to observe high-speed signals, a digital oscilloscope is a straightforward approach.

As Manoj explains, your best bet is to use a digital oscilloscope and connect it to a computer. I would just suggest avoiding RS232 / GPIB if at all possible - most modern units will have USB connectivity, and probably come with LabView / Matlab / general DLL drivers. In this case, you can control the scope from the computer, and let LabView handle all the connectivity issues.

There also also data acquisition cards, which are functionally equivalent to an oscilloscope. We have one from Alazar - it's a solid piece of hardware, but software support is not that great. We run it through Matlab.

A good suggestion from Mr Glebov. USB allows you reduce the programming burden for data acquisition and allows you focus on the application and research.

Manoj Kumar Saxena

Scientist.

Raja Ramanna Center for Advanced Technology. Dept. of Atomic Energy,

Govt. of India

Fiber Sensors Lab. Indore, India.

Dear Mr. Timischl, Unfortunately the suggested URL link is not working. Could you please directly send the pdf to me on my personal mail id: [email protected]

It would be really helpful.

Thanks.

Manoj K Saxena

Thanks guys for your informative reply. 

As suggested by Manoj, I have made a simple transimpedence amplifier with Op-amp (OPA227), resistors and capacitor to see the PMTl signals with oscilloscope. Now I can see the laser pulses with right frequency (7.5 KHz) in my oscilloscope (500 MHz). However, currently I'm facing another problem. My  laser pulses are nearly 7 ns in duration. However, my oscilloscope shows it is more than 500 nm (the RC time constant of my amplifier is ~5.5 nm). Do you guys have any idea what could be the possible reason for this pulse broadening? I have gone through the link that Mr. TimischI mentioned, but did not get any clue.

Any suggestion would be really appreciated. 

Thanks

Kibria

Hello Mohmed Golam,

I think it is problem of the slow response of the system that you are using -

1) Your laser pulses are extremely fast - 7ns

2) Oscilloscope has certain time integration

3) And third and most important - the SLEW RATE of the op-amp OP227 that you are using is also noteworthy - 2.3V/us.

So even though you  might have maintained the the R-C time constant of ~5.5ns in you op-amp circuitry, I suspect it is the op-amp slew rate that is giving you slowed output on the CRO.

Try another faster op-amp (with higher slew-rate).

Best luck!

Prasanna is right: a very much faster op-amp should be employed. Consider that the oscilloscope channel bandwidth of 500 MHz corresponds to about 0.7 ns of rise/fall time. On the other hand, the gain-bandwidth product of the OP227 evaluated at 100 kHz is typically of only 8 MHz.

This means that in small-signal conditions you can thus expect a rise-time of about 44 ns, if the OP227 is used as a unity-gain buffer.

If the OP227 is configured as a 20 dB gain amplifier, then the bandwidth goes down to 800 kHz and the rise-time goes up to 440 ns.

In case of output large-signal conditions, the situation becomes worse, due to the slew rate limitations.

Regards