What is a bipolar power supply and why do we need it?
In the Ra EDM experiment, a high electric field flips sign on a roughly 1-minute cycle. The direction of the field is changed by reversing the sign, or polarity, of the voltage being applied across the “hot” cathode (field is defined by positive-to-negative charge direction). A bipolar power supply is one which can provide both positive and negative voltage output. To fully condition the HV electrodes, we needed to upgrade our unipolar power supply (may it rest in peace) to one that would allow us to condition electrodes in an environment analogous to a real EDM run.
What’s involved in installing a high voltage power supply?
At Spinlab we need several things to use a high voltage power supply effectively and safely:
- A digital “interface” that can send and receive signals from the power supply and relay them to a desktop computer
- Software that can parse the digital interface and remotely control the high voltage power supply
- A physical location for the power supply to be installed
- A controlled test of all the new stuff
I set to designing and building a digital interface. I could have altered the interface box that was used for the unipolar supply, but we may want to use the old supply for testing at some point so I decided on starting from scratch. I put some considerable effort in being tidy with the wiring (cable guides are a godsend).
Jaideep programmed the conditioning software we use to parse the new power supply’s signals. One big change was implementing polarity reading and writing. For the bipolar power supply it requires 2 signals to output a desired voltage:
- A high/low signal telling it to output negative/positive voltage
- a signal telling it the magnitude of the voltage. We can go from one sign to another by hitting the big “Swap Polarity” button on our interface. The LED’s tell us what polarity signal we’re sending and what polarity the power supply is receiving (these should always be in agreement).
A safe place to put high voltage equipment is within a grounded metal cage or “coffin.” We drilled a few holes in our high voltage coffin and bolted the new equipment in. This keeps the power supply cleaner (it’s inside a box) and keeps the conditioning crew safe!
To make sure there aren’t any bugs in the equipment or software, we isolated the power supply from our electrode chamber and hooked up the lead to a high voltage divider (see above). The high voltage divider acts like a normal voltage divider, but it’s unique in its precision and ability to handle high voltages. Its low-voltage output to high-voltage input ratio is about 1:900; this keeps voltages that we read (on the low side) to below 35 V. Since the high voltage divider ratio is known accurately, it can be used to verify the high voltage power supply output. In our test we tested negative and positive polarity. We determined the magnitude of the high voltage power supply from its analogue output signal and measured the high voltage low output directly with a 7-1/2 digital multimeter. Below is a scatterplot of high voltage divider (HVD) output vs. high voltage monitor (Vmon) input. A weighted linear fit gives a slope of 1.12 V/kV, which corresponds to a voltage divider ratio of 1:890. Since the fit agrees with the manufacturer’s specified divider ratio, we conclude that the high voltage power supply supplies the correct voltage and readback signal. We’re ready to condition again!
Bonus Trivia & Pro Tip
Our old single polarity HVPS is from Acopian. The new bipolar supply is the HPR series from Applied Kilovolts. In addition to being bipolar, the new supply (HP030RIP020) is considerably more stable (less temperature sensitive) and has magnetic shielding built into it. In fact, an older version of this supply is the one that the Seattle Hg EDM experiment uses. We tried to order exactly the same model (LSR series), but after almost 7 months, the manufacturer was not able to make a new one for us. Apparently the last one they sold was twenty years ago (I suspect to Seattle group!) and they were unable to reproduce the old design nor update it using modern components. Caveat Emptor: they still list this discontinued model (LS030RIP010) on their website. – RR
Share this:
SPINLAB @ MSU 
You must be logged in to post a comment.