System grounding high voltage supply (8kV), flipping polarity

[ChristofferB]

Hey, so I'm constructing a detector for my DIY gas chromatograph using corona discharge (or rather, the current flowing through it) to detect gasseous chemicals.

The design is simple enough, see attached, and it does produce a nice corona. It's time however to integrate it with the rest of my gear, and heres the trouble:

I would like to be able to swap the polarity at the electrodes. Some chemicals is easier ionizeable like that, same phenomenon as in mass spectrometry.

With this configuration, and the psu neg grounded as shown, can i swap Con1 and Con2 to get -HV at one terminal?

The voltage is selectable, as the CW-multiplier is tapped, but max output is probably around 7-8kV

The box before the transformer is just a simple transistor switching circuit, the HV board + transformer is really from an electric fly swatter.



Thanks in advance, I can't really get my head around it.


--Chris

[jmelson]

If the transformer can safely isolate 8 KV, then it is easy, you just swap two wires.  If not, then it gets a lot harder.  One way to do it is put all the rectifiers on a plug-in board that can be pulled out and turned around to reverse polarity.  I've seen this in some power supplies for photomultipliers.

Jon

[ChristofferB]

I see, well since the transformer is some unbranded thing from a consumer product, It might be wise to assume it's probably not stellar in any regard.

Flipping the diodes is not a bad idea. A bit cumbersome but effective. It's designwise pretty ugly, but if I ran the HV psu off of a battery, non-earthed, wouldn't the problem be solved as well?

thanks!

[PartialDischarge]

Something seems wrong with your schematic. So you have an isolated +-HV, but somehow you want to measure the current that goes through it with a resistor that is referenced to a ground in the isolated primary side?

[PartialDischarge]

You should place the primary ground at Con1 and the resistor in series with the chamber. Then depending on whether Con1 goes to +HV or -HV, you will get a positive or negative reading on the amplifier and ADC.
The transformer will never get at 8kV also

[ChristofferB]

you're absolutely right, I messed up the schematic. This is the actual circuit, as reported.

thanks!

[NiHaoMike]

Have you considered just turning the corona cell around?

You can also take a look at using a CRT flyback transformer as your high voltage supply.

[ChristofferB]

Flipping the cell might work, but that would have the 'receiver' electrode - the one coupled to the ADC - be the needle instead of the plate.

There is something special about the needle-plate geometry of electrodes in this cell, I don't completely get the physics behind it but i'm trying to recreate a cell resembling one in a published paper, so the less i change the better.

A flyback HV supply, or even a spark coil or..pretty much anything other than what I've made would probably be a better HV supply.
I wanted a HV supply that was relatively safe to use, and the draw on the HV in the cell im replicating is in the order of 0-40 uA so I really don't need much.

thanks!