Need to count high voltage discharges using a digital counter

As a science project, I am planning to build a device to count alpha particle hits. The device works by ionization. (please see ) There is a high voltage DC power source. Negative side is a metal mesh. Positive side is a copper plate. Between them there is a space of 1/3". When an alpha particle (such as those emitted by Americium from a smoke detector) hits the arrangement at a right angle from the mesh side, it ionizes a path in the air between anode and cathode, bridging the gap and causing a discharge between them.
I need to very accurately count those discharges over periods of several hours by means of a digital counter (such as those you can buy over Ebay). The counter should be able to be activated by each individual high voltage discharge. Therefore I need a method of making those discharges safely activate a switch that would make the digital counter advance one unit for each discharge.
My budget is very limited. I'd appreciate some input from our experienced community. Thanks in advance.

How high is *your* high voltage?
Any idea what it drops to when a discharge is occurring?
What is the highest number of hits per second expected?

Assuming the high voltage is positive with respect to GND, then a very high impedance voltage divider driving a high speed Schmitt-trigger-input CMOS gate, such as something in the AC series, will give you a signal any counter will be happy with. For example, if the HV is 8 kV and the series current limiting resistor is 20 M (as in the video), then a 10K resistor between the cathode (aluminum disc) and GND will divide the HV down to 4 V, pulling the input of a 74AC14 hex inverter high when there is a discharge. In between sparks there is no current, and the 10 K resistor holds the inverter input low. The CMOS gate can handle over 1 million transitions per second, so the more difficult part is to find a low cost counter that can keep up.

Chip, decoupling capacitors, input protection diodes, $1.00.

This circuit will not capture two overlapping arcs. If an arc in one area starts before an arc in another ares is extinguished, the circuit will see one long arc instead of two short ones. The circuit requires a small time gap between arcs. It can be less than 100 nanoseconds, but it has to be there.

Where are you located?

ak

I think it would be also worth putting a 4.7 volt zener diode in parallel with the 10 K resistor. As there will be some capacitance to ground due to the wiring between the bottom of the 20M resistor and the detector the initial spark will have to discharge the stored charge in the capacitance so there will be an initial current pulse that is not limited by the 20M resistor.

Les.

The discharge most likely produces a fair bit of Rf, you can hear it in the recording, that being the case a simple Rf probe circuti might do the trick, you'd have to mess with the capcitance to get the response right.