Developing HV DC Pulses

John, It's to easy to get complacent about the voltages when your doing a test every 5 minutes, and soon it may be every 1 minute. So, I want a big reminder that HV is present. I did twice, the one day I visited the shop. To that end, this is the blinking red light I have ordered.

Mikek

Hi John I reported back to the designer of the 6BK4 LTspice subckt, his response is below. "I think he probably means the cathode to heater voltage. According to specs, it's 400V -ve and not recommended +ve. When output of about 4Kv, cathode has about 200V pulse which is still within heater to cathode voltage as heater is grounded. If not certain always keep the heater floating, grounded via a cap if hums are detected. The cathode can be shunt with suitable resistor for safety reason and to reduce heater to cathode voltage."

Also he ran the LTspice with the proposed circuit. I have ask to have C4 removed and to add a 25pf load capacitance. And also for an output graph, which he didn't run. He also used 1.5MΩ as the pull down resistor, I don't know that it needs to be that low. Here's the file as he posted.

If the cathode actually pulls itself up to +200, which should be tested, a small zener to ground would be prudent to protect the mosfet.

As far as the "not recommended" goes, I have based most of my career on doing not-recommended things. It reduces the competition.

Maybe just a flyback pulser, basically a car ignition thing, would do the oil separation thing. That would be very simple. The oil drops won't know where the electricity came from.

ps - a little corona never killed anyone.

I'm not sure how I get a DC pulse, doesn't the output of a car ignition ring, A lot!

Yes on corona, at this point I will do simple and better if needed.

I had the LTspicer lower the frequency, he added the 15V zener and a resistor(?). The latest output is here, such a low frequency the rise and fall times are obscured.

When it works. You do have to pay some attention to who giving the negative recommendation, and how they justify it.

There are quite a few legacy designers around, and what got recommended when the 555 was new technology isn't always what an expert would recommend today.

But they will know if the field isn't sustained for long enough to move them around.

Breathing in too much ozone can take a while to kill you - it may not kill you fast, but emphysema isn't a nice way to go.

"In the air, coronas generate gases such as ozone (O3) and nitric oxide (NO), and in turn, nitrogen dioxide (NO2), and thus nitric acid (HNO3) if water vapor is present. These gases are corrosive and can degrade and embrittle nearby materials, and are also toxic to humans and the environment."

It's not nice to give advice that can damage people if they take it seriously.

Are you sure the oil drops expect a DC pulse?

What does "DC pulse" mean anyhow?

Those pulses look pretty nice. 16 volts on the grid does need another power supply.

What's the grid current look like?

I could play with the sim if someone sent me the files.

I'm aware it's not good for the human body, I did get a whiff one time, that's why I ask about corona mitigation. Mikek P.S. My physicist friend had a pet saying, "out of an abundance of caution"

I have requested the file.

John, I would like to know if I can earth ground one side of the Neon Sign transformer. Will that cause a stress between the primary and secondary?

Some neon transformers, namely the higher voltage ones, like 16KV, have two insulated HV terminals. They generally have a secondary winding whose center tap is already grounded to the case. Ohm it. If you ground one side of such a transformer, it would be bad news.

The small ones, like 7 KV maybe, usually have a single HV output and the low side of the secondary is grounded to the case. Again, ohm it.

Modern neons are driven by high frequency supplies. They are small and cheap compared to the old 60 Hz transformers. I think the neon tubes look different with high frequency drive, frankly not as pretty.

I have a saying, "There are a million ways to not get things done."

My first mentor Melvin Goldstein said "There are a million ways not to sell."

I have two 9kV transformers both have a secondary CT. connected to the case.

I have not grounded the case, (as of yet).

If I don't ground the case can I then ground one side of the secondary?

Mikek P.S no word on the LTspice file

I sent the 6BK4 asc. file to your work email. I hope that is what you need. The SUBCKT writer said, if you need anything let me know. So, if you think there is something in the file that needs correcting he's willing to work on it. Apparently, tube subckts are a thing he does :-) Thanks, Mikek

Chinese patent (in English) says a lot and tells me nothing, other than, maybe we're going in the right direction. It says ns to ms rise and fall times and 10Hz to 100kHz frequency.

Here's a tube pulser:

That would make the case kilovolts live to ground, potentially deadly.

Ground the case and use one side. If you need more voltage, use a doubler or a c-w stack.

Got it, tweaked it, posted below.

Does that mean I need 24V to drive the grid? Does that mean I need a 60V power supply? I can do that, just want to know if that's what the circuit diagram means. Mikek

It seems to work well with +24 on the grid. The tube turns on hard and the pulsed grid current is about 25 mA. I assume Dmitry's model handles the positive grid case about right.

Turns out you can use a 50 volt zener from the cathode to ground, instead of the diode+power supply.

Run the sim and tweak!

It might be a tad different in real life, with a filament connected.

The push-pull drive version is nice too.