Inverting Amplifier is clipping the output signal - Help?

[chemical25]

Hi all,

I am working out a way to invert & amplify a signal output from a photomultiplier tube. The signal output is varying negative pulses from 0v to -1v.

I would like to invert that signal & amplify it x10 for later reading with an ADC. I thought it would be a simple inverting amplifier circuit with an OpAmp, but I cannot seem to get it to perform correctly.

With my circuit, while the signal is inverted and the pulses seem to be amplified, the peaks are "flat-topped" and won't go above ~300mv (see Figure 1). I originally thought it was an issue with the slew rate/bandwidth of the OpAmp I was using, so I just received some ultra-fast LM7171 OpAmps (datasheet linked below), yet I am still seeing similar behavior.

Can anyone tell me what I am doing wrong? I am new to the finer aspects of analog design, as I am usually working on digital circuits - I have seen some notes on "input capacitance" and "input impedance," but I do not know if those are the root of my issue... or how to determine "impedance matching" with the output of my photomultiplier tube.


References:

**broken link removed**

My schematic

Figure 1: The output of the OpAmp on my oscilloscope
-- Top Trace: The signal output from the photomultiplier tube that I am trying to invert & amplify
-- Middle Trace: The expanded trace of the output from my OpAmp circuit - **Note the flat top... all large peaks seem to be clipped in similar fashion at ~300mv
-- Bottom Trace: The output from my OpAmp circuit



Any ideas?
Thanks!

 

[FvM]

Many informations are missing:
- any load resistor at the photomultiplier or open (current source) output?
- original PM signal waveform
- oscilloscope timebase setting
- oscilloscope connection: high impedance (1M/10M?) probe? coax cable? length? oscilloscope input impedance?

 

[keith1200rs]

I would have expected a PMT output to be a current pulse but anyway, what is the PMT pulse width? Also, what is the timebase of the oscilloscope traces?

Keith

 

[chemical25]

- The signal used from the photomultiplier is just a BNC socket, which I have split two wires off of - signal is fed through the 1k resistor to the OpAmp, and ground of the signal from the PMT is tied to the supply ground of the circuit.
- Sample PM signal waveform below (See Figure 2)
- 1M probes are attached to PM signal and OpAmp output


Interesting: So, I just hooked a BNC cable from the signal output of the PMT, rather than using the BNC pigtail I made with a 1M probe (signal ground tied to circuit/OpAmp supply ground)... and the peaks are much larger... 0 to -12v now... so what does that mean? the 1M probe showing a much smaller amplitude... Again, this impedence/fragile signal stuff is new to me... I just need to get the signal amplified so I can read it with an ADC.

Figure 2: One of the pulses from the photomultiplier tube **using a BNC cable directly off of the signal output... no 1M probe, as used before
Details:
Pulse width: 157us
Pulse amplitude: -10.28v

 

[pannaguma]

I am not sure if this applies to your case especially since your circuit already has an opamp, but one time faced this signal distortion problem was when I was trying to read some microAmpere current directly with a probe... adding an opamp buffer solved my problem (its simply a negative feedback amplifier with unity gain AFAIR)

 

[keith1200rs]

What is the model of the PMT? It looks like the output could be a current not a voltage.

Keith

 

[chemical25]

You're right - the output of the PMT appears to be a current output, but I believe I am using the OpAmp as a current to voltage converter...

Can anyone help by putting together a simple circuit layout of what should work for me on how to convert the current output of the PMT to 0-5v positive pulses? I am guessing some sort of buffer circuit/inverting amplifier, using some sort of load resistor? Should there be capacitors involved?

Note that when I plug the signal BNC cable directly into my scope I get 0 to -12v pulses, with it on 1M DC coupling...


For the PMT I am using - I am actually using a gamma scintillator probe which has a single BNC cable for the HV supply & signal... here is the circuit I am using to split the signal out of that line from the HV:

 

[FvM]

The measurements you told sor far seem to be inconsistent, it's hard to guess from a distance what's right or wrong. And obviously, there's a lack of understanding how to operate a PMT.

To get high speed signals from a PMT, it has to be loaded by a low load impedance. In contrast, in post #4, you see the PMT current pulse integrated at the cable + oscilloscope input capacitance. It would be more reasonable to switch the oscilloscope to 50 ohm input impedance to see the true waveform.

I have no explanation, why the OP output voltage seems to saturate at 300 mV. As a first point, I would check the probe attenuation, I wonder if the are really 1 Mohm (and 1:1 attenuation). The oscilloscope should have a calibrator output for test.

It may be the case, that the OP has been damaged by input overload, LM7171 is specified with maximum 10 V differential input voltage. Protection diodes may be needed for it when connecting sources like a PMT.

I also suggest to test the amplifier with a signal generator before connecting the PMT.