High Voltage Power Supply (0 - 1,500 V)

I needed a 1-2 KV power supply for capillary electrophoresis experiments. Serendipitously, I saw a couple of high voltage power supply modules in a pile of electronics junk at the De Anza flea market in Cupertino.

I took a weekend to build a lab quality High Voltage power supply from the 1.5 KV module. The first image list the specifications, the second image shows it in operation. Or, check out the YouTube video.

I used the following parts (total cost of < $25):

• The flea market high voltage power supply module ($5)
• A 24-volt wall-wart supply that I got through freecycle.org ($0)
• Two digital voltmeters from ebay ($4, I think)
• A 10-turn potentiometer ($6, this one did not work so replaced with a 1-turn from my junk drawer)
• Two resistors (12k, 15k) and two variable resistors (20k, 100k) bought as assortments from ebay
• Two binding posts ($?)
• An AC inlet socket (harvested from a DVD player)
• A power switch (from a hard drive case)
• Thin plywood scraps to build the enclosure
• Wires, solder, paint, mini-plastic grates, aluminum foil, etc.

Downloaded the Spellman HV power supply module manual. The supply has two 0-10 V monitoring outputs for 0 - 1.5 KV output and for 0-6 mA of output current.

When I connected the HV power supply module (first image), I saw no voltage on the voltage monitor pins. Changing the variable resistance (image 2) did not change the voltage. The small adjustment screw on the variable resistor kept turning more than 10 or 15 revolutions, indicating a defect. I removed the variable resistor and ohmmeter showed a short. Replaced with a standard 10k pot but no monitoring voltage.

Some points on the circuit were showing a short with the ground, a visual check showed that the copper shield was touching one of the diode leads. I created a temporary gap with a plastic probe (5th image) and got an output voltage.

Decided on a 6 in x 3 in x 6 in size enclosure that just fit the HV module board and other components and was approximately the size of my existing power supplies.

The second image shows the enclosure construction plan. Started cutting the 1/4th inch plywood scrap panel till I had the 6 pieces needed. I butt-glued the edges of one side panel (6 in x 6 in) to the top and bottom panels (3 in x 6 in). For the back panel (2.5 in x 5 in), I cut out a hole for the IEC AC input socket. Attached the AC socket and glued aluminum foil to the inside surface of this panel. Aluminum foil was glued to the inside of the C-shaped piece (top + side + bottom). The back panel with its AC socket was then glued to end of the C-shaped piece.

For the front panel (2.5 in x 5.5 in) I laid the components on it till I was happy with the layout. I then drew the cut-out dimensions on the back surface of the panel and cut these with a drill press, a saw attachment on the dremel, and/or a chisel. The front panel was also coated with aluminum foil and then glued to the box sides. Once the glue was dried I did a test fit of the front panel components (last image).

The first image shows the layout of the label. The laser printed page was glued onto the front panel. The specifications label was glued onto the back panel. The outside of the open-sided box and the separate left-side panel was painted with black acrylic paint. The inside foil covered area and the labeled panels were covered with acrylic varnish.

I like using binding posts that are the standard 3/4th inch apart but was not sure how safe these would be at 1.5 kV so created a spark barrier from a small piece of L-shaped plastic (last image).

I opened up a 24 Volt Power supply. Removed the AC connector by de-soldering the two pins (image 3) and soldered in two AC wires where the AC connector pins were and also drilled in a couple of holes where there were no copper traces for attaching the PCB to the back panel. I wrapped a layer of shipping tape around the edges of the 24 Volt power supply.

I connected the "Live" terminal of the AC power socket to the power switch and the other terminal of the power switch was connected to live wire of the power supply module. The neutral wire from the power supply-board was connected directly to the AC socket.

I hot-glued two 2 inch square plastic grids (from a dollar store) to the back panel to separate the back of the 24 V power supply PCB from the aluminum foil. The board was then fixed to the back panel with two short screws.

For the digital panel meters I would have to scale the 0-10 V monitoring voltage to represent 0-1.5 KV and to represent 0-6 mA.

The internal resistance of the ebay digital panel meters was about 80 k ohms so I calculated resistances for a voltage divider that would reduce the 10 V to 1.5 V. I would need a 68 k ohm resistor hooked up to a 12 Kb resistor. So used a 100 k ohm variable resistor and adjusted this to get a 1.5 V reading when I input 10 V on top of the voltage divider. Same for the current sensor, here I put a 20 k ohm variable resistor in series with a 15 k ohm resistor.

I hot-glued the variable resistor to the back of the volt meter and then soldered the fixed resistor in to one pin of the variable resistor. The other end of the fixed resistor went to an empty ground hole in the voltmeter PCB. I then glued these small digital volt-meter assemblies into the front panel. It was hard getting the glue gun nozzle in to that tight space therefore the messy glue job.

Now connecting the high voltage power module. First soldered the low voltage connections to the pins on the HV module PCB. Then soldered the earth wire from the IEC AC socket to the copper ground shield. And finally the high voltage leads were soldered to the binding posts.

A pre-test of the 10 turn 5 k potentiometer showed that one of the end-leads was open. Removed it and replaced it with a single turn 5k ohm potentiometer from my parts bin.

Did a careful check and followed it up with an ohmmeter check to see that there were no shorts.

To the bottom of the HV voltage module I had hot glued two of the plastic grids to create a gap between the bottom of the PCB and the foil-covered bottom of the box that was also covered with 3M shipping tape.

I wedged the HV module PCB into the enclosure and was now ready for some testing.

I applied an additional coat of paint, let dry, followed by acrylic varnish.

Plugged the unit it in and saw that the meters came on and the voltage increased when I turned the voltage adjust knob. I then connected my multimeter (set at 1000 V DC) to the binding posts and tried to match the readings between the multimeter and the PSU voltmeter. I tweaked the screw on the variable resistor glued to the back of the mini-digital voltmeter till the readings matched. Tried these at different values and the values matched quite well. The last image shows a voltage out of 0.90 kV and 0.04 mA.

So, one weekend, a small number of parts, and a really useful lab quality HV power supply.