Pulse-energy monitor
We would like to know how much energy is being delivered in each pulse in order to:
We will be sensing each laser pulse with a fast photodiode for purposes of triggering the timing electronics/APDs and measuring the start time with high signal-to-noise. The photodiode puts out a negative pulse about 1.0 ns in width (when fed a 40-ps laser pulse at 786 nm), and sends this to the 9327 amplifier/discriminator. The discriminator works optimally at 30 mV pulse-height, but has a 5× attenuator on board, so that 150 mV pulses become optimal when this is set.
The photodiode will be fed a splinter of laser light: that which does not reflect off of the dichroic turning mirror at the laser output (see this description). Even though this is only 0.01%—0.1% of the laser light, it will be heavily attenuated so-as not to send the photodiode off-scale. I plan to use a 10× attenuator to get down to 150 mV, so we can expect a 1.5-volt 1 ns negative pulse (AC-coupled) to come out of the bias tee looking for 50-ohm termination.
A fast integrator that can integrate and hold the corresponding analog voltage for some 40 ms would do the trick. This way, we give the computer plenty of time to record the value for each pulse, and still have 10 ms to reset and wait for the next pulse. The integrator should scale so that a 1-volt 1-ns pulse (remember, it's negative) would produce something like a 5-volt output (variable gain to get it right). Actually, it would be smart to tune this to the same number that a power meter placed in the main beam would measure (something like 2.4 W). This way there will be no interpretation necessary, and we'll know right away when the laser needs to be tuned up.