Hello,
I need to build an amplifier for a photomultiplier tube. I have scoured the net for suitable circuits but am struggling since there seem to be so many solutions.

I want to make something that is low noise and I will be sampling the output at speeds up to 5 MegaSamples per second.

I am likely to require a gain of about 300.
Should I amplify in two stages. Perhaps two inverting amplifiers.

What sort of OP amp should I use? I have seen these which others have used for similar purposes
LT1012. http://www.linear.com/product/LT1012

Would an instrumentation amplifier be better.

In speaking to people, the frequency should be specified. Since I am sampling at 5 MS/s at the very most, I think that the most I would need to filter at is about 2 MHz. What would be the best way to add a filter to the circuit (active/passive) and how do I calculate the resistor/capacitor configuration?

The PMTs I want to use produce somewere between 2 uA and 100 uA current output maximum depending on the type. An example is H10770PA-40.
These are relatives of the following so are likely to have similar specifications
http://www.hamamatsu.com/resources/pdf/etd/m-h7422e.pdf

Can anyone point me in the right direction? I am thinking of two stages of inverting amplifier, say gain of x3 then x100 but I don't know how to design the filters, where to put them in the circuit, or what value range of resistors to use to give the right gain with lowest noise. I know how to calculate the gain for an inverting amp but not whether I should use low ohm values or high ohm values for best results.

I have a basic knowledge of electronics and can wield a soldering iron but not much experience in designing circuits, particularly ones for fairly high spec low noise applications.

Hope you can help,
Thanks,
Best wishes,
Nick H

 

Hello,

The gain of a PMT is dependend on the supplied voltage on the PMT.
Read the attached PDF for more info.

Bertus

 

Hello,

The gain of a PMT is dependend on the supplied voltage on the PMT.
Read the attached PDF for more info.

Bertus

True but the current output still has to be amplified. I plan to control the gain of the pmt with a digital to analog converter. That part is sorted.
Nick

 

True but the current output still has to be amplified. I plan to control the gain of the pmt with a digital to analog converter. That part is sorted.
Nick

Might be worth searching Hammamatsu for appnotes.

 

Hello,

Did you have a look at the PDF?
Have a look at the pages 180 and on.

Bertus

 

I work with PMTs daily. What are you trying to measure?

You can use a FET input amplifier such as AD8067

or you can use a preamp from Cremat:

http://www.cremat.com/

 

Thank you Bertus. I have read the handbook from Hamamatsu. It was one of the many resources I found online. Very informative but the electronics circuits are mainly theoretical with principles rather than details.

To Mr chips, the PMTs are on a microscope. I wish to sample the outputs as voltages at speeds up to 5 MHz. Actually, I am sampling one point every few microseconds. Ideally, I would like to integrate the output signal over a gated period of 1-2 microseconds every 10 microseconds. It seems boxcar averaging is a way of doing this but this seems complex and perhaps something to think about after I have these amplifiers working.

I am upgrading from standard PMTs to InGAsP versions. The old ones have output currents of up to 100 uA and the newer ones up to 2 uA. I need to change the existing electronics that were made for me some time ago by an electronics tech who has since retired and I thought I should take the opportunity to update those circuits which never seemed to work very well. I thought better better quality components and op amps would help. There are no filters either and if I am only sampling one point at a time, then noise is a problem.

One other thing.... I have not bought the InGasP ones yet and one company that sells these with their own microscope suggests using photon counting versions of the PMT. I don't wish to photon count and the company that sells them also do not photon count so I wonder why people would prefer photon counting pmts over standard versions when the signal is simply being converted into a voltage and sampled at a high rate as a voltage. Would there be any difference in the electronics?

This is getting a bit complicated.... I hope you have the patience to stay with me and help.
Thanks,
Nick

 

You still have not said what you are trying to measure. I can only assume that you are trying to measure optical density.

Why do you need to sample at 5Msps? Why is the light intensity varying? Are you scanning on an X-Y microscope stage?

Yes, I was going to ask you if you knew about the difference between photon counting and current measurement.

Photon counting is used for measuring very low light levels. At higher intensities individual photons add up to give an integrated current. The electronics to measure the two are very different. One uses digital circuitry while the other uses analog.

In order to reduce thermal noise you have to cool the PMT. I have done this with a water cooled TEC (thermo-electric cooler).

 

I have built a multiphoton microscope that uses acousto optical devices to scan a pulsed, infra-red laser over a specimen. The signals coming back are from fluorescence and represent light intensities changing over time. In principle, it is possible to collect at very high speeds although in practice individual measurements are made every 10-15 useconds to give the acousto-optic deflectors time to settle before acquiring the data. Other methods might collect faster and so I want to design something that could be used for a range of purposes.

The light levels are not hugely bright but in this case, I will not be using cooling because I need the PMTs really close to the sample. Apparently, as long as you don't mind a large dark current (I can offset this anyway), then these InGASP PMTs seem to work quite sensitively even without cooling. Photon counting is not that great because over a short period of time, even collecting at 40 MHz which is possible with Hamamatsu electronics, only a few counts are measured. This means that the apparent bit level of a resultant image that is formed by plotting light intensity against position of the laser is less than 40 counts per pixel. Slowing things down is not an option.

Best

Nick

 

I would start with a current to voltage preamplifier using the AD8067 followed by one or two stages of voltage amplification (and low-pass filter if required).

 

Thank you Bertus. I have read the handbook from Hamamatsu. It was one of the many resources I found online. Very informative but the electronics circuits are mainly theoretical with principles rather than details.

Try this:-
http://en.wikipedia.org/wiki/Jim_Williams_(analog_designer)

Can't remember whether he ever did PMT apps, but he did publish volumes and volumes of analogue appnotes.

Maybe search the usual suspects like NS, AD, LT etc and see whether they suggest any of their OP-AMPs for PMT amplifiers, firms like those are usually good for a few example circuits.

 

Try this:-
http://en.wikipedia.org/wiki/Jim_Williams_(analog_designer)

Can't remember whether he ever did PMT apps, but he did publish volumes and volumes of analogue appnotes.

Maybe search the usual suspects like NS, AD, LT etc and see whether they suggest any of their OP-AMPs for PMT amplifiers, firms like those are usually good for a few example circuits.

Sorry for the delay in responding to this. I followed your suggestion found the following application note:

On the basis that simple is good, I adapted the circuit in figure 90. It is basically an inverted OP amp circuit. I used a slightly different chip (OPA 2107) as I had something to hand. I will replace it with an LT1355 which I see someone else has used for PMTs and it has better high speed characteristics. If I feel adventurous, I might try the next circuit in the note which allows gated integration of the signal.

Thanks,
Nick