Making your own ferrites?

High frequency, antennas

Re: Making your own ferrites?

Postby chrismb » Tue Oct 18, 2011 4:55 pm

Doug Coulter wrote: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),


Not sure I follow what you mean. Volts/turn is volts/turn? Max flux generated is Av.Volts/turn/frequency/cross-section(/4). I'm looking at designing for 50V/turn and 500kHz. So if my max flux, in ferrite, is say 0.05T then I'd need a core cross-section of 5cm^2, unless I can find a material with higher saturation. Like the iron powder, but I think that would have prohibitive losses, so maybe [hence me looking] there is some option in the middle-range, as I don't need the big ferrite ur at that frequency.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 7:32 pm

Re: Making your own ferrites?

Postby Doug Coulter » Tue Oct 18, 2011 6:54 pm

What I meant is that if you get 1v/turn at 10khz (without saturation), you'll get 2v/turn at 20khz. Scales directly, all else equal. Losses due to hysteresis also scale with frequency, directly, since going around the loop is a cost-per-cycle thing. Running things near saturation usually increases the core loss per watt through the system as a percentage of total power and at some point, core heating kills your transformer even if you don't care about efficiency very much - it can be hard to get the heat out so things don't get hurt, though some cores actually do better when warm.

You can also do better with a guaranteed minimum load as noted above.

I'm not expert enough to say if using a bigger core at lower B will reduce losses enough to be worth doing it - it also increases the copper losses (since the windings around a larger core get longer), which are very significant at those frequencies - the reason Litz wire was invented to reduce skin effect losses.

The reason people are saying "ball mill" is that the tinier the core particles, the less eddy current losses in them. This is why they spec volume resistivity in the core specs I uploaded above.

Mo perm is the best, you can get it in powdered cores with all kinds of u specs, nothing much has significantly lower losses. You don't see it much, since it's quite pricey compared to the rest, but in things where you're pushing the limits of what's possible and cost isn't much object, that's what people use.

If you read the old engineering books about transformer design (principles are the same since a long while back) you find that it was then a completely cut and try thing -- there are too many crossing curves to optimize it completely feedforward though the math is a little better now, it's still more work than making and testing a few. You can plot the historical ratio of iron to copper by dating transformers made in mass production in history, as cost was obviously one of the curves....and transformers were made to cost the least possible. You can use more core to get less winding (to a point) or vice versa if you don't care about losses. Most manufacturers didn't as the customer paid that in use, not in sale price (still true for induction motors).
So when transformer parts were expensive, they tended to get extra hot.
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: 2968
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Previous

Return to RF

Who is online

Users browsing this forum: No registered users and 1 guest

cron