We know exactly what conductive coolant causes [shortens LPS life] but we do not know why. More often than not chronic LPS failures have been associated with conductive coolant. Unfortunately we don’t know exactly what is the right range of conductivity… we do know that distilled water treated with algaecide has a good working conductivity.
My over simplistic theory is that the water jacket full of water capacitive-ly shunts to ground some of the current from the LPS. It is not uncommon to stick your hand in a coolant bucket while operating and get a tickle. Since this current is not in the cathodes circuit it does not show up on the meter and the operator does not see that the actual current being drawn from the supply is greater than it should be run at.
So maybe when running the machine at 20 ma maybe 10 ma is being robed (because the coolant is anifreeze) and the LPS is trying to put out 30ma which is above its recommended running current?
This seems fantastic but consider that:
- The effective area of the jacket exposed to the voltage and the water is pretty big
- The voltage on the tube is large 20,000 VDC
- The max current the supply can handle is 30ma, so 10ma is 30% of the supply’s capacity
- You can easy imagine 100’s of volts drop across the various effective capacitance’s. One user measured 300v across his coolant bucket.
So you could easily imagine a large plate capacitor with 20,000vdc across it coupled to earth ground through all of the coolant loops surface area.
I have not been able to test my theory cause I have not gotten a hold of an old tube yet to set up a proper test environment.