0
\$\begingroup\$

I am trying to explore good options that allow me to control the voltage output of an HVPS from 1KV to 10KV DC voltage. I have very little experience dealing with voltage range above 1KV. I understand that you need to be extremely cautious when you work with HVPS. Any thoughts on control mechanisms on the output voltage and any safety precautions will be more than welcome! Thanks!

\$\endgroup\$
  • \$\begingroup\$ Does your high voltage power supply have any adjustment or 0-10V interface even? \$\endgroup\$ – winny Sep 6 '17 at 20:47
  • \$\begingroup\$ What's the max wattage expected on your load? Or, the max current? \$\endgroup\$ – wbeaty Dec 10 '17 at 9:38
1
\$\begingroup\$

What's the max wattage expected on your load? Or, the max current? Can you tolerate some 120Hz hum, or do you require a cleaner signal?

Also, why not just find an old used controlled-hvps, with 0-5V or 0-10V ctrl voltage? Spellman CZE1000 or X3000 are 10watt, 30KV, max 300uA output. They use a 24V 1.5amp DC supply. They're often $100 on eBay, sometimes much lower. Actually these two are those sealed boxes found inside the expensive rack-mount Spellman supplies, the ones with the big panel meters. They can be controlled with DC signals, or just use a couple of 10K pots (to set Vout and current-limit.)

Also, for 30watt and below, you can find little HV modules on eBay, search Emco and Ultravolt.

If you must build your own from scratch ...typical modern supplies control the HV by controlling the input to the step-up section (control the 120VAC with a dimmer, or adjust the regulated DC that runs the HV.) For low noise they'll use a power-oscillator driving a high-freq transformer and diode/capacitor voltage multiplier. A high-ohms divider (100M or 1000M resistor) measures the output and provides a feedback signal for your regulator.

But for old-school low-watts supplies, the usual method was "shunt regulator" using a vacuum-tube triode to partially short out the HV output! Sorta like an adjustable 10KV zener diode. These gave a clean output without needing a (noisy) high-freq oscillator section.

For more than 30watts or so, projects with funding would instead use a "plasma generator," a high-power RF source at 13.56MHz or 27.2MHz. These can create discharge without chamber electrodes, just a tuned coil outside the glass envelope. Lots of items on eBay from ENI Co., search: ENI plasma, or ENI rf

\$\endgroup\$
0
\$\begingroup\$

Our standard approach at the plasma lab was to run the primary of the HVPS from a variac (some of which were rather large and expensive variacs in order to supply the needed current.)

I suppose variac is a brand name, and "rotary autotransformer" might be more generic, but it's a commonly used name for them. Anyway, we simply reduced the input voltage to the HVPS and got a proportional decrease in the output voltage.

Be careful - it does not take much of a charged capacitor to do you harm when it's charged to kilovolts.

\$\endgroup\$
-1
\$\begingroup\$

In college I assisted the manager of the physics storeroom, with his 5,000 volt regulated (constant current) for Carbon Dioxide laser. Once the CO2 had arced over from the highvoltage, and the plasma formed, he had to limit the current. He used a Triac in the Transformer primary, and did not attempt any regulation at the high voltages.

He only had 120 regulation opportunities per second. What is your desired control bandwidth?

\$\endgroup\$
  • \$\begingroup\$ Huh - we just used a big honking HV resistor to limit the current on our CO2 laser... \$\endgroup\$ – Ecnerwal Feb 8 '17 at 3:32
  • \$\begingroup\$ @analogsystemsrf I am building an HVPS for a corona device, my lab didn't give me any specific bandwidth or sampling rate requirement, what would be a good sampling frequency for current limiting in your experience? \$\endgroup\$ – Andy Wang Feb 12 '17 at 10:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.