It is pretty consistent stuff. That is why the industry has used it for decades. Cincinnati Milling Machine (now Milacron)made the first machines that made the first 4 foot by 8 foot sheets of FR4 for PCB manufacture back in the late sixties.
Anyway, embedded cap plates are easy to make and characterize. One must be sure to place and manage the "access node(s)" carefully.
We used to use 25kV rated coil assemblies on, I forget the voltage, but something like 5kV. EHT does indeed demand plenty of margin.
NT
But it doesn't have an exit command. It's an infinite loop.
I was suggesting more complicated structures too, brainstorming. But those might not be necessary.
Equal-resistors is a special case, but it's exactly my current situation-- I made a higher-voltage resistor by connecting several in series.
This could all be air-wired or the intermediate connections made over slots, but I'd rather not require hand-assembly.
Cheers, James Arthur
I'd think other ratios, as long as they're not too wildly different, could be made by varying the area of the bootstrap halves and resistor connection (ratios matter, absolute value doesn't).
As I mentioned, my load is capacitive, in the 10 pF ballpark. So, in an ideal world, charging it to 1KV and then discharging it takes zero energy. Charging a 10 pF cap to 1 KV only takes 5 uJ.
Maybe 10 square inches of PCB!
The 10 pF cap will be a dummy load/pickoff, on a breadboard, so it's no big deal if it blows up. The actual load is an electro-optical thing that can probably stand 10s of kilovolts.
Trust me... it isn't zero. Maybe you are only delivering for a very short duration?
Yeah... that's not driving at a half an amp.
The best capacitive "test load" we came up with for HV was simple.
Take your favorite HV coax of whatever voltage and find out what its capacitance per foot value is and cut to length appropriately. Just remember to leave a stub of the center wire jut out past the main body so you can seal it off.
They are very accurate and very repeatable. Ceramics can garner issues as test loads over time and use.
DimBulb has always had a problem with power and energy.
Sorry, can't share that one.
I did the breadboard with a 10 pF cap as the dummy load + top of the capacitive divider, 1000:1, but ground loops make a lot of noise; I have insane dI/dT swarming around. I'm thinking that just 4.95K into the output connector, then to a 50 ohm scope, would be better, at
100:1 and no ground loops. The 4.95K could be three or four 1206 resistors in series. That would load my 1KV pulse at 200 mA, peak 200 watts, which I could tolerate on the breadboard maybe. All it could do is blow up.Have something to add? Share your thoughts — no account required.
Ask the community — no account required