Making your own ferrites?

High frequency, antennas

Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 10:56 am

I've recently been planning high frequency switching circuits, a bit like the ones I have made already (500V/6MHz/1A) but at lower frequency/high current. This is both for my own stimulation, and also, hopefully, I'll get some useful power supplies at the end of it.

The subject I have bumped into along the way is the use of transformers and inductors. Figuring I don't really know much about these suckers, I took a look on ebay for a power transformer to test things with, and found a guy selling 13 of them, bobbins stripped from switch mode supplies. I bought those, and a stack of other inductors the guy had collected. Around 4 kg of ferrite altogether, I'm told. It is on its way.

So this will hopefully be my 'fast track' dive into the world of high frequency magnetics, and I plan to figure out how best to measure inductance and B-H curves, and the like.

One thing I have figured so far is that the majority of ferrites seem to be actually too high permeability for the intended power/frequency range I am looking at. My basic calcs show that in the range I want to experiment in (500kHz to a meg, or so - hence I put it here in the 'RF'), most ferrite is too lossy and the reactances too high anyway.

So, [if I didn't already have enough projects already] I am considering buying a kg of iron oxide, one of manganese oxide, and some hard-setting gypsum, and mixing them to cast my own ferrite. I want big pieces of ur=~50 or so, to try in this HF range, so I can make smaller coils than air-cored, but less lossy that regular ferrite with air gaps.

Well, that's the punch line. Is this an interesting experiment to try, or is there info out there that already shows this is a crash-and-burn route to a waste of time (probably quite messy, too!!)?
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Making your own ferrites?

Postby Doug Coulter » Tue Oct 18, 2011 11:55 am

I personally wouldn't try that one. Think "hot isostatic pressing followed by controlled annealing in a hydrogen atmosphere". You might do well without that, maybe not. I've been burned on most of my attempts at do it yourself cores. Your luck might be better, but I wind up buying them.

Ferrites span the range of u, so maybe you just got the wrong sort. You can often do fine, just don't push the B limit too hard and the losses go down (most of the losses are per-cycle and per B swing, see Pressman or Terman). They're not linear, which leads to some losses, but the upside is they don't saturate hard and sudden either - which saves you a lot of semiconductors while learning things about what that can do to them (along with stored energy).

Fine tape mo-permalloy (Arnold magnetics) is best for a lot of things, and luckily you don't need too much (it's real expensive). Almost no hysteresis losses, and with super thin tape, low eddy losses too. A ring 1/4" cross section, 2" diameter will do 500W and up.

If course, if you're going to blow sq waves into this, you need 5mhz low losses too to handle the edges and the harmonics they create, or a filter between the switches and the "tank circuit".

These guys don't have the lowest prices, but ship anywhere:
https://www.amidoncorp.com/

You might do a bit of googling to find more-local suppliers. Most of the guys who make this stuff just have one big factory and ship worldwide from it -- and some of those are on your side of the pond. The big problem is finding a distributer for small amounts, none of the retailers carrry much, as there is just too much custom work and too many to make profit - most of it would sit on the shelves forever.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 12:17 pm

Doug Coulter wrote:You can often do fine, just don't push the B limit too hard and the losses go down.


Well, that's exactly the thing I want to explore. By having a core that is 'not much material', would I not increase the saturation limits, plus if I am making the stuff myself in bulk I can make a huge crossection, thereby meaning I can run much higher V/turn that conventional ferrite will permit?

Bottom line is that if I could get away with one turn primaries and still have efficiency, then I'd be happy!! Avoiding running the core with high flux is the wrong way to the way I'd like to push it!!

Hey, look, I would have expected someone to be making this type of material if it was a useful way of doing things, so I am suggesting this is no more than an experiment which will take me (oblige me) down a route of figuring out how to measure BH/losses/saturation etc.. The thing is, I might even come to a useful conclusion/point, because I'm interested in the ~1MHz region, whereas for most folks the 20-40kHz range is real easy to switch and matches the ferrite performance (which I guess is no mere coincidence!).

If you have any data tables on coercivity for a big range of common materials, in the mass of stuff you've collected, Doug, please do let me know. The way it reads to me, as far as I can figure it, is that very fine iron oxide tends to have lower coercivity (compared with just 'iron oxide' - how small I do not know) and that manganese oxide exhibits diamagnetic effects which reduce the ur and saturation, but tend to counteract the coercivity of the other ferrite components. Obviously, ferrite manufacturers aren't going to just post up their secret sauces, so the only route is by experimentation. Unfortunately, I have a very limited range of substances I can conveniently buy.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 1:52 pm

Doug Coulter wrote:These guys don't have the lowest prices, but ship anywhere:
https://www.amidoncorp.com/


OK, so [hat-tip to that] a bit of random hunting shows up 'material 67' with ur=40.

But unfortunately looks like they only make rings of it - looks like E cores come in material 77 only. Not that that doesn't look like a great material (max flux density 0.46T, nice if that were through its full operating frequency range - though doesn't specify what frequency that's valid to) but too high ur, I think.
Last edited by chrismb on Tue Oct 18, 2011 2:01 pm, edited 2 times in total.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Making your own ferrites?

Postby johnf » Tue Oct 18, 2011 1:56 pm

Chris
I like Ferrite type 43 for most things, it does well for transformers in the MHz range and is an excellent de-coupler for baluns up to 500MHz.
If this is still too high in u then iron powder is a better choice and is almost impossible to saturate --this you can make yourself out of ironsand and powdered glass. The ratio glass to ironsand gives the u figure -you do need a furnace and a ballmill.

Type 65 ferrite is also good --check out motorola app notes from Helge Granberg using transmission line transformers MRF151G, MRF141G
And I forget the part number that used the HOG pack.
PS Helge was to RF what Jim Williams was to analog
johnf
 
Posts: 433
Joined: Sun Aug 08, 2010 3:51 pm
Location: Wellington New Zealand

Re: Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 2:00 pm

Thanks, John,

johnf wrote:you do need a furnace and a ballmill.


hmmm... I was hoping just to cast it into hard plaster....would that be no good??
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Making your own ferrites?

Postby johnf » Tue Oct 18, 2011 2:10 pm

Maybe for low u epoxy resin and ironsand --but you will still need the stuff to be graded for fineness
johnf
 
Posts: 433
Joined: Sun Aug 08, 2010 3:51 pm
Location: Wellington New Zealand

Re: Making your own ferrites?

Postby Doug Coulter » Tue Oct 18, 2011 2:14 pm

Volts/turn is also a function of frequency as well as u (and magnetic path length). In the 3c9x (x = 2 or 3) stuff I have, I'm using 8v/turn at ~~18khz for a 1" cross section (1x1), it would scale as you go up in F more or less. Low u stuff means that in general you have to wind around the whole core, or lose a ton. I found this out when trying to use some amateur radio balun cores with the primary on one side, the secondary on the other for DC isolation. That one just does not work, far too much field loss - low u implies that it's not good at guiding field around corners.
Even with the high u stuff - I had to make my 3 turn primary go most of the way around the core - putting the 3 turns tight on one side had much too much leakage inductance to the secondary (but a little is good if you want to drive with a sq wave, but don't need a sq wave output) Both Spellman and Glassman use that trick in their special transformers, so their switches don't try to drive what amounts to a short circuit with the fast edge frequency components well above the self-resonant frequency of the transformer. In this case, the leakage inductance is the needed low pass filter between the switch and the transformer.

At any rate, at 500khz, volts per turn isn't going to be much of a limit for most reasonable things.

Trying to get super high turns ratios is not a reasonable thing (and why I've launched some threads in the attempt to attract some real experts who can tell me how far I can push it best case), and is in general limited by self resonance of the many-turn winding. Going to higher u to reduce the turns doesn't really buy you much either, I've found. In fact, this truth is why there are such things as CW voltage multipliers, rather than just winding a super-transformer and using a single rectifier -- even at mains frequencies.

Remember that a transformer has the most "B" in it when unloaded. This is why most of the X ray transformers we've bought can't work for a fusor. They saturate too easily unless they are fully loaded and burn up on even short runs if you don't limit the input current - some heat significantly even when fully loaded in a one second X ray exposure. Drawing current from the secondary cancels some of the primary's field is how that all works, similar to the concept of back emf (but different, that's a gross analogy). So things that have a ??? load (or even open circuit) are much harder to do and much more expensive than things with a well-defined load - see the prices of AR wideband amplifiers and compare to a ham transmitter...about 30::1 or more ratio there.

You might fool with grinding up some ferrite and adding more binder (epoxy) to reduce the u if you want to. I've used a rock tumbler with alumina media to fine-grind things like that, takes a lot of patience but not much sweat. Ferrites in general are good to the mhz region at least. Did you check specs on the link I gave you yet?

Here's some stuff downloaded from there, gheesh.
FerriteProperties.PDF
(34.88 KiB) Downloaded 442 times

SoftFerrites.pdf
(64.28 KiB) Downloaded 429 times


They also do iron powder, mo-perm powder (the best) and you-name-it. No need to measure this stuff's BH and hysteresis loops; unless that's your main project. Just get something with specs and do the project. Call them if they don't have what you need. You might be able to social-engineer a couple samples, I've done that. You sometimes have to do a little work to get what you want in that game.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Making your own ferrites?

Postby Doug Coulter » Tue Oct 18, 2011 3:26 pm

I'll second John on the types 43 and 65. 43 is probably best for what you want (if I understand that). They use it for AM radio loopsticks, among other things, and it gives nicely high Q in those. Nice low loss stuff. This low u stuff you can take bits and glue them together to get shapes you want (within reason) without losing much due to a little more gap in the system -- they are mostly gap (due to the binder or inert stuff between the particles) anyway. I have seen cases where people who don't like that have used epoxy with a slurry of ferrite dust in it to do the gluing as well. There's probably a commercial product like that, but I never found anyone selling it. I've seen it in various things that have ferrites in them though.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 3:45 pm

Doug Coulter wrote:Did you check specs on the link I gave you yet?


Yes. Did you see my post above?

Actually, what I was interested in seeing was a list of magnetic data for regular materials available to me, as chemical feed. And, yes, one of my interests is trying to evaluate these material properties for myself so I gain a deeper understanding of what it all means.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Next

Return to RF

Who is online

Users browsing this forum: No registered users and 3 guests