Any radmon experts here? My unit has failed after water ingress

My radmon plus has been outdoors for about 2 years and recently water has entered the enclosure and got on the electronics.The esp8266 appears to be transmitting data to radmon.org so that all looks OK, but the high voltage supply doesn't seem to be working. The OLED display appears to be displaying the correct HT voltage, but when I measure it using the resistor string with my multimeter it reads about 7 volts, touching the probe on the terminals while making measurements will also make it give a false reading of a high number of counts.That means I'm probably getting around 70 volts? Looking at the schematic, the unit is supposed to be sending 3.5khz pulses from the arduino GPIO pin 5?Nothing appears to be showing on my oscilloscope that I can tell when probing that pin.Any ideas on what else to measure?
When i touch the negative pin of the GM tube terminals, it causes the unit to register a reading.
Ive gone over the solder joints and given the whole board an ultrasonic clean to get rid of corrosion, etc. And lastly,  is there any better quality enclosure with a window? The one I bought with the clear window went all crazed and yellow on the lid and started to break down from the sun's UV.Have added photo of board running. Screen has suffered a bit of damage but still readable.

I was wondering that, should I be able to see a pulse signal on my oscilloscope if I probed GPIO 5 I think?
Not sure if the transistors on the board that drive it are damaged or not, but either way cant get anything to show on my scope.

I have had a radmon+ for ...years, and it has been running in a very humid environment (rh 100% most of the time) - as a result the printed circuitry has peeled off in some places and discretely corroded in others, causing numerous difficult to locate failures. The most recent failure displayed all the same symptoms as yours currently. No HV could be detected anywhere on the board, I changed over the ESP, then Q1 and D2, checked the pot. No joy. The only fault I could find was the printed copper circuit from R6 through to the anode screw terminal was corroded in a few places, so, using my 10M ohm anode resistor I joined the end of R6 to the base of the anode screw terminal - It makes little sense, I know, but it hasn't missed a beat since. You may do well by following each circuit and checking continuity between individual components - lead to lead.

344 Jussie. Sounds like you have been giving that counter a hard time! Good job on keeping it running.... under water! 

nzoomed, I have the GK Radmon Basic. It seems the differences, board aside, are in the Arduino code and the plus is browser configurable. I was going to suggest uploading my version of code to try and diagnose as it keeps the HV on all the time, but not sure if it would work on the plus or not. It's been a while since playing with the GK Radmon but I remember that the stock code will go into a deep sleep for 5 minutes, wake up, powers the HV and starts to count for 15 or 30 seconds, then submits to radmon.org, then goes to sleep again for 5 minutes and turns the HV off. Just a sanity check; you have been checking the PWM whilst the counter is awake?

This is from the GK website:

A count and sleep cycle keeps power usage to a minimum. During the count cycle HV is supplied to the tube and events are counted. At the end of the cycle the kit connects to WiFi and sends its readings.

Following the transmission, the RadMon will enter sleep mode. The power demand drops sharply. At the end of the sleep cycle the RadMon wakes up and repeats the cycles.

I wrote new code for my GK Radmon as I wanted it to act more like the NetIO GC10 where it is constantly sampling over a minute and calculating CPM based on 60 seconds of counts. My HV stays on all the time and the counter never goes to sleep as the power supply was always available - I don't run it on batteries/solar. Maybe you can use your browser to stop it going into sleep mode?

I also checked the PWM on mine with a scope and the trace is below. I connected my scope to the female pin header next to the ESP pin headers, so is coming right from the ESP and nowhere else on the board. If I was testing the ESP I would remove it from the counter and power up on an external 3.3v supply and write a very basic code to simply test the output on GPIO5. Maybe test for high and low outputs, then test for PWM. If it has failed on GPIO5 it may be possible to use another GPIO instead of GPIO5, but that would require some modification to the circuit.

Edit: Maybe not possible to use another pin for PWM. It looks like all of the pins are used on the ESP after looking at the schematic.

 

 My board has similar damage to yours juzzie but the traces appear OK, just the solder mask lifting in places.
here is a photo of my scope output on GPIO5 while counting, looks like this is OK?
What should i be testing on the board if im getting no HT?
I see some sort of signal on one of the pins of Q1, just needing to work my way between GPIO5 and the HT.