I need help with a transformer design / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

I need help with a transformer design

1 2 next

Move Thread

LAN_403

GrantX

Mon Dec 05 2011, 04:13AM

Registered Member #4074 Joined: Mon Aug 29 2011, 06:58AM
Location: Australia
Posts: 335

Hey everyone.

I've finally gotten sick of MOT's, and wish to build my own mains-frequency HV transformer. I have a few concerns though, mostly about the core material. As I want it to be a relatively high power (10kVA+) transformer, I'm going to need a fair bit of core material. I can get my hands on a literal ton of mild steel, but would it have suitable properties for a transformer?

Below is my current design, and I already know my core will probably have to be at least twice the size.

Thomas W

Mon Dec 05 2011, 12:04PM

Registered Member #3324 Joined: Sun Oct 17 2010, 06:57PM
Location:
Posts: 1276

hmm, the best place to get a transformer core is buying a site transformer (one of the HUGE ones such as:

) on ebay or off a contact :)

GrantX

Tue Dec 06 2011, 12:41AM

Registered Member #4074 Joined: Mon Aug 29 2011, 06:58AM
Location: Australia
Posts: 335

Tom Williamson wrote ...

hmm, the best place to get a transformer core is buying a site transformer (one of the HUGE ones such as:

) on ebay or off a contact :)

I tried getting my hands on an absolutely huge arc welding transformer, but my uni decided to keep it at the last minute (the bastards...), and I haven't had much luck on ebay, since most of the cores are overseas and thus cost hundreds to post :(

So I'm guessing mild steel has a horrific hysteresis curve for transformers?
Because at the moment the only realistic option (financially) is mild steel :(

Would it help if I simply made the core many times larger than in the design, say about 40kg of mild steel?

Ash Small

Tue Dec 06 2011, 09:18AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

You may find this link useful:

GrantX

Tue Dec 06 2011, 10:03AM

Registered Member #4074 Joined: Mon Aug 29 2011, 06:58AM
Location: Australia
Posts: 335

Ash Small wrote ...

You may find this link useful:

Thanks, I think that helped:

Ok. From what I understand, mild steel is quite lossy, so the core will have to be much, much larger, also to achieve the desired voltage output I'll have to add quite a few extra turns to both secondary coils to compensate for the loss in the core, correct?

This is going to be the second transformer I've ever built, the first being a tiny fluro tube igniter wound around a ferrite stick in one of our first basic electromagnetic classes, so I'd appreciate as much feedback as possible. Especially since this will probably be a mildly expensive project, haha :P

Steve Conner

Tue Dec 06 2011, 03:24PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

OK, for a start 3mm is much too thick for the "laminations". They are usually about 0.5mm. And the core has to be closed, a metal "stick" won't do.

I've seen a terrifying picture of an African made DIY welding transformer, where the core appeared to be made from cut up beer cans, or paint cans or some other really thin sheet steel, mashed into a pile of flat sheets. Nevertheless it seemed to work. Surely you can do better with the resources of a developed country at your disposal

Dr. Slack

Tue Dec 06 2011, 07:38PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

To be a bit more precise about what a metal stick will and won't do for you

With a metal stick, you need about 50% of the ampere turns you would have needed to get a given magnetising flux through your coils had there been no magnetic core at all.

With a closed core, you need about 0.05% of the ampere turns, if you've done a good job of reducing the unavoidable air-gap to less than a gnat's. That's why most people go for closed cores.

GrantX

Wed Dec 07 2011, 05:42AM

Registered Member #4074 Joined: Mon Aug 29 2011, 06:58AM
Location: Australia
Posts: 335

Ah, thank you. I had seen a transformer on youtube built around a single 'I' core, which seemed capable of high power output, but the entire video was in German, so I couldn't understand the specifics.

Closed core it is, then. Would an EI core work with the primary wound on the middle leg, and the secondaries on the outer legs?

How many kg of core material would you suggest for 15kVA at 50Hz?

Can the laminations be pressed together directly, or do I need material between them?

Thank you for your help so far.

Patrick

Wed Dec 07 2011, 07:31AM

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

XravenorX wrote ...

Closed core it is, then. Would an EI core work with the primary wound on the middle leg, and the secondaries on the outer legs?

Can the laminations be pressed together directly, or do I need material between them?

Half the cross sectional area of the two limbs will have half the flux as is in the center limb. so the number of mag line changes per second are fewer, so youll have half the volts per turn for this idea. i think thats not good.

the laminations are usually meant to be high resistance from sheet to sheet with laquer coating, or else the core becomes a one turn short or whatever, the laminations also prevent the eddy loss problem.

lo freq AC is lame for this very reason, you need 40,000 lbs of steel, 20,000 lbs of copper, and 400 gallons of oil to per MW, or whatever. instead of lets say 800lbs of ferrite at hi freq.

Dr. Slack

Wed Dec 07 2011, 08:13AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

XravenorX wrote ...

Closed core it is, then. Would an EI core work with the primary wound on the middle leg, and the secondaries on the outer legs?

Yes, it would work in principle, but there is one important consideration

The two half-windings on the outer legs *must* be balanced in flux and hence voltage, which means they must be connected in parallel. This stops you using it in what I presume was your preferred centre tapped scondary output connection.

1 2 next

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