Power Supply Upgrade

[Emma Black]

This is currently in the pondering stage, will update this thread as the upgrade starts to happen down the line. I would be keen to get a bit of a sanity check from those with a lot more experience.

I am using a 110-33KV 50hz non-centre tapped potential transformer, fully rectified. Its specs work out to 15ma at the HV end. I assume this is designed for 24/7 operation so can likely be pushed a couple of times this for shorter periods. Have three of these so very much inclined to keep using it.

From the other thread I mentioned three options for upgrading, two of which are to A) multiplier at 50hz B) second transformer in between the bridge and the PT.

B, after playing around with some numbers, seems to need either a massive core or a very large number of turns when at 50hz.

For A, if I was to accept say 10%-15% ripple using a single stage multiplier the capacitor size looks more reasonable. Obviously still quite dangerous and potential for damaging things. How much of an issue would the ripple cause, inclined to think not much based on the supply currently?

Been using this for the multiplier calculations:
https://www.extremeelectronics.co.uk/ca ... alculator/

[Bob Reite]

Ripple is not an issue as people using old X-ray transformers just use the raw unfiltered output of the full wave rectifier. It is critical that an appropriately sized ballast resistor be used in series with the supply output to limit the current in case of a fault.

[Matt_Gibson]

If you have three of them, maybe you could get away with putting two in series and then rectifying? Not sure about their insulation ratings…

-Matt

[Emma Black]

I did think about this Matt, it would certainly be the cheapest option. Unusually perhaps I actually have not only the spec sheet but a test report for each transformer.

Because they are supposed to be used in reverse they are designed to cope with voltage spikes in the line. These in particular can deal with 70kv for 1 minute or even higher for a shorter duration. I assume that this relates to the duty cycle i,e, heating of the core but what I am unsure about is how this would marry up with using the transformers in reverse.

Maybe a silly question but if I ran both with 220v would the insulation issue would be the same? I.e:

[Richard Hull]

I do not think it would make much difference how you wire the primaries as they are ground referenced, as will be your HV on one side at the fusor.
Only a test will tell the tale related to insulation.

Richard Hull

[Matt_Gibson]

I’d put two of these in series and rectify before messing around with a multiplier running at 50hz. I’d steer clear of the multiplier route until you start looking at 20khz or higher using a switcher+ferrite transformer combo.
-Matt

[Bob Reite]

With either of the two primary configurations, I would use a full wave bridge rectifier on the output. That will give over 93 KV peak if I remember the math correctly, and without the danger of large capacitors as used in a voltage multiplier.

[Richard Hester]

I would also chime in to discourage the use of a 50Hz voltage multiplier - the caps required would be huge, and the stored energy extremely dangerous.

[Emma Black]

Thank you for all the replies, it testing goes well, two in series and then rectified would work nicely. Without the need to store dangerous amounts of energy or spending a small fortune on capacitors. I also have everything already.

Cant imagine I would need more than 60kv for now. Before the rectifying this puts roughly 42kv on the 36kv rated transformers, my feeling is that this will be likely ok.

[Richard Hull]

with secondaries in series that is 66kv rms at full primary input voltage. Full wave rectification with no caps for filtering that would mean 1.414 X 66 = 93.3 kv peak voltage to the fusor at 100hz ripple. A Variac on the primary will tame this to a less stressful level of 40-60 kv peak applied to the fusor.

Richard Hull

[Richard Hester]

At those elevated voltage levels it may be time to consider some X-ray shielding, as the SS shell of the fusor will start to become transparent to x-rays at the input voltages you're considering. A properly placed "shadow shield" might be sufficient. If you can manage it, distance from the X-ray source also helps.

[Bob Reite]

I have found that at voltages around 25 KV to 30 KV X-rays will start to penetrate the stainless steel of a typical vacuum chamber. On my original fusor I added lead shielding. Dougs fusor is operated by remote control from 100 feet away.

[Emma Black]

Depending on which meter I use, I can detect the xrays at between 30-34kv through the shell.

The data cabinet I'm using to contain everything has 4 layers of lead roof sheet installed in the door and single sheets in the side. This gives the cone of protection and even at full voltage I only get background at the workbench about 6 feet away. I suspect that even at the future higher voltages this is going to be just fine.

Fusor has now been moved so that the viewport points away from the door, obvious now of course but unintuitive when first putting everything together.

I have also upgraded the viewport to a proper leaded glass affair (expensive but nice) and this cuts the soft xrays right down, otherwise they scatter around inside the cabinet quite dramatically. At one point before the viewport upgrade, the cheap webcam I am using would display a lot of snow and occasionality glitch out.

At full voltage and with the door shut, a meter placed directly against the old viewport got a ridiculous reading of 2.1 cGy/h or 20 mSv/h. This was using an army ion chamber type meter.

[Richard Hull]

In all cases with folks working their first fusors in the 25-35kv range, X-rays will be the big radiation issue. Neutrons will never be an issue until you are in the 10e7 range which means 50-60kv plus. Just as you start to worry about neutrons, the x-rays will become ever more fearsome and dangerous beyond belief.

Tackle x-ray issues first. If you up your game with voltages over 50kv. X-rays remain a primary worry at a new level of danger. Fortunately, more lead solves all such issues

Richard Hull

[Richard Hester]

Emma - glad you have enough lead around to sort the x-rays...

[Richard Hull]

Thankfully, it doesn't take much thickness of lead to make a system safe even at 60kv.

Richard Hull

[Richard Hester]

If I ever get around to building a fusor, my first option For powering it up will be a Sedecal X-ray power pack (Federal connectors and all) that's supposed to be good for 120 kVP. Whomever sold it on Ebay wasn't thinking straight, as I got it for ~$200 plus freight. I've seen them going for much, much more than that lately. It's a little larger than the one Mr. Hull has powering his fusor. I probably need to dip into it to turn the rectifiers around and maybe strip out the filament transformer.

[Emma Black]

A quick test with the two PT's in series was successful, 66KV without any arcing issues. This was with no load and before rectification.

For this time around I would prefer a dry system, rather than an oil tank so I have been experimenting with casting some diodes in resin.

First I made up some diode strings taking care to use lots of solder to smooth out all the joins, there is a string of 8 12kv diodes here. Then 3D printed a hollow "bushing" shape to save some resin. This then has the diodes installed inside and is filled with resin, once set its very solid.

The plan is to then install 4 of these as the bridge, with large spheres on each end. Everything with this supply is going to be spaced out much better and without all the rough edges I had before. Of course wont be any options for repairing these unlike with oil so I will make a couple of spares.

[Richard Hull]

66kv is right on the edge of dry work in HV as corona is a big issue creating all manner of ionized air current paths in and around conductors. All the best of luck in this effort.

Richard Hull

[Emma Black]

All things have been busy here with work, so there has been little time until now to make some progress on the upgrade.

This ended up being a bit of a battle, with the first attempt at a fully dry system resulting in quite considerable corona noise and current leakage.

So I increased all the sizes of the spheres to smooth things out, which helped. Then I swapped out the "bushing" on top for a new one filled with oil. All the 3D printed parts were printed hollow and then filled with resin. Finally I just spaced out the components even further, this supply is now quite massive.

I've ran it under load up to 60kv without major issue but over this the feedthrough on the fusor starts to arc, but thats a problem for another day. There are still some tweaks to make but overall quite pleased with how things have turned out. The extra voltage and current seem to be doing wonders for my fusion numbers.

Before wiring: