FAQ - Mark Rowley on the 30 & 60 kV precipitator supplies

If you have a question about this topic, the answer is probably in here!
Post Reply
User avatar
Richard Hull
Posts: 12604
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

FAQ - Mark Rowley on the 30 & 60 kV precipitator supplies

Post by Richard Hull » Fri Sep 11, 2020 8:51 pm

The following FAQ is authored in its entirety by Mark Rowley. Pay close attention to his caveats on the use and operation of these supplies as they require special care in operation. Many thanks to Mark for taking on this rather full exposition on these interesting power supplies. As always, in any FAQ, Reply with specific questions or additional information based on your own experience or corrections to any errors found. RLH
These switcher type supplies are quite common on eBay and range from $40 to $150. They are specifically designed as a replacement psu for a yet-to-be-identified electrostatic air purifier system. There are several variants but we’re mostly familiar with the 30kV and 60kV versions. Both advertise a theoretical peak current rating of 10mA.

For the purposes of general applicability, this FAQ will focus primarily on the 60kV supply.

Cautionary Advisement:
These power supplies are extremely fragile and unforgiving. The operational specifics outlined in this FAQ are not optional or arbitrary. Any omission or lessening of the requirements guarantees significant damage to the supply. Additionally, these supplies have questionable quality control. Several people have reported failures for no apparent reason.

-You have been warned-

See photos of the two types of supplies below...........

What kind of output does this type of supply provide?
These supplies provide a 100% pure floating output. There is no electrical commonality between the input and the output.

Do they really put out 60kV?
Yes, and then some. Exceeding the 60kV rating risks sudden catastrophic failure.

Do they really supply 10mA?
These supplies were designed to operate within excellent static field control conditions found in a professional grade precipitator type air filtration system. Under such perfect conditions, yes they may be able to attain 10mA. However, such a system cannot and should not be compared to a fusor. Comparatively speaking, a fusor is the polar opposite of such an electrically stable HV system. That being said, 5mA should be considered the absolute ceiling or red-line limitation.

Can I use this with a demo fusor?
No, they are not suitable for demo fusors. Demo fusors operate with comparatively poor vacuums which will exceed the current draw limitations of these fragile supplies. Attempting to do so will likely result with the destruction of your supply. Additionally, demo fusors should not be operated in a capacity to produce X-rays. The precipitator supplies can create unsafe X-ray conditions whereas a neon sign transformer is unlikely to do so. If you’re going to assemble a demo unit, avoid the precipitator supply and use a neon sign transformer, MOT, or OBIT. Lastly, demo fusors typically lack the proper high voltage routing to prevent arc over. And if that happens, one may as well get the trash can ready for a destroyed supply.

What size fusor will these work with?
As of this writing the supply has successfully powered two neutron producing fusors. A 2.75” conflat cross fusor with an inner chamber diameter of 1.36” and the other being a specialized chamber with a 2.3” inner diameter. Larger diameter voluminous fusors (4”, 6”, etc) typical demand much higher current ratings than this type of supply can provide. It it’s highly recommended to not exceed an inside diameter of 2.5” with a precipitator supply.

What conditions cause these supplies to fail?
Arc-over or deliberate spark testing is a death sentence. Failure after an arc can happen immediately or it can slowly dwindle from full output to 50% in several minutes. Damage from such an occurrence can be shorted secondary windings in the output transformer, blown rectifier diodes within the sealed output transformers, shorted MOSFET’s, blown IR2153 driver chip, something yet to be discovered, or a combination of everything listed. To prevent arc over, extreme attention must be given to how you route HV to the grid. Ensure that your HV feedthru and routing cables can handle a 20kV safety margin over your maximum desired voltage. For example, if your fusor is designed to limit at 40kV, make sure it’s arc-proof to 60kV. In my case, I’ve used 40kV wire insulated within 1” PVC white conduit which is then kept a minimal of 6” away from any structure. Between those 6” distances are thick HDPE or poly type plastic barriers. Additionally, the entire power supply is (and should be) submerged in mineral oil. The HV leads from the supply must be kept as far apart as possible and not routed over the top of the supply prior to their exit points from the mineral oil tank. I have yet to attempt a dry system utilizing static field control as you get no second chances if an arc-over occurs. Mineral oil, however, is as close to a safety guarantee as one can get.

The next killer is over aggressive powering up. These supplies demand very gentle and slow voltage increments as you increase your neutron output. As pointed out by others, operating a fusor is equally if not more complex than the build itself. This portion of the FAQ is to only warn the operator to take it easy and very slow.

Remember, general fusor operation is a totally different topic which is outside of the scope of this FAQ.

If you’re operating a 2.75” cross or similar, you should never need to exceed 3.5mA. In fact, 2-2.5mA is more than enough to get the neutron counter chirping to a respectable level (providing gas and other requirements are properly employed). If you’re over 25kV at 2 or more milliamps with no neutron detection, you have other issues.... it’s not the supply. So don’t keep pushing it up higher in hopes to make neutrons or you’ll be soon hoping to find a good deal on a replacement supply. It will fail.

Another death nail would be activating the fusor with over 60mTorr of pressure. Pressures higher than this will cause the power supply to incur a large power demand that will easily exceed 10mA. Typically this is an unavoidable fact during the initial phase of plasma ignition. But at over 60mTorr (especially with air), the recovery is lessened by the higher conductivity of the plasma. This type of pitfall offers a higher survival rate but why chance it? There’s just no need to turn it on above 60mTorr.

Last but not least is over voltage. This is a difficult condition to prevent as unexpected plasma extinguishment is very common when learning how to operate your system. When plasma extinguishes, the no-load conditions causes the voltage to instantly ramp up to 20-40% higher than what you were previously operating at. (The 20-40% range is dependent on where one has the trimmer control set on the main psu board...more discussion on this later). If the trimmer is set too high, the supply can easily jump up to 70kV causing anything from arc over to over taxing the rectifiers or worse. Once you become proficient at operating your fusor, minimizing the occurrence of extinguishment becomes fairly easy.

How do you operate the supply?
There are three main controls on the board (two potentiometers up front and a small blue trimmer pot in the middle. The front left control is the current limiter, front right is the main voltage control, and the trimmer is for adjusting startup voltage. Important to note is that these supplies are incapable of being set to zero volts. At a minimum, they float around 15kV when energized under no-load conditions. After testing these supplies it was found that startup voltage should be set to a no-load value of 23-25kV. This is done by adjusting the blue trimmer pot AFTER both of the main controls are turned fully counter-clockwise. Improperly adjusting the trimmer pot to voltages past 25kV risks enabling the maximum output to dangerously exceed 60kV! Once the proper adjustment is made, running the supply is just a matter of switching it on and adjusting the main voltage control pot. The current limiter should be left in the full-stop counter clockwise position. In that position, the supply is at its pre-programmed current limitation of 10mA. Unfortunately, this limitation hampers fusor operation as small plasma and gas variations cause very brief(milliseconds) current demands exceeding 10mA. Once over-current is detected, the supply goes into shutdown mode. To reactivate the supply, turn off the power, rotate the voltage control fully counter clockwise, wait 5 seconds, and turn it back on. If your fusor enters shutdown mode too frequently, there are two options. The first requires disabling the automatic current limiter. It’s a simple operation. On the upper left side of the power supply there is a small black inductor coil with a red wire going through the center.

See Photo below of limiter modification

re-route the red wire around the inductor (as shown in the photo below) to disable the current limiter. Once complete, determine the size and value of a suitable ballast resistor and add it to your fusor. The other option requires a slightly more complex modification to the psu as lightly discussed in Rex Allers post on Precipitator Hacking. His highly detailed report about the electronic design is mandatory reading before any attempt is made to use these supplies. The following link will redirect you to this thread:


The post includes a discussion about the possibility of changing the value of a resistor in the current limiter circuit to increase the level of current needed to initiate shutdown mode. As of this writing no one has yet to try the modification although it seems likely to work very well. If employed, there is no need to add a separate ballast resistor to the fusor.

Misc Thoughts
Although I had decent success with this unit powering the 2.75” cross (1.36 inner diameter), I’ve since learned that size of a fusor poses inherent risks to the supply. Its small size can contribute to internal plasma arcing to the wall, electron runaway effects within the grid stalk, and an overall lack of stability. Those fluctuations pose a constant risk of plasma extinguishment and other conditions which are rough on the supply. It wasn't until after building the larger 2.3” inner diameter system where it became known that the increased size totally stabilized operation. This change allowed the supply to safely and easily initiate neutron production capable of a wide range of activation experiments.

From what has been learned so far, a 2.3-2.5” ID chamber seems to be the sweet spot with this type of supply. Any bigger and current demands become excessive. Any smaller, instabilities occur putting the supply at increased risk. To be fair though, a small conflat cross system is very stable if you’re happy with just making easily detectable neutrons of a low count rate. You could literally switch it on, get it up to 25-30kV / 2.5mA and it’ll sustain an easily detectable neutron count till you run out of gas. But if you want higher numbers and you start cranking up the input, instabilities become an annoying and risky factor.

If you choose to use one of these supplies, it’s almost required the fusor be built to suit the supply, not the other way around. And if done correctly it’s a very cost effective option.

Mark Rowley
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

User avatar
Mark Rowley
Posts: 661
Joined: Sun Dec 29, 2002 5:20 am
Real name: Mark Rowley
Location: Sacramento California

Re: FAQ - Mark Rowley on the 30 & 60 kV precipitator supplies

Post by Mark Rowley » Mon Nov 16, 2020 3:59 am

On 12Nov20 during a high output activation run I barely raised the deuterium flow while the voltage was in the 50kV range. Even though the increase of flow was slight it was enough to apparently spike the current past 10mA for a fraction of a second. This instantly over taxed one of the four transformers causing either the secondary to short out or the internal full wave bridge rectifier to fail (autopsy forthcoming). Either way, this resulted in its failure as the supply was limited to ~10kV instead of 60kV with almost no current capabilities. Replacing the transformer quickly fixed the issue.

Diagnostics was a simple task. Using a cheap $25 ZVS driver each transformer was tested for their standard 15kV output. The bad transformer only supplied about 1kV.

Moral of the story, lower the voltage to a safe level before any increase in deuterium flow. In fact, probably turning off the power and cautiously walking it back up is the safest bet.

As a side note, this particular supply has been operating without issue for over a year.

Mark Rowley

User avatar
Bob Reite
Posts: 436
Joined: Mon Aug 26, 2013 1:03 am
Real name: Bob Reite
Location: Wilkes Barre/Scranton area

Re: FAQ - Mark Rowley on the 30 & 60 kV precipitator supplies

Post by Bob Reite » Wed Nov 18, 2020 6:06 am

A few of us have attempted to "remote" the voltage adjust pot. That is extend it with wires to a safe distance from the supply so that it could be adjusted while the system was "live". This resulted in instability and eventual damage. As I recall it was determined that the control actually varies the switching frequency. Voltage is raised by moving the frequency closer to the self resonant frequency of the transformer secondaries. Anyway it would be better to just add a long fiberglass extension shaft to the pot, so that the circuit is not disturbed.
The more reactive the materials, the more spectacular the failures.
The testing isn't over until the prototype is destroyed.

Post Reply