Passive conversion of TTL to NIM pulses

What is a NIM pulse ?

Most of the commercial counting card on the market expect NIM pulses at their input. More precisely, NIM pulses, standing for Nuclear Instrumentation Module, are negative pulses between 0 and -800mV.

However, most of the microcontroller and  some avalanche photodiode output TTL (Transistor-Transistor logic) level pulses, that is to say, typically, +5V positive pulse.

One has to be very cautious since feeding a counting card with positive pulses may destroy the electronic inputs of the counting cards. A minima, the positive pulse will be discarded by the counting card.

Consequently it is compulsory to convert TTL pulses to NIM pulses. Reducing the amplitude of the pulses to 5V to 0.8V is the easy part and can be simply done with a voltage divisor (or even better with a PI in order to keep the 50 Ohms impedance), but inverting the pulse is trickier.

Simple RC derivative passive circuit

Of course, one can use active circuits with an opampYou have to carrefully choose the bandwidth of the op-amp since MHz pulses can have harmonics at very high frequency for instance, but passive solutions do exist since it can be bought from picoquant or B&H for hundreds of euros.

Actually, as it has been confirmed to me by B&H, this circuit are simple RC derivative circuit.

Fig 1 – RC derivator circuit, the Schottky diode is facultative and is here to reduce the amplitude of the positive pulses.
Passive_TTLtoNIM_pulse_inverter

The RC filter, used as a high-pass filter (see fig 1), can act as a derivator when the 1/RC resonant frequency is far lower than the pulses frequencies Indeed, the tranfert function can be then approximated by a mutiplciation by jω, that is to say a derivation. In term of filtration, only the high frequencies of the sqaure pulse are keeped, that is to say the very onset of the pulse.. Derivation come also with a large voltage reduction since most of the frequencies of the square are filtered by the high pass filter.

The filtered square impulsion hence become an alternance of positive and negative spikes. Positive when the square pulse is rising and negative when the square pulse is falling. Positive alternances can be dimished using a low threshold diode, typically a Schottky diode with a 0.2V threshold.

Once again, one should carefully check what is the maximum positive voltage that can safely be directed to the input of the counting card and confirm that positive pulses of around 0.2V won’t destroy the card.

As for the cut off fc frequency of the filter, it has to be low compared to the frequency of the pulse in order to act as a derivator, but not too low because it would decrease too much the amplitude of the pulses. In the commercial passive pulse inverter of B&H, the cut off frequency is close of 15MHz.

2 thoughts on "Passive conversion of TTL to NIM pulses"

  1. Nathan Fuchs says:

    Hello, I know that we can calculate the values with fc= 1/(2πRC).

    However, what values would you recommend for R (and C) if I want to cut at f=15mHz (for a 40mHz signal)?

    Thank you

Leave a Reply

Your email address will not be published. Required fields are marked *