I'm having trouble with Dave's schematic.
That was intended to be a starting point
for a design. It assumed that there was a 200 Volt source and that one of the Nixie digits would be fully illuminated at 2 milliamperes with 135 Volts applied to the Nixie anode and ground supplied to a cathode. That may or may not be true for any given type of Nixie. Do you have a data sheet?
Even if you have a data sheet, you might find that it is necessary to do some subjective testing to see if the appearance meets your expectations. (And to see whether the approximations that I assumed for the design are valid for the components that you are actually using.)
Yes, the starting point is now with R2=330K and R3=3K3
To see if the supply is doing what I think it's supposed to:
1. Disconnect everything from the Arduino. Disconnect the Nixie tube's anode from the power supply.
2. Connect a 68K Ohm resistor to the point where you would normally connect the Nixie anode (Bottom of R4 on the schematic). Connect one of the Nixie cathodes to ground. Measure the voltage across the 68K Ohm resistor. Should be about zero.
3. Connect +5 Volts where you would normally connect the Arduino anode output signal (Left-hand end of R1)
Measure the voltage across the 68K Ohm resistor. Should be somewhere around 125-130 Volts. (At this voltage the resistor will be dissipating close to 1/4 Watt, so it will feel pretty warm. Two 33K 1/4 Watt resistors in series would probably make a better load than a single 68K 1/4 Watt resistor. The exact value isn't terribly important for this test.)
4. Now, test the Nixie: Disconnect the anode driver circuit from the Nixie.
Apply 200 Volts to one end of a 39K 1/2 Watt resistor. The other end of the goes to the anode of the Nixie. Connect one of the Nixie cathodes to ground. What does it look like? Vary the voltage, if possible, to get the brightness that you like. (If the 200 Volt supply is not variable, you can play around with the resistor value. I would expect it to be in the neighborhood of 39K)
Bottom line: Current through Nixie is (High voltage supply value - voltage across Nixie) / Resistor value.
5. Given the required Nixie voltage and current, we (you, actually) should be able to fine-tune the driver resistor values R3 and R4 to get good results.
6. I have been assuming that in the actual circuit the cathode voltage is zero when a given digit is enabled. The final step is to activate whatever cathode driver you are using instead of directly connecting a Nixie cathode to ground. The cathode should go really close to ground potential.
Anyhow, when you have verified power supply component values, then you might be ready to attach to an Arduino.
It's kind of hard to troubleshoot by remote control and, since I don't have the same setup that you are working with, I might not be able to get to the bottom of everything, but the idea is to thoroughly test the driver stuff before you connect to an Arduino.
The MPSA42 and MPSA92 are 300 Volt devices and will not (should not) break down with a 200 Volt supply. However...before connecting anything to an Arduino I would test the power supply and control circuitry separately and make sure there are no sneak paths (due to design errors or mistakes in connections or anything else) that would apply improper voltages (or allow excessive currents) to Arduino pins.
With power supplies, sometimes "smoke test" is, unfortunately, more than a figure of speech.