DC TO AC FAST SWITCHING

[josedavidch]

I have an application where I need to surge a gas discharge tube* (GDT) with high voltage DC current to set it in conductive mode (turn it on). As soon as the GDT is on, I need to fast switch to 120 VAC in less than 500µs (avoiding the GDT to go OFF state).

What can be used to switch from DC to AC in such fast way? I have been trying find a device, but no lucky so far.

 

[Skogsgurra]

This sounds like something people do all day - and night. Not knowing how they do it and not knowing if your application is a standard application, my thinking goes like this: Why so fast? The 120 V AC is below 50 V (that's where gas discharge usually stops) for around 0.8 milliseconds twice per period. So the 500 microseconds time seems to be just about right.

Second: Do you really have to switch? If you use an old fashioned inductive ballast, then you can have 120 V applied all the time and then apply a short high-voltage pulse across the tube to ignite the gas. The ballast will keep the short pulse from being lost to the mains (and also protect other loads that do not like kV pulses). As soon as the gas is conducting, the 120 V wil take over.

How to do it with an electronic ballast is beyond me.

Gunnar Englund

http://www.gke.org

 

[cbarn24050]

Two mosfets connected source-source, gate-gate will switch AC or DC between the drains. Speed depends on the driver you use, 500uS should be very easy. The two gates need an isolated drive.

 

[josedavidch]

Thanks skogsgurra

The given proposal is quite simple and just confirm my thinking. It is hard to believe, some times, that solutions can be that easily applied. Sometimes one believes it should be a more hitec solution, but then, one has to remember that solutions should be as simple as possible.

I was thinking in switch cause I wanted to make sure not AC will go thru DC gas starting supply.

My solutions goes like this:
1-To ignite the gas, apply a pulse 2 times the GDT dcbd (direct current break) from a DC-DC high voltage converter whose output is in parallel with an HV capacitor (cap needed to provide GDT the extra boost of energy required to ignite the gas). The converter is a 3 mA xformer where output (0-3000VDC) is proportional to input (0-12VDC). The HV will be connected to GDT using a HV diode in series with a 1M resistor big enough to dissipate power during the GDT ignite.
2-The AC, always present and connected at the same node as the 1M resistor will be ready to conduct current thru the GDT as soon as it ignite the gas.
3-1 donut current sensor will detect current flow thru the GDT, disabling the DC HV supply, just to make sure we disconnect it when not required. The


A secondary solution will be to use a small step up AC transformer from 120VAC to up to 2000 VAC at a very low current. The ballast, as skogsgurra suggest, will keep this xformer isolated from the mains and then, we don't need to switch nothing that all, cause the current always prefer to go the easy way, and it would be the GDT, not the HV AC circuit isolated by the ballast.

skogsgurra would you comment on my text?

Regards

JDC

 

[OperaHouse]

Why not just use HF AC to ignite the tube. Both sources could be connected together all the time. Turn off the HF when low voltage when LV current is detected.

 

[josedavidch]

OperaHouse

What is a HF AC and how would it works?

Is that a High Freq AC device?