Developing HV DC Pulses

Do I still need a negative supply to pull the grid negative?

OK, rejecting the tube in series with the load idea.

Ya, musta been a glitch in my microsoft calculator. ;-)

Part of the reason I'm leaning towards a tube is I don't want to be troubleshooting why I'm popping transistors or mosfets, I think the tube is just more robust in the face of problems, but I could be wrong. No one has addressed the 1300H secondary that I will be unloading. Since this will now be a filtered bridge output maybe the filter caps would limit any spike, but that has spooked me on Solid state. I'm willing to go SS if my spike issue is addressed. On the other hand I'm think of an over build for later schemes, so maybe tube, or 30kV of mosfets. Actually a filtered 30kV neon transformer is more like 42kV, but we could limit it.

I'm gathering supplies to setup 3 vessels running at increasing voltages to see if production is linear with voltage input and to see if there is an increase of voltage that has diminishing returns. Will build a resistive voltage divider to pick voltages from. Someone mentioned a ground plain under the resistors, that seems counter intuitive. Should I mount the HV standoffs on Lexan or a copper PCB, and then do I earth ground the pcb?

27,000Vdc, Max Plate current 1.6ma, OK, not a lot of overkill, but cheap.

I'll wait a bit and see if there is further discussion, and might order later today, along with an octal socket. I'll probably double up, is it possible to series two tubes if we later want higher Voltage of parallel for more current.

Driving the grid positive would turn it on harder. Tubes are depletion/enhancement devices. The rated max plate current need not be respected here.

Series would be very tricky. Do you want more than 27 KV?

One trick in driving Pockels Cells is to drive the two electrodes antiphase. That could be done with two grounded tubes. Not a bad idea even at lower voltages. Maybe.

Not now, we are working max 9kV times 1.414 or 12.7kV with the current neon sign transformer. but there are 30kV neon sign transformers for later. Just thinking about when we upscale and have more space between electrodes, we would need higher voltage to get the same V/cm electric field.

I'm starting to think any spike from unloading the 1300H secondary is, "in my mind only", as no one has address it.

Mikek

My first job interview, I told the guy that I preferred tubes to transistors because transistors were expensive and tubes were hard to blow up. He said "that won't do" and dismissed me. I said the same thing at the next interview and that guy laughed and hired me. I designed $200 million worth of stuff for him.

Actually, I just remembered, I did design two tube-type HV pulse generators for microwave spectroscopy, high school summer job. Both were 2-tube half-bridge topology, one box with hard tubes and one with thyratrons. Envision HV-insulated filament transformers and pulse transformer grid drives. Envision thyratron shoot-through.

1300H?

Using a DE-5000 at 100Hz, primary is 340mH and secondary is 1271 Henry. Approximately 295kΩ output impedance.

Mikek

DE-5000?

DE-5000.

Got it, but, the tube curves show up to 7 V negative to pinch it off. Do I need a ± supply to switch it on and off hard. Mikek

It's the original General Radio desktop Impedance Meter design, now cloned. IET bought the wreckage of Genrad, and moved manufacturing to China. I'm guessing that the DE-5000 is made by that original Chinese manufacturer, now freed somehow.

The patents are informative:

US4181949, US4196475, and US4342089 to Henry P.Hall of Genrad Inc, all in the 1980s.

Joe Gwinn

No, there is a simple, elegant circuit that needs just one small power supply, plus the HV of course.

Sloman says that I am a hack who can only make minor tweaks to my own circuits (which is an interesting bit of logic itself) so I think he should contribute the first schematic.

Of course, he won't use any of my dumb suggestions.

(I haven't used tubes since I was a kid. A 6BK4 actually makes sense here. No mosfet is going to have 27 KV Vd-s or 1 pF drain capacitance.)

Even if the neon transformer secondary is 1200 H, which I guess it could be, that's only 500K ohms at 60 Hz. Shouldn't matter.

Neon sign transformers are soft, by design.

What actually matters here is leakage inductance, probably a lot lower than 1200 H. Measure the secondary L with the primary shorted.

Those little RLC meters can be wildly off, too, measuring L.

I have made my own stab at a design, not understanding bias polarity and looking at a website for some info. I'm putting it out for corrections to learn something.

Mikek

The secondary measures 604 H with primary shorted. What does that mean?

Mikek

Let's give Bill 24 hours, plenty of time for something simple like this, and if he doesn't invent something good, I'll try a sketch.

Meanwhile, buy some resistors and a few 2N7000's.

That means that the primary and secondary coils are loosely coupled magnetically. Neon transformers operate into almost a short circuit, so they are designed to have high output impedances.

The 604H value is relevant here, since the primary will be driven from a low impedance AC supply. 600H is only 220K ohms at 60 Hz. Should be fine in your application. The rectified and filtered HV DC might droop a little when you pulse hard.

Don't see why; your current is a milliamp (or less), so simply driving grid and cathode in opposite senses will give you +12 to -12 off of a single +12 power source.

I'm getting info that you don't want the grid to go positive, just to 0V and negative a few volts, in relation to the cathode. As I understand it, with the grid at 0V the tube has no current flow, as the grid voltage goes more negative current starts to flow until it is turned all the on. Is that correct? Also, what is the voltage drop across a tube in full conduction? Does it get pretty close to 0V, or 10V or 100V? Mikek

Positive on the grid would help drive the plate closer to ground.

Look at the 6BK4 transfer curves. Plate current is increasing hard as grid voltage is increasing past 0 volts. We are literally in uncharted territory.