I'm working on a series of discrete-based preamps for PMTs, with slew rates of 500-1000V/usec and rise times of around 5 nsec. These will be based on garden variety parts. I am in the simulation stage at this point, with promising results. When I have actual hardware tested, I will post. The 3-transistor Hamamatsu preamp that Richard Hull has posted is actually not too shabby, and uses readily available discrete parts. Rise and fall times are 100nsec,limited by the compensation capacitor, and it it has a true current input and transimpedance characteristic, which makes it a natural for a PMT preamp. It has a few other good and bad points, which I will discuss when I post on the circuits I'm presently designing.
I am a bit concerned about the speeds of the usual PMT preamps, which appear mostly to be optimized for use with NaI scintillators, which have a fluorescent lifetime of a few hundred nanoseconds. Plastic scintillators are considerably faster, on the order of ~5nsec. This may be mitigated somewhat by the fact that a recoil proton in the scintillator plastic will take a finite time to be stopped, leaving a trail of ionization (and light output events). I need to stop soon and calculate the path length for an average recoil proton in scintillator plastic (and collision events/second) to see what sort of a light streak is generated. I suspect that what I'm going to end up with is some sort of fast, moderate gain current input preamp followed by a charge sensitve amplifier. I have seen the LM359 around, but have not actually used it for anything of the sort you describe. I suspect you'll want to start with something faster, with a short settling time. Richard Hull has suggested one of the video amp ICs. One other you might want to try for starters (if you can find it in your area) is the Analog Devices AD846. This is a wide bandwidth current feedback opamp with fast slew rate and settling times. Linear Technology makes some similar devices. I think both AD and LT sell small quantities directly these days. LT even has on-line shopping.
A parting shot - I think that discrete amplifier are still used for nuclear preamps beacuse one can individually optimize each gain stage and pay attention to such things like well mannered overload characteristics, symmetric shew rate, ruggedness, and such. ICs offer a lot of convenience, but at the cost of compromises inherent in the nature of a device that must be integrated on a single piece of silicon. I'm doing the design I mentioned above for these reasons (also to see how far I can push a discrete design).
Created on Thursday, June 14, 2001 1:42 PM EDT by Richard L. Hester