Integrating a pulsed photodiode signal

[DrDunk]

Hi All,

I am trying to build a circuit that will do the following;

1) Take as input a gaussian shaped waveform, which is a the output of a photodiode that has been hit by a pulsed laser
2) Convert this input to a square wave output, the amplitude of which is proportional to the area under the incoming peak (i.e. integration)

I have come up with the following circuit, however I am not sure if it will work. I was thinking of using an OpAmp in integration mode, and use a monostable (7422) and switch (4066) for the reset/integration. The monostable will allow me to change the integration window.

Any help would be greatly appreciated! Do you think it will work?

Cheers,

 

[ericgibbs]

Hi All,

I am trying to build a circuit that will do the following;

1) Take as input a gaussian shaped waveform, which is a the output of a photodiode that has been hit by a pulsed laser
2) Convert this input to a square wave output, the amplitude of which is proportional to the area under the incoming peak (i.e. integration)

I have come up with the following circuit, however I am not sure if it will work. I was thinking of using an OpAmp in integration mode, and use a monostable (7422) and switch (4066) for the reset/integration. The monostable will allow me to change the integration window.

Any help would be greatly appreciated! Do you think it will work?

Cheers,

hi,
Looks ok in principle, the problem maybe the switching speeds of the logic and 4066 analog gate.

What range of laser pulse widths are you expecting.??

 

[DrDunk]

Hi Eric,

Thanks for getting back to me so quickly, and for your comment.

The laser width is 100 femtoseconds, however I was going to use a photodiode which has a rise time of 1ns. My thinking is that the energy from the laser pulse will appear as the voltage developed in the photodiode circuit, and the rise time of this peak is not a concern.

I was going to use the sync out from the laser, which comes some time before the actual laser pulse (typically on the order of microseconds), to trigger the monostable. This would then activate the analog gate well in advance of the pulse from the photodiode coming.

What do you think?

Cheers,
DrDunk

 

[crutschow]

Unless your 100 femtosecond laser pulse (with an implied frequency of about 5 THz) has a large amount of energy it is unlikely to be seen by the photodiode. Do you know it's energy?

 

[DrDunk]

Hi Carl,

We already see the laser pulse on the photodiode on an oscilloscope. The energy is on the order of 0.5-1 mJ per pulse. We have had to reduce the energy by using filters so that the photodiode is not saturated.

I think that this might be ok, what do you think?

Cheers,
DrDunk

 

[ericgibbs]

hi Dunk,
It must be a very high spec scope that you are using in order to see a 0.1pSec pulse.

Is the purpose of integrating the detected pulse to determine a range/distance measurement.?

 

[DrDunk]

Hi Eric,

I wish the scope was that good! Actually, the shape of the laser pulse we see on the scope is determined by the response time of the photodiode, which is at best 300ps rise and fall time.

The purpose behind all of this is to construct a spectrometer for transient absorption experiments. In essence, we will excite molecules to a higher energy state, and then record the absorption of light from that state to an even higher state. The reason for these experiments is to determine the rate with which the molecules relax back down to the ground state. It's pretty cool stuff (at least we think so.. If you hadn't picked up on it, we are in academia..).

We will use two of the photodiode setups I have shown above, one will integrate a laser pulse which has passed through the sample of molecules, and is called "I". A second set up will integrate a laser pulse which will pass through nothing, and this is a reference measurement called "Io". The absorption due to the sample of molecules is given by Abs = log (Io/I)..

The reason behind integrating the signal is that it will produce the square wave, and we will then sample this square wave with an AD converter (Data Translation, DT304). We can sample 1000 shots a second comfortably, and then average them. If we were to try and sample the gaussian pulse of the photodiode directly, then there is too much irreproducibility in the signal we would get, if the time profile of the peak is slightly different each time and we sample the same point in time.

Hopefully this makes a little more sense?

Cheers!
DrDunk

 

[ericgibbs]

hi,
Thats pretty clear, sounds interesting.

The laser work I have done is pulsed semiconductor lasers, pulsed on for about 25nSec at a prf of 400/sec, approx 10 Watts at 904nm.

On of the problems we found was the response rate of our PD and amplifiers.

The response time/rate of the PD and amps is a function of the signal strength.

This response time can be in the order of 5nS to 20nS, the 'weaker' the signal the slower the response of the PD and amps.

To overcome this we amplified the signal from the PD in 2 linear amp stages before attempting to sample/hold the signal.

We used a high speed A2D converter to digitise the S/H signal.
During the calibration stage of the equipment we used an iris to incrementally reduce the laser pulse onto the PD, from this we produced a 255 step 'correction' table.

During use/actual measurement, the correction table was subtracted from the actual range figure to give a corrected range.

I guess you may need to consider this as a requirement in your project.

We found that the 74F series of logic gates gave the best results.

Hope you can follow this OK.

 

[MikeMl]

I think you are on the wrong track with the photo-diode amplifier (integrator vs charge amplifier). Read this current thread on some other Forum.

 

[DrDunk]

Bit confused, am I on the wrong track?

 

[ericgibbs]

Bit confused, am I on the wrong track?

hi,
I will let Mike explain his post.

As you already may know there are two basic ways to connect a photo detector to an amplifier.


Attached a pdf which may help, sorry if you have already covered this ground.