I did some measurements using 3.1nF caps that had no voltage label, they came off a CW multiplier with 8kV diodes. So, they probably didn't meet the Voltage I put across them with the doubler suggested. It all worked OK at 5kV and 1ma load. But I pushed it to 8kV where one of the caps arced, so I'm done until an Amazon delivery today of some 20kV 4.7nF caps. It's looking promising, but I want to find out the droop under load with 4.7nF caps. Mikek
Arrgh! I'm pretty far in and realize the setup we have has a 12MΩ load, The proposed 10MΩ dropping resistor going to the tube will reduce the output to 4.4kV. I could increase the resistors in the 8 resistor voltage divider, but the risetime with 20pf per vessel is already 250us, worst case. I can lower the 10MΩ dropping resistor to 6MΩ and raise the voltage to 12kV to get back to the desired 8kV. That does push the limit of tubes 1.6ma average current, but if we don't go over 50% duty cycle, I'm thinking it's still OK. Your thoughts? Also, Can this be made into a tripler and quadrupler?
That's just the first stage of a Cockcroft-Walton multiplier; the 'CT as a reference' is ambiguous; your center tap on the secondary is ground, and the symmetric ladder will generate (+) and (-) legs, which could be applied to two HV electrodes to do the electrophoresis (alas, the switching of the negative leg isn't trivial with your vacuum tube).
There are *so* many ways to make a simple thing hard.
Why bother. A Cockcroft-Walton multipler driven by a fast oscillator - 10kHz to 100kHz - can charge up and discharge fast enough to deliver the on and off times that Mirek wants without any fancy switching.
And so many ways to screw up if you don't bother to understand what you are doing and skimp on the detailed design.
OK, I have a double sided tripler working, (although low volts, 5Vac in 15vdc out) I don't get all the ripple? Here are the wave forms of each side independent and loaded. The second picture is with the diode on the right side connected and all the load moved to the upper side and center tap/NEG. The voltage does go up, i.e. it will supply more current. I thought the ripple would be reduced, it is actually increased. Any thoughts?
P.S, guys remember, I'm the fishmonger, I'm doing the best I can with what knowledge I have. I don't want to take the time to develop a high frequency supply, and I probably don't have the ability. John, "There are *so* many ways to make a simple thing hard." I don't know what you are addressing, Is it the double sided tripler? I may not need it, but I'm close on getting the 12kV. Or is it the 12MΩ load requiring a higher input voltage? After experiments, we may be able to reduce this to one vessel, but for now we want 4 vessels run at 4 increasing voltages.
Forgot to add the doubler schematic.
Does that make 3Vp?
No sorry, left over from copying the schematic.
5Vac in 16.8vdc out, 2.8kΩ load Waveform below.
Ah, simple output cap, I tried that earlier, but it killed the output, I expect i had it misplaced. Mikek
I have a 6kV input, as initial test. I'm using a 2.5MΩ dropping resistor an a 4.5MΩ load. That voltage divider gives me 3,844V on the plate. When the grid is energized I get a 3,678V pulse riding on 166Vdc. About 1.5ma through the tube. Still need some additional parts to finish the tripler and I need to put a 30 watt 6MΩ resistor together. I'll see what I can get, I can probably get by with 20 watt, but I'd rather run it cool.
Thanks for all the help I was given to get this up and running.
Mikek
Have something to add? Share your thoughts — no account required.
Ask the community — no account required