BIG inductors - DIY style?

[max_torque]

Say i wanted a 10mH inductor, that can carry 100A DC without saturation, is there a "DIY" way of making one of these?

It's obviously going to be pretty big, an aircore would mean excellent saturation characteristics, but massive size and a large series resistance, because it's need a lot of turns and hence a long length of wire.  But if not an aircore, what could provide a suitable core for this inductor?  I'm not really that worried about its loses but i'd like to be able to just air cool it etc


 Anyone made such a thing?   

[Marco]

100kg worth of laminated E cores?

[johansen]

so that's 50 joules worth of energy stored.


ok, so, assuming losseless cut E cores.. a stack of core such that the center leg is 80 mm by 80mm.. the core will weigh about 23.5 kilograms.
filling it  70% full of copper will require  13 kilograms of copper.

a physical air gap of about 5mm will result in 10mH for 110 turns of wire.
at 98 amps, you will lose 285 watts in the copper at a flux density of 1.5T

an average 25 kw pole transformer could be used if you can gap the core. no need to remove the primary coil either.. the secondary should be able to handle 100 amps continuously across the 240v winding. but you may not have enough turns.. either way its a good starting point. once you get the core out then disassembling it should be easy.. rewinding a new coil isn't...

[metacollin]

Sorry, couldn't resist .  Original idea from Drew at tooth paste for dinner.  But with pan flutes. 

[daqq]

Buy several reels of SMD inductors, use an unholy amount in a series/parallel combination, get monstrosity.

[metacollin]

Seriously though, I have to ask, what the hell are you doing? That's going to be one big ass choke.  Are you trying to become a super villain or something?  I'm pretty sure 50J of magnetic field is one of the ways people turn into super villains. 

Anyway, people do make stupidly large ballasts and stuff that are this large.  You can do it with a toroidal core.  Do not use E core laminate construction type cores, you'll have so much eddy current loss you'll not even care about saturation, because power dissipation will be the main factor, and it will demand much larger core. 

Hunt down a 45-50A variac toroidal core.  Actually, its better to go by weight.  You want a variac that weighs 80-85 pounds.  The core is what needs to be intact, if the wiring or other crap is broken, all the better, hopefully it will be cheap in that case .  The toroid construction will cut your eddy losses in half, and its nice and compact for lugging.  Be careful though, 85 pound toroids are the devils from which permanent back injury and hernias are born.

You'll need to air gap it, which is not easy, its a ribbon of silicon steel wound up like a watch spring in the core.  But it can be done, the http://4hv.org guys do it, I would ask them for guidance.  Make sure whatever gap you make is filled with something strong - that core is going to be crushing inwards like only 80 pounds of 1.4 tesla electromagnet can. 


Or you can just follow my flow chart.  That works well too .

[mikerj]

Assuming this is for low frequencies, you need to find a honkin great big laminated E core to get your inductor down to a reasonable number of turns.  Decent welders usually have large inductors on the output to stabilise the arc current, and very large transformers, so a broken/discarded welder might be a good start.  You might be able to use microwave transformers as well, though I suspect you'd need to string some in series to get enough core material for 10mH at 100 Amps.

[peter.mitchell]

get a bit of copper pipe and start sliding toroids over it until you reach your desired inductance.

[johansen]

Assuming this is for low frequencies, you need to find a honkin great big laminated E core to get your inductor down to a reasonable number of turns.  Decent welders usually have large inductors on the output to stabilise the arc current, and very large transformers, so a broken/discarded welder might be a good start.  You might be able to use microwave transformers as well, though I suspect you'd need to string some in series to get enough core material for 10mH at 100 Amps.

electrical machines get more efficient as they get bigger.

the inductor i mentioned in my post would weigh 37 kilograms.
to use microwave oven transformers, he would need about 37 of them, and it would weigh a lot more than 37 kilograms.
to use microwave oven transformers somewhat efficiency, he would need to pack the cores in a line.. which would end up being about 2.5 meters long.--this is actually a plausible option because he could use 1 gauge aluminum wire to carry the current. you should be able to get 4 turns through all the cores.
gapping them all at about half a millimeter.

[calexanian]

What frequency are we talking? Actually what you are asking for is not particularly complicated, but may be very expensive.

[metacollin]

Wow, lots of bad advice in this thread.

Assuming this is for low frequencies, you need to find a honkin great big laminated E core to get your inductor down to a reasonable number of turns.  Decent welders usually have large inductors on the output to stabilise the arc current, and very large transformers, so a broken/discarded welder might be a good start.  You might be able to use microwave transformers as well, though I suspect you'd need to string some in series to get enough core material for 10mH at 100 Amps.

No.  Laminated E core is the worst choice as core geometry goes, and is especially bad for a high current inductor.  This isn't iron powder or ferrite, its not some homogenous  magnetic material. Silicon steel is not magnetically isomorphic.  It gets much of its flux density from the grain of the metal being oriented with the field. Due to the shape of the core being rectangular, at most 40% of the core material has its grain oriented usefully with respect to the current.  This is why the flux density is so low.  The amp turns for a required inductance will be the poorest of any steel core because of this.   But yes, he would need a honkin great big one, that's for sure.  It would be twice the size and have much higher eddy losses than if he were to use the appropriate core for an inductor.  Which is a ribbon-wound toroidal core, the kind used in all variacs.  100% of the grain is oriented correctly, and they can handle 2T and per kilogram, will have twice the amp turns for the same inductance as E cores.  Oh, and have much lower eddy losses (again, due to the shape).  And its easy to find a core that large, you can even buy a commercially available variac with a suitably large core for under $1000.

Decent welders have large split-core saturable core reactors.  If you think you're going to get a core that can take a ton of magnetizing current with out saturating, you're in for an unpleasant surprise.  A saturable core reactor is basically a leaky common-mode choke, and sometimes it has a 3rd winding that is used to apply various amounts of DC bias to the core, letting you vary the leakage inductance by putting the core more or less into saturation.  Common mode chokes can handle high common mode currents because the AC flux cancels out, and 200 amps could be going through its windings but as far as the core is concerned, there is no current at all.  It has some leakage inductance on purpose, often quite a bit, unlike a normal common mode choke.  This acts as a series inductor with the load, and causes phase shift.  It limits the current by giving the welder output terrible power factor, basically.  And the cores are split, there is a seam splitting it in half.  This matters because, again, the laminates are not magnetically isomorphic.  The split core is split because the grain is oriented in opposite directions for each half, so the amount of magnetizing current such a core is capable of is significantly worse than a regular E core, and an E core is not the right core to begin with. 


And someone mentioned 2.5 meters of MOT cores or something.  Are you kidding? Just use a toroid core, its ready made and a practical size =P.


http://www.surplussales.com/Variacs/Variacs-3.html Not in the UK, but most of those would probably work.  I think you could get away with the 28 amp one, but you might not.  A bit bigger would be best. 

[PeterFW]

No.  Laminated E core is the worst choice as core geometry goes, and is especially bad for a high current inductor.

My knowledge of this subject is fairly limited, i just ask because i am curoious and most definately because i want to nag
Most high power inductors i have seen were transformer coils, that is laminated e-core if i am right?
But those were only low frequency audio inductors from passive filters in huge subwoofers, from the time before active filters were cheaper.

If they are the worst choice, why were they used?
Size or because it is cheaper?

[johansen]

if you can afford to spend 10 dollars a kilogram on some nice 1.9T metal.. go ahead, it would certainly be better than 36 microwave oven cores.

you'll still need a 20 kilogram core to store 50 joules at reasonable losses.. but it really depends on what "reasonable" losses means.

a 30 amp variac core is about one half to one third the size you need.

no, you can't just unwind it and then coil it up with say, 2mm thick paper for a gap. --this insulation material would have to be a tapered spiral, maximum at the inside diameter (shorter magnetic path) and thinner at the outside diameter (longer iron path means less air gap for the same permeability)
you'll end up with a core that's about 90% air and be flat like a washer, and very large.. like 4?! times the diameter you started with at minimum.

you will have to cut the core on a bandsaw to put a gap in it.


It does not matter if you use cheap E core material for a choke because again, electrical efficiency increases with the physical size of the object.
I have some E cores from a welder, 26 kilgrams worth and the core cross section is about 2 by 4 inches.
filled with copper, the copper would weigh more than the iron, and because of this, it would be practical to run the cheap E core metal at 1.9T at "reasonable losses"

Toroidal transformers are intrinsically different and usually you can't get more than 50% fill factor, where as you can get 90% with E cores.

For large inductors the cheap metal doesn't matter. By gapping the core you're reducing the permeability from 2000 to 200 or less than 100.
of course we don't like hysteresis and eddy current loss but that is situation dependant.
Microwave oven transformers typically run the core at 1.7T off the line anyways.. and the magnetizing current to get them that high is about 500 amp turns for the ungapped core*.. which is about half what they would be run at to store 1 to 2 joules per mot core. 1200 amp turns with an appropriate gap is about the maximum for "reasonable losses"
so yeah, 36 microwave oven transformers configured as back to back like this [ ] discarding the "I" sections is actually totally appropriate, but still sucks
I have 4x mot cores in my welder with 2 pieces of paper for a physical air gap (more on one side than the other to get a swing choke effect) wound with 9 turns of 1 gauge aluminum i think. it works well. and very cheap.

also, metacollin go back and recalculate the total volume of metal in an E core that is used in the suboptimal dimension... you'll find its less than 25%.
the E cores are stamped in semi grain aligned metal, and the "I" section is cut from the slot where the windings go.. so most of the "I section" is in fact used in its optimal direction.


BTW, to illustrate the point that as you increase the size, things get more interesting, the power grid folks use solenoidal inductors without an air gap.. hundreds of kilograms of copper wound around a cylinder of metal. but here's the key: the cylinder of metal has a toroidal winding on it that is used to saturate the core without applying but maybe 1000 amp turns per meter (no air gap) and they can vary the inductance by saturating the core with relatively little power consumption.
but the main inductor is wound on the outside of the core and the "air gap" is on the order of 10 feet. --they use these things for tunable harmonic traps.

^this is totally unpractical on a 10 kilogram core and inductor for example, but when you get larger systems electrically they become more efficient.
and the "q" of large inductors is often so high that you need de Qing resistors anyways, so you might as well throw cheap metal at it.

*might be wrong on this, most of them pull 4-6 amps from the line at 60hz but not all of that is magnetizing amps. anyways, their effective utilization is around 2 joules per core.

[mzzj]

Wow, lots of bad advice in this thread.

Including yours 
Toroid core is nice for transformers but nonsense in big-ass high current inductors.  Never seen big toroidal inductor in any actual use.
4hv teslacoil-hobbyist probably use whatever they can get their hands on 

[johansen]

Wow, lots of bad advice in this thread.

Including yours 
Toroid core is nice for transformers but nonsense in big-ass high current inductors.  Never seen big toroidal inductor in any actual use.
4hv teslacoil-hobbyist probably use whatever they can get their hands on 

mostly due to the cost of winding it.

as far as i know, the most efficient topology for bulk energy storage where both ferrite and powdered metal cores sort of overlap.. is the pot core. very high fill factor for the copper bobbin.. and optimal use of the magnetic materials.. form fitted to the spread of the magnetic flux. the copper bobin can have its optimal 1:1 cross section and the inside diameter can be optimized.

now try and make a pot core from amorphous iron ribbon lol...aint going to happen.. so we're stuck with tape wound E cores.
ahh! but the tape wound E core can be made with a circular core cross section to reduce the copper losses! but now we're adding labor again...

btw, large transformers have E cores built up from laser cut, then ground and polished grain alligned steel sheets.. built up in hundreds of unique sizes to get a circular core cross section.. and they operate at over a kilowatt per kilogram power density at about 1 watt per kilogram total losses.

And so again, "reasonable losses" is totally arbitrary.
Wasting 400 watts continuously to store 50 joules of energy might be ok for a welding choke.. but unacceptable elsewhere.

[mzzj]

now try and make a pot core from amorphous iron ribbon lol...aint going to happen.. so we're stuck with tape wound E cores.

Without need for air gap you could go bit creative and make sort of inverted toroidal 
Copper winding in a shape of toroid core and core material processed to wire and wound around the toroid just like copper in usual transformer. 

Too much work to calculate if there would be any benefit but for sure it would be good for eeeexpensive high-end audio amplifier just because being exotic. And expensive.

[johansen]

now try and make a pot core from amorphous iron ribbon lol...aint going to happen.. so we're stuck with tape wound E cores.

Without need for air gap you could go bit creative and make sort of inverted toroidal 
Copper winding in a shape of toroid core and core material processed to wire and wound around the toroid just like copper in usual transformer. 

i tried using a 2 kilogram spool of bailing wire as the core for a toroid, even though the wire was pretty well oxidized, it was still effectively a solid blob of metal.
i didn't bother trying to separate out the differences between hysteresis loss and eddy current loss because the losses were so high i'd have to compensate the copper losses needed to make the measurement.

i have kicked around the thought to use pallet strapping steel tape as a transformer core. its 1$ a pound and often you can get it for free.
but it would have to be annealed and i don't have a big enough furnace to do an entire 200 pound roll of the stuff at once... yet.

[max_torque]

The application is an output filter for a large(ish) battery emulator, that uses an IGBT bridge at between 5 and 15KHz fundamental PWM frequency, to "chop up" a DC link of between 40 & 400Vdc.

Currently i use a 3 phase thyristor drive, firing into a small inductor and there is quite a bit of capacitance on the output to get a low voltage ripple, but the limitation of the maximum control frequency due to the 50Hz input, and the capacitance makes dynamic output voltage control difficult.


Moving to a high frequency modulation of the applied dc link voltage with a larger inductor but much less capacitance should allow much more dynamic control of the output voltage

[jaxbird]

Say i wanted a 10mH inductor, that can carry 100A DC without saturation, is there a "DIY" way of making one of these?

It's obviously going to be pretty big, an aircore would mean excellent saturation characteristics, but massive size and a large series resistance, because it's need a lot of turns and hence a long length of wire.  But if not an aircore, what could provide a suitable core for this inductor?  I'm not really that worried about its loses but i'd like to be able to just air cool it etc


 Anyone made such a thing?   

This is a cored 8.2mH (15awg wire) that will probably carry 100A for a short duration:



To220 part for size:


Series resistance is about 330mR

[technix]

The application is an output filter for a large(ish) battery emulator, that uses an IGBT bridge at between 5 and 15KHz fundamental PWM frequency, to "chop up" a DC link of between 40 & 400Vdc.

Currently i use a 3 phase thyristor drive, firing into a small inductor and there is quite a bit of capacitance on the output to get a low voltage ripple, but the limitation of the maximum control frequency due to the 50Hz input, and the capacitance makes dynamic output voltage control difficult.


Moving to a high frequency modulation of the applied dc link voltage with a larger inductor but much less capacitance should allow much more dynamic control of the output voltage

Can you increase the frequency and/or increase the amount of phases? My dual Xeon motherboard have a buck converter capable of 200A using 3MHz switching frequency and 24 phases, and that will not ask for those monstrous inductors (and you may be able to get sway with using a pair of TO-220 MOSFETs per phase instead of a pair of huge IGBTs)

[T3sl4co1l]

The application is an output filter for a large(ish) battery emulator, that uses an IGBT bridge at between 5 and 15KHz fundamental PWM frequency, to "chop up" a DC link of between 40 & 400Vdc.

400V at 5kHz and 20% ripple should be under 500uH.  How the hell do you get 100mH?!

Currently i use a 3 phase thyristor drive, firing into a small inductor and there is quite a bit of capacitance on the output to get a low voltage ripple, but the limitation of the maximum control frequency due to the 50Hz input, and the capacitance makes dynamic output voltage control difficult.


Moving to a high frequency modulation of the applied dc link voltage with a larger inductor but much less capacitance should allow much more dynamic control of the output voltage

Just what are you trying to do here?  None of these numbers sound the least bit sane...

Can you increase the frequency and/or increase the amount of phases? My dual Xeon motherboard have a buck converter capable of 200A using 3MHz switching frequency and 24 phases, and that will not ask for those monstrous inductors (and you may be able to get sway with using a pair of TO-220 MOSFETs per phase instead of a pair of huge IGBTs)

Uh, well, your motherboard does it at less than 10% of the voltage, too...

Equating a special purpose design at 12V and 3MHz, with a monster at 400V, 100A, and any frequency, is some pretty serious cognitive dissonance there...

You're talking moped to Lamborghini.  Well, not quite that many orders of magnitude.  More like peppy RC car versus truck hauling ten tons uphill.

Tim

[eneuro]

Seriously though, I have to ask, what the hell are you doing? That's going to be one big ass choke.

I have no idea for the moment what inductance has this MOT primary, however one need to remove this crappy coppered aluminium <1mm in diameter wires to be able let decent current flow througth it since oryginal winding had 2.3 R and is rated for  230VAC 50Hz


Anyway, at higher frequency this MOT could be probably quite good for cooking if oryginal windings changed with a few turns of copper pipes  in  high frequency induction oven

May the Force be with us 

[T3sl4co1l]

With a gapped core, you'll get whatever inductance you get (up to a maximum "ungapped" value in the mH), saturation current inversely proportional to inductance, and saturation flux proportional to the core area and turns (vaguely guessing around 0.14 Vs for a JP/US (~120V) model, double for EU (240V)).

Obviously, without the secondary and shunts, you have the area for about 2-3 primaries, so if you happen upon multiple identical transformers, you can put more together and fill up the winding window to get more current * flux capacity.

Tim