What kind of jacks/terminals would be best for a bobbin? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

What kind of jacks/terminals would be best for a bobbin?

previous 1 2 3 next

Move Thread

LAN_403

LutzH

Wed Jan 20 2010, 08:27PM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello:

Yes on the PDF design this is true, the rings are split with an angle cut, as one can see.

The one that I was questioning was not the transformer on the PDF, rather it was one that I have from a CT scanner HVPS. It has a large ferrite core, and 2 sets of windings. The final HV output wire on each secondary coil is soldered to a 2cm wide copper foil band which wraps around the secondary to form a full turn, and it looks to be connected at the ends as far as I can tell. It is hard to tell because they coils are then wrapped with insulating tape, and a small square is cut out to expose about a square cm to make the HV secondary connection. The same transformer also has a copper foil layer between the windings.

If I remember from a while back Harry, or someone else, explained this one time, something about it having a shielding effect, to reduce the interwiding capacitance. I wish I could remember exactly, but this person explained why it is not a problem, even though it appears to be a shorted winding. This transformer has 20 turns of heavy Litz wire on each primary, and about 800-1000 turns of 26-28g wire on each secondary in layers, with insulating tape between them. Each secondary is wound similar to the flyback that Steve Ward built for his CWVM, with each new layer having a few less turns than the previous one. The sad thing is that I have no idea ref. the final output voltage of this transformer. Normaly in HFHV medical applications the inverter runs off the rectified line voltage. So the input should be 1.4 x 220-240 V (RMS to Peak Rectified). I think 440 would be a bit much, so I have ruled it out, and 110 is unlikely due to the heavy current draw required. So in a nutshell I think that it was designed to put out 15KV, or 30KV, if the secondaries are connected in series. The HF which was then fed to a VM circuit to get the full 120-150KV. I did manage to find out that the intended frequency which was 40KHz.

The transformer layout looks similar to the one pictured by jpsmith123, except that it has no bobbins, and they used insulating tape, with dabs of epoxy to keep things in place. The use of the grounded copper foil shield, and the copper foil HV output ring makes it interesting. I will try to dig it out to post a picture in the near future. Maybe for voltages under 20-30KV the bobbin is not needed, and good quality insulating tape is enough? As far as enclosures go Patrick nailed it, the ideal fusion of functionality and art. It has me thinking about some variation like a section of SS ISO 160 vacuum pipe with end flanges, the router with a pattern cutting bit could make short work of the end plates, maybe with some epoxy glass green board, or even neon acrylic :)

Sorry for the confusion, I maybe should have been more specific in my posting, thanks for helping clear this up. Maybe the same person who explained the copper foil shield concept before, could give us a refresher :)

Patrick

Wed Jan 20 2010, 10:07PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

A single loop around the center coreleg would be a dead-short, a single loop around THE whole transformer all (three legs in a E-E shape or both in a U-U shape) would be fine because ther flux wil not coulpe much out side the window (remmeber flux terminates ONLY ON current!

i will post pics in a few hours so lutz your idea may be ok.

-Patrick

jpsmith123

Thu Jan 21 2010, 01:23AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Lutz I've already looked into the possibility of using vacuum hardware as a transformer "tank"; so far to no avail. I was trying to find an "ISO-F" full nipple in aluminum, and I didn't find it.

I found some half-nipples in #304 SS, and I found some blank flanges in aluminum, and everything seems very expensive.

My guess is that if I were to find one, it would be at least $200 to $300, which is too much for me.

Maybe a stock pot or something would work, but then a flange would have to be made and welded on...I could see it becoming a costly PITA.

LutzH

Tue Jan 26 2010, 01:08AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello: Not to be the "fly in the ointment" but using vacuum is very hard due to the somewhat porus nature of some ferrites, and a whole bunch of other problems, unless you have it on a fast pump all the time. Maybe I miss-understood, and you were going to just use the vacuum fittings, but filled with oil, or insulating fluid. In this case I would search under "ISO Vacuum", and "NW Vacuum" under bussiness & industrial on ebay. You can also try searching under "Sanitary Fittings" dont ask me why, but I find them this way at times. If you do not want to use oil then I would try the following options:

1. Insulating Gas = SF-6, or even just Nitogen under pressure, you can even use off the shelf insulating fiberglass ANSI flanges, and pipe as a container. Normaly it is rated to 125PSI minimum, and it is available in all sizes shapes and fittings including blank ends for your feed through conductors. Pure N2 is about 115% of the air value, add even only 15% SF-6 and you have 2 times the air value with no pressure required. Do a yahoo search under fiberglass fittings /pipe, it comes from 1" all the way up to industrial 32" size. Add a tire valve with some epoxy, and you are in bussiness with a 80-90PSI N2 fill from your local tire shop.

2. Fluorocarbon insulating fluids, these evaporate without the mess of oil. Look for the following on ebay: Fluorinert, or Galden Fluids. I have some Galden HT 110, this stuff is amazing, it will handle something like 30KV/mm. I paid $50 on ebay for a whole gallon, normaly its about $500 a gallon.

I think the fiberglass pipe option is the most easy, with 60-90PSI N2. If you do not like oil. Good luck

jpsmith123

Tue Jan 26 2010, 06:03AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

I would never try to use vacuum insulation for something like this.

In any case, I originally planned for the xfmr and multiplier to operate in the open air, but the 2CL2FM diodes I was planning to use for the multiplier seem physically rather small for their rated voltage (they're about half the length of some other 20 kv diodes I have laying around here) and I'm therefore somewhat reluctant to use them in an open environment...I can see them flashing over on a humid day or from some accumulated dust/dirt. Moreover, according to the specs, it seems the diodes may dissipate a few watts of power. If running for any length of time at moderate to high power, I can see these small diodes getting rather hot.

So, given the short creep distance and small surface area for cooling purposes, I'm beginning to think it would be better if I were to immerse the multiplier in oil...and if I do that, I might as well put the transformer in the tank as well.

I'd prefer to use an aluminum tank of some kind...but I can't find anything anywhere at an affordable price. Unfortunately though, even if I had a big lathe and could make it myself - just the price of a one foot piece of 8" Al pipe is over $100 from what I'm seeing.

LutzH

Tue Jan 26 2010, 07:13AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello:

You are right about the 2CL2FM diodes, they WILL flash over any where near their rated voltage. If I were doing open air I would go to

, and order some of their 1/2" polyethylene U channel. Then I would order a "sample" of HV potting epoxy from a manufacturer. Use the one with the highest thermal conductivity. Then I would place the diodes into the channel, and fill it with the epoxy to form longer insulated rectifier units.

To keep the diodes centered you can mix 2 batches of epoxy, pour one right away into the U channel mold to a little under 1/2 of the total desired level, let this layer gel just enough so that the diodes do not sink into it. Then lay down the diodes, and top it off with the second layer of epoxy from the newer more liquid batch.

US Plastics also sells those clear plastic tube containers, with flexible vinyl cap ends, page 169 of their 2009 catalog, the 1/2" OD tube is 31 cents per foot, with the end caps at 6 cents each. You can cast the diodes into these tubes using the same epoxy. The cool part is that you can put a pinhole in the end cap and push the rectifier lead wire through it for a positive seal, or even cast in a threaded brass screw etc, to fit a rounded anti-corona acorn nut.

The same company has just about every plastic imaginable in many forms. Including fiberglass, phenolic, and PVC structural forms. For a real cool look why not assemble the entire VM inside of a large diameter clear acrylic tube, or better yet a clear PVC tube (Yes they have clear PVC tube up to 6") If you fill the tube with insulating oil, or gas then no insulating of the diodes is needed.

If I were building a VM with under 500KV output then I would use an air insulated design, with cast HV epoxy potting of select components, like the diode strings, and protective resistors. Then I would finish it off with an output Toroid for 250KV or less, for 250-400 I would go with a SS metal sphere terminal, because the toroid get s real wide at 400KV+.

These VM's can look as cool as they perform especially with round plastic sheets as individual stage supports, with equi-potential metal rings around them. There is one seller on ebay who has a bunch of clear round 12" Polycarbonate disks for sale for cheap. If they were 14" I would have bought them for my VM stages (I need a large center hole for an accelerator tube). Any way you go US plastics will be very handy :)

Another idea I had was to enclose the entire diode string inside of clear plastic (Epoxy glue bondable plastic) tubing sections, which zig-zag up the column via glued in conductive copper elbows which also form each stage connection (A tiny hole in the elbow bend could be soldered from this point, to the diode string inside. For a FW design you could use copper + for the center section, and elbows on the sides. This tube could be inside potted with HV epoxy, a section at a time, or the whole thing could be filled with oil when done, and capped. The only problem is replacement, but normally if you blow diodes, you have to replace all of them anyway, or at least a full section.

I hope this helps....Lutz

PS HV Epoxy is about $30/gal the last time I checked.

jpsmith123

Wed Jan 27 2010, 05:09AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

It seems that the "mechanical" issues of construction, e.g., what to use for a tank, are always the the biggest PITA, doesn't it?

Anyway, unless I can find some kind of an affordably priced, suitable "tank", in the near future, I suppose I'll finish the transformer for operation in open air.

I'm actually planning to make two sets of HV secondary coils for testing - one using magnet wire and potted with something, probably epoxy, and the other one using teflon insulated wire and not potted.

With regard to the terminals, I did a quick electrostatic simulation of a typical spark plug terminal at a potential of 10 kv, a few inches away from ground, and it seems it would be below the corona threshold if the edges were rounded slightly.

Thus I'm thinking about machining some terminals out of brass (part of which would be embedded inside a cpvc sleeve), making the profile the same as a spark plug, and then using standard spark plug wire and associated hardware to to connect the transformer and multiplier (which will be equipped with identical terminals).

I wonder if you can generally buy bulk copper-conductor-type ignition wire, connectors and insulating boots, separately, from local auto parts stores?

Edit:

@Patrick, as far as the cpvc goes, if I assume a dissipation factor (DF) of 0.015, and I operate at 25 khz, using a layer winding with 1500 volts/layer or so, I calculate I'll have less than 200 mw of dielectric loss per bobbin. BTW, I'm not exactly sure what the DF of my cpvc is, as it seems to vary with manufacturer, but I think a value of 0.015 is a good estimate for a conservative design.

Turkey9

Wed Jan 27 2010, 05:54AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

All modern spark plug wire is a kind of resistance wire made of carbon to reduce interference with the car radio. I don't know how easy it would be to buy copper stuff at an auto parts store. The wire is easy enough to take apart so I bet you could make your own cables.

LutzH

Thu Jan 28 2010, 06:51AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello:

NAPA auto parts here sells HV silicon rubber insulated ignition coil wire, the racing folks drive this item, because they care more about a hot spark, than EMI affecting their stereo.

For contacts, and HV anti-corona stuff, there is a company called Liberty Brass:

They are a veritable HV anti corona construction part goldmine. They have brass balls in any size configuration that you want, including ones with threaded holes etc. If you go to the site check out the 1 1/2" cluster body, this thing is like it was "tailor made" for multiple HV connections in the 20-50KV range, man would that item look slick in an air insulated VM! For hollow balls check out:

If you are willing to use oil, go to Walmart, in the housewares dept., they have a stainless steel bucket, with a clear lid which has a cam lock lever seal, complete with a clear silicon seal ring. Your transformer will fit in it, and you could just drill some holes in the plastic lid for your connections, and seal it with silicon. It has a nice brush finish, and it will set you back about 20 bucks, unless you switch the bar code with the smaller one (Just Kidding) Sorry I just do not like Walmart, so I could not resist a little humor. I once bought an expensive Sony camcorder there, and they would not take it back after only 5 weeks! Also they just ran a story on the news about some preacher in England who is getting a lot of heat for suggesting that it is not a sin for a woman to steal food for her child, as long as it is from a corporate owned store like Walmart, and not from a family owned one. My friend used this SS container for his x-ray power supply VM, I think it holds about 3 liters, all in all it makes a very nice tank, and its a pretty good value for the money. You may want to put a nylon zip tie on the lever to prevent unintended "pop opens" an the resulting possible oil spill. For small transformer like flyback size ones, I have used the new fancy sealing Rubbermaid containers from Walmart, but for a "bigun" like this the larger solid SS one is the way to go. Hope this helps...

jpsmith123

Fri Jan 29 2010, 07:05AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Lutz do you know if any of the bulk wire sold by NAPA has a copper core? They refer to some of their wire products as a "metallic" core - actually they say "tinned metallic". They state (and the picture shows) that it's Belden wire (e.g., their part# BEL 734803), but I couldn't find any information about it on the Belden web site.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob