Need to make an 8mH inductor which works under 13V peak

[peter-h]

I used to know all this years ago... how to work out the flux density, and choose a material which doesn't saturate, but I have long forgotten!

It needs to be good with a 13V peak square wave, 500Hz.

So we have 13V across 8mH for 1ms, which from v = L (di/dt) yields 1.625A. A quick hack in LTspice confirms this, for the first 5 cycles. I had a lot of trouble generating a square wave from -13V to +13V so I am not sure what the current waveform will look like when you switch to -13V when the current flowing is still 1.625A; I suspect it will not ever exceed 1.625A later though. With a 13V peak sinewave (a predefined function) it looked to be just under 1A, but all positive which is obviously BS.

The bit I have forgotten is how to calculate the flux density in the core. I would prefer the whole thing to be something the size of an RM10 core
https://www.farnell.com/datasheets/528671.pdf
In 3C95 material you have 5500nH/T^2.

A toroid would have less stray flux, and again there is a vast choice. I have a bag of these from many years ago
https://www.tme.eu/gb/details/tn16_9.6-3f3/ring-ferrites/ferroxcube/tn16-9-6-6-3-3f3/
which with 83 turns would achieve 8mH.

What I don't get is how much current this will carry. 83 turns at 1.6A is 132 AT which sounds an awful lot. I thus suspect I will need a bigger core, probably iron.

Looking on Ebay for ready made stuff, 10mH, I see e.g.
https://www.ebay.co.uk/itm/115852064002
but I am instinctively damn sure that won't carry the current.

Another option is some toroidal transformer, and ignore the existing winding and put more turns on it. Ebay is full of vintage transformers but most are pretty big.

I was going to wind the TN16 with some turns and see what it does. I have an HP 3314 pulse generator and a power amp which can output 9V peak.

The 8mH needs to be +/- 0.3mH. It was determined experimentally using this amazing thing
https://www.ebay.co.uk/itm/224101042113
but that obviously contains massive magnetics - it weighs about 1kg.

I also have an LCR meter.

I'd be grateful for any suggestions.

[Terry Bites]

https://www.electricaltechnology.org/2014/02/maximum-flux-density-bmax-calculator.html

[Conrad Hoffman]

It's all a balancing act, but I think you can avoid saturation with a small gap. Need more turns though.

[Slh]

B = L*I/(N*A) is what you're after.

Using your 1A, 83 turns and 8mH and Bmax 300mT you need an area of 320mm2.  Alternatively, the tn16 core is going to have 4.8T.

I'd suggest a steel core if you can get one. It might be easier to find a powdered toroid core from someone like micrometals. You're probably looking something on the order of a three inch diameter one.

[PCB.Wiz]

You can 'sanity check' any choice by searching for commercial inductors.

eg I find 8.2mH and 10mH in Vishay IHB-6 series, rated to 1.9A and 1.8A

8.0mH +/- 0.3mH is non standard, but standard values are close.

ie, you could buy one that is close, and remove a few turns.

[peter-h]

Interesting - didn't know these existed e.g.

https://www.mouser.co.uk/ProductDetail/Bourns/1140-822K-RC?qs=EnuEBe%2FYa%2FrMV5h7dv14gA%3D%3D
https://www.mouser.co.uk/datasheet/2/54/5700_series-777305.pdf

They are not small! But I can use them. 3" diameter will be too big.

Also this explains why previous attempts did nothing - they must have just saturated.

I think the above are both ferrite. I use Allied Signal amorphous metal cores at work, in a PSU in a product I designed 25 years ago (sold 10s of k) and that was used because it runs with a narrow duty cycle (wide input range flyback) and ferrite, while working fine, ran much hotter. But these cores need a lot more turns; 500Hz would be a large one.

I can see why iron is much more compact. A quick google:

The typical saturation flux density of Power Ferrite material is under 4000 gauss (400mT). Whereas the saturation flux density of MPP material is 7000 gauss. High Flux is 15,000 gauss and Iron Powder is 10,000 gauss.

Problem is... not easy to buy blank toroids in iron, it seems, in 1-off.

[johansen]

https://www.bridgeportmagnetics.com/bmg-product/c-cores/

I think their second smallest core which weighs 99 grams could do it. CD10x12.5×25

[jbb]

I can see why iron is much more compact
…
Problem is... not easy to buy blank toroids in iron, it seems, in 1-off.

Iron (or similar) is where it’s at for this job; the high saturation flux makes a massive difference. (I once got a ferrite 10mH 10A RMS (20A peak or something) made as a lab item and it was huge).

Potential core types:
- laminate iron in EI sections with an air gap. Note that the gap will get you some control over L and reduce the risk of saturation, at the cost of some more turns of copper.
- nanocryatalline/amorphous iron sections - probably U cores - with an air gap. Note in this application it’s may not perform any better than standard iron.
- something like the Magnetics Inc Hi-Flux material, which can - allegedly - be bought in a variety of core sizes and has air gap built in

[johansen]

- nanocryatalline/amorphous iron sections - probably U cores - with an air gap. Note in this application it’s may not perform any better than standard iron.

The energy storage is according to flux density squared.

A core that can be pushed to 1.9T will do a lot better than one that is limited to 1.5.

[peter-h]

Reading the above, this one must be iron
https://www.mouser.co.uk/ProductDetail/Bourns/5725-RC?qs=m14s2R35DUZfXqUOkLnMFw%3D%3D
even though they say nothing about the material.

[CaptDon]

We used 10Mh chokes which were wound on an open ended laminated bar. The total size of the bar was about 5/8" X 5/8" X 2.5". They were 'off the shelf' stock from Mouser or Digikey. The open bar style prevented hard saturation. I don't remember the number of laminations, probably around 15 or so. They were used in a locomotive fuel injector simulator. B.T.W., G.E. locomotive injectors (Made by Bosch) can stand 16" tall or more.

[jbb]

A thought for the peanut gallery: would it be feasible to buy a 50 Hz power transformer using a laminate iron core & modify it??

[peter-h]

Certainly I could buy a small mains toroid (say 230V to 9V) and either ignore the windings or remove them (with a hacksaw) and then wind some turms on it.

What concerns me a little is that - as posted above - the small signal used by an LCR meter will be stuck down in the hysteresis region.

[MarkT]

So we have 13V across 8mH for 1ms, which from v = L (di/dt) yields 1.625A. A quick hack in LTspice confirms this, for the first 5 cycles. I had a lot of trouble generating a square wave from -13V to +13V so I am not sure what the current waveform will look like when you switch to -13V when the current flowing is still 1.625A; I suspect it will not ever exceed 1.625A later though. With a 13V peak sinewave (a predefined function) it looked to be just under 1A, but all positive which is obviously BS.

For a perfect inductor, square wave voltage gives triangle wave current, and yes the current will not go negative when starting with a full cycle of square wave, as the positive half cycles increase the current and the negative decrease it back to zero.

For a real-world inductor the losses will mean the overall waveform decays till the average current is zero, more lossy the faster it will do this.

For least peak current start the waveform with a half-length half-cycle.

[peter-h]

Really interesting!

For a perfect inductor, square wave voltage gives triangle wave current, and yes the current will not go negative when starting with a full cycle of square wave, as the positive half cycles increase the current and the negative decrease it back to zero.

Is that for a symmetrical about zero input? Like +13V to =13V? When I used a sinewave like that, the current was also sine but delayed - I do understand that. I would have expected a square wave to produce a triangular current waveform symmetrical about zero.

[Slh]

If you start with half a positive (or negative) period then it will be centred around 0.

If you go straight in with 50% duty cycle then the positive pulse takes to +I_peak and the negative pulse takes you back down to 0A - you need it to be on twice as long to get to -I_peak. Then you want the third pulse to be the same length as the second to get back to +I_peak.

I generally find that slowly ramping the voltage on tends to work well enough with a 50% duty square wave centred around 0V.

[Randy222]

Does that pc app Coil32 help at all? At least in construction?

[peter-h]

I ordered that 8mH one from Mouser. I would assume they measure the L at a large signal.

[Randy222]

I ordered that 8mH one from Mouser. I would assume they measure the L at a large signal.

A chart of freq sweeps at various levels would be nice to have.

[peter-h]

That inductor from Mouser has arrived. As expected, iron core toroid, not bad for 8 quid!

[Randy222]

That inductor from Mouser has arrived. As expected, iron core toroid, not bad for 8 quid!

Iron?
Perhaps a ferrite formulation?

[peter-h]

It could be...

[Randy222]

It could be...

My guess, a ferrite powder formulation mixed into an epoxy and then pressed into a form. Yes, "iron", but not just a slug of Fe metal.

I made a few small cores in the past
carbonel iron powder and polyurethane resin

fun projects

[T3sl4co1l]

If it's yellow and white on one side, that's mix #26 by Micrometals, and widely copied elsewhere (I don't think the numbers or colors are standardized at all, but many use them the same for direct substitutes).  It's... not a carbonyl iron I think, but finely powdered, not alloy, and is pressed and cured with a binder.  It's pressed pretty tightly with little binder, giving high permeability.

It's also a notoriously low-Q material (~10 at 100kHz, dropping quickly above there), so not much better than a slug of metal, but it's good for filters where ripple is negligible, and okay for converters in CCM.  Hence their prevalence in computer PSUs (mostly older ones by now, though you still see them, or #52 (green/blue), in modern two-switch and full-bridge forward converters).

Tim

[peter-h]

Yellow one side, white on the other, yellow on the periphery.

I need to pot it in a box because it is far too delicate like this.