Brute forcing high voltages with two MOTs (not what you think)

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Post Reply
David Vermilion
Posts: 7
Joined: Mon Sep 16, 2019 4:07 am
Real name: David Vermilion

Brute forcing high voltages with two MOTs (not what you think)

Post by David Vermilion » Tue Sep 17, 2019 12:33 am

Okay so I am David Vermilion, I've been lurking around here for about two months (had lurked around a few years ago too) and finally decided to build a fusor. Well the biggest challenge is obviously the power supply. So I read many a forum post on the subject. Including the recent ones about precipitator PSUs and their unfortunate failures. Well I gave it a thought and concluded to use two identical MOTs to build a ~30Kv transformer with 0.03-0.04mA use a variac for control, a voltage divider to measure the voltage and string some cheap 20Kv diodes in series to get a bridge rectifier.

Now I know exactly what experienced users are thinking. "This again? You cant get anywhere near 30Kv with MOTs in series and even if you did you'd break down the super thin insulation and would need about 15 of them sunk in mineral oil with a coolant loop to dissipate all the heat.

This is where an idea hit me.

Why not get two identical MOTs (serial numbers, wattage etc) and cut them apart like some people do to make current transformers for homemade welders?
Except I will join both E pieces end to end and use the very large available space to wind a secondary from 32AWG wire! (In hind site I should have opted for #34AWG) The two E cores would allow me to have much more room to put a custom wound secondary coil. I'm not a big fan of complicated and noisy switch mode supplies/flyback transformers (K.I.S.S) and as many many users have pointed out NSTs wont cut it due to the "low" voltage and minuscule operating current.

So that's what I've set out to do.

Now time to give credit where it is due.
I was googling the subject for information and happened upon a webpage that had nearly everything I needed to know and more!
https://ludens.cl/Electron/trafos/trafos.html This webpage is extremely useful. This guy provides a lot of crucial information for rewinding, repairing, or building from scratch transformers. This guy is amazing.

A problem he never mentioned but that I read about in an ancient textbook (Elements of Electronics 3rd edition, Hickey and Villines, 1970) is that there are magnetic circuits just like electric ones (magneto-motive force just like electromotive force) and that there is an magnetic analog to resistance called reluctance. Meaning that the path of the magnetic circuit has a certain amount of reluctance (resistance to having a magnetic field pass through it) and that doubling the length of that path doubles the reluctance even if an object has high permeability. Now the magnetic flux induced in a transformer is measured in Amp-turns or, if its an old book, Oersted as opposed to gauss or tesla (current I multiplied by the # of turns of the primary or I*T)

So I would be forced to use both primaries in parallel because the primaries are #14AWG only coated in magnet wire and can only handle 6 amps in order to drive my transformer, which would eat up the valuable space. Note, a series connection would give me more volt-turns and would saturate the core wasting lots of power as heat, when I need amp-turns. So that's a problem...I dont have as much space as I though I would. Thus why I wish I had ordered #34 wire instead of #32 (I have 25,000 feet of it).

So I did the voltage math.
The primaries have 103.5 turns each (they are driven in parallel for more amp turns than volt turns) and when I want 30Kv I will run it at 130v (my cheap variac can do 130). So 30,000v divided by 130 is 230.769, this is our primary to secondary turn ratio. So 103.5x230.769=23,884.59 turns.
I haven't built it quiet yet. I'm working on trying to get the bobbin for it correct because the dimensions are very tight (again wishing I had ordered #34 wire). I wish I could share photos but the attachments tab says they are too large. Which is disappointing because I took photos for this very post that I wanted to make over a month ago but the site was going through issues/revisions so I waited.

For anyone concerned with my safety, I DO OF COURSE plan on using NMN laminate paper for the insulation between the core and the secondary, use layers of kapton tape to separate each layer of windings, and submerge in mineral oil while vacuuming out the air bubbles under heat/vibration. Also plan on soldering large So dont worry I've got insulation covered and if heat becomes a concern I will rig up a timer-relay circuit for the power. I dont see how running fusor like an oven could hurt anything but neutron count over time, so just extend the operating time from 30 minutes to an hour and just let it chooch away while exposing a bubble tube.

Please leave feedback letting me know your thoughts.

-Dave

User avatar
Mark Rowley
Posts: 446
Joined: Sun Dec 29, 2002 5:20 am
Real name: Mark Rowley
Location: Sacramento California
Contact:

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by Mark Rowley » Tue Sep 17, 2019 5:25 am

"...Including the recent ones about precipitator PSUs and their unfortunate failures."

Dave, that is old info and the precipitator supplies do in fact work. Yes, there were some failures in the beginning but at least in my case they're resolved. Just had a very successful 45 min run with mine about 90 minutes ago.

Unlike a classic transformer, the precipitator supplies introduce voltage and amperage a little differently and extreme care needs to be taken to avoid a flashover arc. Definitely an operational learning curve but nothing insurmountable. And it's important to note they are probably best suited for a 2.75" system. The higher current demands of a 6" sphere may be too much for one, although it'd be nice to see a dedicated attempt.

No one else here has published any attempts at fusing with a precipitator supply.....yet (as far as I know). Just a matter of time before we see more.

Mark Rowley

User avatar
Rich Feldman
Posts: 1142
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by Rich Feldman » Tue Sep 17, 2019 5:17 pm

Welcome, David.

Applause for learning about coil-and-core transformer making.
MOT's are, indeed, a good source of free core parts and primary windings. As you know, putting two E core blocks in series won't increase the 60 Hz volts per turn above the MOT design point, at the knee of saturation. You just get a bigger winding window, and the need for more no-load ampere-turns in primary to magnetize the longer flux path. Note than in original application the MOT primary conducts at least 12 amperes. Combination of magnetizing current and N2/N1 times the secondary current, at very different phase angles WRT the voltage, so simply adding the current magnitudes will overestimate the total.

To wind almost 25,000 turns of fine magnet wire, with 30 kV between start and finish, is a pretty high hurdle for a beginner.
I would think that 20 kV RMS (28 kV peak, with FW bridge rectification) is enough for fusion. You'd get close to that with a NST, over-volted to the same magnetic flux density as a MOT. ( Uncooled MOT's overheat in an hour while NST's run for decades. )

As a sanity check, before any fine wire winding, I would assemble the cores and primary windings. Then measure the magnetizing current as you ramp the primary voltage from 0 to 130. Then install a test secondary winding of 10 or 20 turns, physically distributed to match the intended HV secondary, and see what volts per turn you actually get in that zone of the core window. (The flux coupling ratio could be significantly less than 100%.)

Good luck, and let us know how it works. Glad to see you consulting other transformer-oriented sites.
Last edited by Rich Feldman on Wed Sep 18, 2019 2:49 am, edited 1 time in total.
Mike echo oscar whisky! I repeat! Mike echo oscar whisky, how do you copy? Over.

User avatar
Rich Feldman
Posts: 1142
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by Rich Feldman » Wed Sep 18, 2019 2:39 am

p.s. We want to see your pictures.
The easiest way to suit the 500 kB limit: re-take the pictures, after setting your camera pixels count to something like that of an Iphone _display_ or laptop PC _display_, which is plenty for any forum that isn't about photography.
If you're happy with existing pictures that just have too many pixels, you can crop and scale them for free using things like Windows Snipping Tool, or online photo-shoppy editors like Pixlr.
Mike echo oscar whisky! I repeat! Mike echo oscar whisky, how do you copy? Over.

David Vermilion
Posts: 7
Joined: Mon Sep 16, 2019 4:07 am
Real name: David Vermilion

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by David Vermilion » Wed Sep 18, 2019 11:47 am

Thank you Mr.Feldman! I'll try that!

Rex Allers
Posts: 387
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by Rex Allers » Wed Sep 18, 2019 1:07 pm

Big pics...

Rich's suggestion may work depending on the settings available on the phone or camera. I think I've seen cases where the smallest setting was still pretty big.

You can also resize an image in one of the photo apps.
If you have a Windows PC, Irfanview is a (great) viewer which has some editing abilities. Resize/resample is the feature needed. Crop might also help before a resize.
https://www.irfanview.com/

I also tried on my Android phone and succeeded (different app).

I know there are photo editing apps on Linux but I've never done that.

I don't know the Apple world but there are no doubt apps there to do resize.

Let us know if you need more details. I can describe the steps I used in Irfanview if you need that.
Rex Allers

David Vermilion
Posts: 7
Joined: Mon Sep 16, 2019 4:07 am
Real name: David Vermilion

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by David Vermilion » Thu Oct 03, 2019 8:32 pm

So here are the photos. My apologies for it taking so long. Life happens you know. I had to redirect my attention to some E bay listing for important parts I needed for the fusor because if I didn't nab them up quick someone else would. I also took time off of this project so I wouldn't get burnt out. As for winding the coil I've only gotten ~250 turns on it so far and will be working on it for a while. The winder I built goes too fast and I will have to find another pulley to reduce its speed to something reasonable. It's a sewing machine motor and the shaft is 1/2" so finding a 6" pulley within my budget will take some time. Something I dont have much of.

Something interesting I found was that the secondary coils of these transformers look like they are made of aluminum and not copper. They were a shiny silvery grey and haven't turned red or green as copper does.
Attachments
0903191842b_HDR[1].jpg
aluminum windings question mark.jpg
0905191943_HDR.jpg
double E core.jpg

User avatar
Rich Feldman
Posts: 1142
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Brute forcing high voltages with two MOTs (not what you think)

Post by Rich Feldman » Thu Oct 03, 2019 10:41 pm

Thanks for sharing the pictures.
I've seen MOT's with aluminum primary coils, but not secondaries before now.

Re. your last picture with the stacked E cores:
Have you measured the magnetizing current (vs. input voltage) with both primary coils in parallel, or with a single coil ? Does it match your expectation?

If you use two primary windings, it might make a significant difference whether they are at opposite ends or right next to each other, especially when you measure the induced voltage in a test secondary winding. That's because of flux leakage, which isn't accounted for in basic 1-dimensional core reluctance model. It might be less than 5% with your gapless core geometry.

If that setup were on my bench, I would start with a test winding of 10 or 20 turns that's very compact, say 1 or 2 mm in both thickness and length.
Measure its voltage as you move it to different places in the core window.
While core is magnetized to your intended B_max level, at which the permeability may have rolled off to much less than its "normal" value.
Mike echo oscar whisky! I repeat! Mike echo oscar whisky, how do you copy? Over.

Post Reply