Okay, so you are injecting a constant average current; did you have to adjust that resistor value to match the other components? I’m also curious about whether the 500 V level drops when there are a lot of counts, and what the pulse length is - it looks like the output RC constant is about 10 us? Does that set your dead time, or is the tube a limiting factor?
Anyway, this all looks very fun, and I am definitely going to have to try building my own Geiger counter soon. I would like to try doing it without the 555, using a PWM output and analog feedback to set the output voltage, if possible, and maybe a hardware counter on the AVR for the counting. There are plenty of uses for one of these around the house: checking for radon, measuring the potassium content of bananas, analyzing granite countertops, etc…
-Paul