Understanding my own design - moisture detection circuit

[Tonkabot]

I have this circuit I desiged that tells how much resistance there is between the connector (both contacts are connected to one sense wire) and ground.
Then, if the test point (a brass elbow in a co2 gas stream) gets wet (blood, most likely), we know the machine has had a biological contamination.

So normally the ADC pin (A1_Fluid_Sense) reads about 32000 When J5 sensor is hooked up and open.   it reads 38000 or so when J5 is open.
it will read 0 if you dead short it to ground.

My problem is that D34 is getting blown.    it is just a sod-323 shottky.   I had thought that some transorb or something would be more appropriate there, but I never did figure out something that worked.   The intention of D34 was to be like a 5v transorb, but I couldn't get that to work.  We need to keep the ADC pin safe from any crap that might blow the processor.
D34 has not blown much (I think about 3, over a year of 200 machines running), but that is too much.

Oh, Vusb is 5.2volts.


Not sure I am getting the schematic png attached...

[ConKbot]

Is ESD building up from a dry gas flowing? The diode should be sufficient for charge on the wire itself, but is a nearby sections of plastic tube charging up and then discharging into the sense wire?

[NiHaoMike]

Try a larger diode and add a 1k resistor in series with pin 1?

[Tonkabot]

I didn't think there could be a static build-up because there is only 75k to the 3.3v rail and 100k to the 5.1v supply.   I'm thinking that that should eliminate a static potential build up.

Brynn

[Tonkabot]

The client is pretty resistant to making board changes...   However it would be pretty simple to incorperate a series resistor in the sense wire harness, and I don't see any drawback to making such a change permanently.   I'll be trying that out, and it seems like it will help protect the diode on the board.

Brynn