X-ray transformer - zero output

[Rich Gorski]

Taylor,

I agree with Dennis. I would also build this into a plastic container although it is true that all commercial HV transformers I've seen are built into metal tanks that are well grounded via welded studs. If you decide to use a metal tank make sure the transformer, diodes and wiring are mounted on a sturdy insulating platform and secured inside the metal tank so things can't move around. Also make sure the grounding wire is secure and can't come loose.

One issue with plastic containers is to be sure it won't collapse when under vacuum. Maybe test the strength of the tank by filling with oil (without the transformer inside) and evacuating it. It will help if you have a clear window or use clear plastic so you can watch for air bubbles forming. Then you know that the vacuum level is good enough and degassing is working. Plastic tank and lid also makes it easier to handle the high voltage electrical feed through since these are mounted in an insulator. Once the transformer is in the tank and the oil degassed you don't want to pull the transformer out again else you will have to go through the degassing process again. The main thing you want to accomplish with the degassing is to remove any air pockets within the transformer. Vacuum degassing should do this but moving and tilting the tank around will help. Once the degassing is complete and the transformer is completely submerged the tank will come back up to air pressure. ATM pressure is fine as long as air pockets have been removed and the transformer stays submerged.

Rich G.

Ps. The post from Andrew Seltzman that you referenced, shows that he built his transformer in what appears to be a 5 Gallon bucket and had all the electrical feedthroughs through the bucket lid. The post did not indicate whether he went through the vacuum degassing.

[Rich Feldman]

Got to add usual comment. Pad-mounted utility transformers and serious x-ray transformers
have their tanks evacuated before any oil is admitted. Rectangular tanks are reinforced to withstand vacuum.
No bubbles ever.
WHen you add oil before vacuum, you can't get 100% removal of bubbles. Bubbles expand only so much.
Vacuum pressure at micron levels, above the oil, is meaningless for bubbles more than a few mm below the surface.
Surface tension also limits bubble expansion, even if the vacuum were perfect and you could suspend gravity for a few minutes.

[edit] Many amateurs have been successful after dunking HV transformers into oil, then using vacuum and heat and vibration and tilting to remove most of the air bubbles. (And gas dissolved in the oil, which is handled separately by professional transformer servicers.) But at this forum more than 10 years ago, I'd bookmarked a couple of posts where XRT saga ended with destruction by arcing under the oil.
I never got that far myself, while scrounging a succession of XRT's over the years, and reporting on several without ever developing kilovolts.

[Taylor Romain]

Here are some progress photos of the transformer mounted inside the container. The setup includes banana plug connectors for AC input, voltage measurement, and current measurement. Given the limited space between the required oil level (which must cover all components, including resistors) and the vacuum port, there may be some challenges during the degassing process.

Heated mineral oil will be added to the box, and I plan to attach an Edwards FL20K foreline trap, which uses activated alumina as a sorbent material, to the box. A KF25 hose will connect the trap to an Edwards pump (12.1 CFM, 32-micron ultimate pressure). The pump’s gas ballast will help prevent oil or water from entering the inner chamber and contaminating it.

[Dennis P Brown]

I pulled a vacuum on my transformer then open a valve and allowed oil to fill the chamber holding the transformer. So much of the water vapor and most of the air was removed.

[Taylor Romain]

Did you find any oil wanting to travel through the vacuum system when dripping oil through the valve?

[Rich Gorski]

Taylor,
Looking at your photos there’s a couple of suggestions I have.

1) Having wood inside the tank with mineral oil is probably not a good idea. Wood might leach out who knows what that could contaminate the oil and drop its breakdown characteristics. I assume the wood pieces in your photos are for support etc. Switch to acrylic or HDPE plastic. Acrylic is pretty cheap, easy to drill, cut and machine and you can get it in rods or blocks.

2) I see banana jacks used for terminals. I hope one of them is not for the high voltage output. HV transformers always have the high voltage terminal well separated from any low voltage terminals using a tall ceramic feed through. The HV feedthrough is on one side by itself and the low voltage terminals on the other side. The simple way to have the HV exit the tank is to solder (or bolt) the high voltage cable directly to the transformer HV connector and come out through the top as a high voltage flying lead.

3) It would be best to have the full wave bridge also in the tank. Make sure there is room for that and also room for the high voltage measurement circuit. Of course you could mount the bridge and measurement circuit in another tank with oil but then you will have several high voltage flying leads going around your work space.

4) There will probably be a layer of air at the top of the tank so make sure that the high voltage output lead wire is fully insulated where it passes through the air.

Rich G.

[Rich Feldman]

Here's another anecdote about letting oil slowly be sucked into an evacuated tank, like Dennis did. More than 10 years ago, a reader named Carl (not Willis) graciously gave me a gallon of Shell Diala AX electrical oil, from a 5 gallon can he'd procured. We had a nice talk in his garage, in Milpitas (IIRC), and he showed me some HV transformer stuff. He said that at work, they would admit oil into evacuated transformer tanks through a spray nozzle near the top. Oil falling through vacuum in fine droplets was a final opportunity for residual gas or moisture to escape. (I still have the gallon can, unopened.)

I guess Taylor's banana plug sockets, like fluted HV bushings, are chosen to be vacuum tight. Beware that the lid will not be flat while pot is evacuated. Will pot be strong enough for vacuum when mostly full of oil that's "hot" ? There might be opportunities for internal bracing/reinforcement of large flat sides, but that would be a lot of trouble. Says someone whose closest attempt at vacuum potting was never finished. viewtopic.php?t=4050

Rich G is probably right to urge caution in using wood. But In large utility transformers, wood and paper are among the most common solid insulating materials. In many cases, they're dried for days under vacuum, before first exposure to oil.
https://www.electricalclassroom.com/ins ... nsformers/

An Internet search for transformer wood brings up special resin-impregnated plywoods for example:
https://www.mueller-ahlhorn.com/en/proc ... rmer-wood/
https://www.roechling.com/us/industrial ... ansformers

[Rich Gorski]

Here's another method used to degas liquids... ultrasound.

Crazy idea?... Maybe attach two piezoelectric tweeters on the outside of the tank. One on each side. Maybe 25 watts each at 50kHz and turn them on during evacuation. Similar to an ultrasonic cleaner but here the oil would be under vacuum and heated.

Rich G.

[Rich Feldman]

Sounds promising. Dont go overboard on vacuum strength unless oil is being circulated; 10 microns is about the same as 10 torrs for oil at the bottom.

Oil and windings can be heated at will using DC current from outside, noting that time constant may exceed 1 hour. I think that won't work for secondary on far side of a full wave bridge rectifier. But DC output could be shorted, and heat power dialed in with low voltage AC on primary; a ballast R wouldn't hurt.

[Daniel Harrer]

A suggestion in case you cannot figure out a way to evacuate your box for filling: evacuate and fill with oil in another container. Then re-add air quickly move the transformer over to the already (mostly) pre-filled box. This way the oil got into all the finer crevices and due to surface tensions should mostly stay there for short times. At the end use every other method you can think of (ultrasound sounds interesting, but I have no idea how well it works).

For that initial fill you could use something similar to my transformer's chamber, without the cables:

I built my (quite similarly looking) x-ray transformer into what once was a cheap vacuum chamber from eBay: effectively a large stainless steel cooking pot. I made my own acrylic lid where all the cables go through, as well as a one-way (to avoid wet air drifting into it later) valve where one can apply vacuum. To keep things airtight I added standard silicone on top of the pot.
To fill under vacuum I drilled a ~8mm hole at the top of the steel pot and soldered a brass connector onto it. Thus one can evacuate through the valve, then add oil through the brass connector.

[Dennis P Brown]

Rich Gorski is correct. Wood is a bad idea for high voltage work. Besides the contamination issue (water no less), it conducts at high voltage (again, water). I also agree with all of Rich's suggestions. They are very good ideas with any transformer going above 25 kV. The others have good points as well.

Walmart sells a all glass candy holder (w/lid) that is super thick and was big enough for my x-former. I used a plastic (thick) cover, drilled a feed pipe and used a valve to admit oil. Another (largest) hole with a copper tube (epoxy sealed) was used for the vacuum pump access. It was very simple, cheap and easy to make. I use synthetic auto oil since that is readily available and has excellent HV properties. My pump was a standard two stage oil pump so certainly below 10 microns was achieved.

[David Kunkle]

Rich,

Do you know of anyone actually doing this for degassing electronics in oil?

Because I have personally burned holes in thin metal under water (and unfortunately, not on purpose) doing essentially what you propose. Ultrasound gets absorbed at the the interface of 2 different densities- like oil and electronic components. Without the transducers constantly moving, the energy tends to get directed to and absorbed at certain spots. In therapeutic ultrasound, patients will get a nasty little burn at the periosteum (bone/ muscle interface) if the tech stops moving the sound head for long. I have a pretty good idea what it would do to electronics.

[Rich Gorski]

David,

I know very little about ultrasonic degassing except that its used often in industry. I have no experience with it. I think the process of degassing a liquid for electronics means you do the degassing BEFORE you insert the electronics. I just mentioned ultrasonics as another method of degassing.

Honestly, I would proceed without the ultrasonics and maybe even without vacuum degassing. Basically, insert your transformer and any electronics in the tank, fill with heated oil, shake and tilt it around to release any trapped air pockets. Then make all your connection to terminals. If you want to pull vacuum on it this would probably be time to do it. FINISHED... TEST AND SEE HOW IT WORKS. Run it for a while at 1/2 rated voltage like 30kV and see how things behave in the tank. The electric fields and the RF inside might also help to degass the oil.

I know you want to get to 60kV but first see if you can get to 35-40kV without problems. 60kV will take some time and you have to approach that level slowly correcting things as problems appear. You can do an awful lot at 40kV including activation.

Rich G.

[Daniel Harrer]

I wouldn't just test and see how it goes. If insulation fails due to air pockets the resulting arcs can and probably will damage the transformer. At half the rated voltage it might survive (maybe even at full), but it is a gamble. And if breakdown happens then it is too late. I am not sure if there is any realistic way to measure if the insulation will hold as-is; an insulation tester for such voltages probably exists but is very expensive, more than a new transformer.

I've also heard that various people were successful with just placing the transformer in oil for a few weeks in various positions. So the air can mostly bubble out of all but the smallest pockets, some also dissolves into the oil. When I tried this even under mild (~1 mbar) vacuum, I kept getting more and more bubbles each time I pumped a bit more, but always from sharp edges. So I think that I saw dissolved air or water becoming gaseous again. After a month of this it worked at 30 kV (exact rating unknown, but at least 40 kV). The internal voltages are maybe twice that because it is center-tap grounded.

But in the end and for full voltage I did what I wrote in my previous post.

[Rich Gorski]

Daniel,

I’ve done some research on the oil degassing process for HV transformers. Here’s a procedure that could work. It requires two oil tanks both that you can pull a vacuum on. One tank is an oil reservoir tank (RT) and the other is the transformer tank (TT). These are connected by a hose with a shut off valve. The upper RT will gravity feed pre-degassed oil into the lower TT. The upper RT will also need a valve on top to let air in as you fill the lower TT with oil. Here’s the step by step process (more or less). At this point the transformer is in the TT, all electrical connections made and ready to pull vacuum.

1. Fill RT with proper amount of oil.

2. Start vacuum pump which is connected to both tanks. The oil in the RT will start degassing and the air will be removed from the TT. I have no idea how long to let it pump. Maybe when the bubbles start to diminish or stop entirely. If you try ultrasonics the transducers would be on the RT.

3. Slowly open the valve between the two tanks and start filling the TT. You might have to start letting air into the RT to allow flow.

4. When filling is complete do your best to tilt, shake, vibrate the TT and take note of any air bubbles that are released. Hopefully you will see none or after enough shaking they will stop.

5. Seal off the TT and your done.

Good luck.

Rich G.

Ps. For what its worth, at the university I did two oil fillings without any type of oil degassing. One was a very large 3' x 3' x 3' HV transformer that we needed to rewind for a different turns ratio. The other was an aquarium tank to house a HV multiplier. Both were capable of > 100kV isolation. Even without any degassing both worked without an isolation or arcing problem within the tank.