How can I invert a negative signal pulse?

I am working on a multi-channel analyzer and have an output coming from a photomultiplier tube. The output is in the form of negative pulses of varying amplitude between 0 and -3v. They are very sharp and have a ~2uS fall time.

I would like to convert that signal to positive pulses so that it can be read with an ADC.

The key is that the conversion needs to be proportional, so that I can determine the pulse heights.

Due to the sharpness of the pulses, after inverting the signal to its positive equivalent, I would also need a way to hold each peak at its highest point until it is read by the ADC, and then have my microcontroller throw some sort of reset to clear the peak being held - not sure how to do that.

So, any help would be greatly appreciated - I am great with digital circuits, but this one analog obstacle is holding up my project...

Thanks!
- Josh

That sounds like a job for an Op-Amp. Check out Figure 30, here: https://www.analog.com/static/imported-files/application_notes/28080533AN106.pdf

There are probably 100's of designs to get the same inverted function. TI and National also have a wealth of information of Op-Amp applications.

John

Using the suggested figure to simulate (see below), I am not seeing the output as the absolute value of my signal... Rather, on the output, I am only seeing a slight positive bump and a very similar signal to my input. Is my simulation setup wrong, or is the circuit not capable of supporting my input signal?


The suggested figure:


My simulation:


The input signal on the simulated oscilloscope (made to be very similar to the signal I am looking to measure):


The output on the simulated oscilloscope:



Thanks -

I suspect the intent was for the JFET to have its Drain and Source connected together (i.e., diode connection) (see:https://www.ti.com/lit/an/sboa058/sboa058.pdf).

In any event, the crux of my suggestion was to use an op-amp. That specific circuit may be more complex than is needed in your situation. As mentioned, there is a large number of compendia with op-amp applications from the major manufacturers.

John

I don't see, that the problem calls for an absolute value amplifier. Pulse inversion will be achieved by a basic inverting amplifier. Pulse stretching respectively peak hold is a different topic. It's a standard function in classical nuclear electronics and you'll find a lot of related literature. Available ADC speeds have considerably increased over the years, I guess that todays fast ADCs will mostly obviate the need for peak hold circuits.