[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] RF transformer design




-----Original Message-----
>From: Finn Hammer <f-h@xxxx>
>Sent: Dec 21, 2007 1:14 PM
>To: Tesla list <tesla@xxxxxxxxxx>
>Subject: [TCML] RF transformer design
>
>All,
>
>One thing has been high on my wish list for quite some years now.
>
>How to design a RF power transformer, on a ferrite core, without 
>resorting to trial and error.

Aren't you a bold one!


>
>My experience with transformer design is limited to 50Hz, and Many of 
>you all remember how I built the Porkchop, but that was simple.
>
>My present needs are in the 20-50kHz, 15-20kW range.
>
>I have been offered a limited time trial on IEEE enterprise, with 10 
>free downloads, and perhaps there are papers there, describing a usable 
>approach.
>Pointers to papers are invited.
>If there are any of you that can suggest text books on the subject, I`d 
>be grateful too.


In my opinion, there's only one book worth getting.  E.C. Snelling, "Soft Ferrites: Properties and Applications" published by Butterworths, the 1988 second edition preferably. It's essentially a book length ap note from a specialist at the manufacturer.

It's out of print, and quite rare.  However, this opens an opportunity.  In the U.S. at least, copyright law has a specific exception allowing a research library to make a copy of a book that is rare or hard to get.

It is a WONDERFUL book with everything you could possibly need to know about designing transformers, inductors, etc with ferrites and powdered metals.  Every detail, idiosyncracy, etc is covered, obscure side effects, etc.



The other approach is to trust in one of the superduper FEM codes and hope they built their material property libraries right.  At least your trial and error won't involve tedious winding, just tedious mouse clicking.

There are other reference works out there, but they seem oriented more to just getting the first rev in the ballpark, and then it's up to you to do some empirical cut and try.

In high performance HV transformers for PWM switchers, particularly, it's very much an art.  At work, we buy Traveling Wave Tube Amplifiers for deep space comms.  A TWTA is the combination of the actual tube (the TWT) and the HVPS.  Both are essentially works of art and craft, with a healthy dollop of engineering.  For the HVPS side you have to deal with dielectric stress and thermal dissipation, as well as whatever weird performance requirements the customer puts in. The guys making the HVPS  base their designs on long experience and a heap of in-house specialized modeling codes to do this, and even then, you wind up making a couple tries before you get the final design.  Stuff like parasitics is too hard to predict, and you have to tweak the HVPS outputs anyway to match the tube's unique characteristics.

And even companies who generally do a good job occasionally "lose the recipe".  In the 80s, Sony, which had a reputation for rock solid reliability in CRT monitors and TVs (I have a 30 year old Trinitron that still works just fine),  had a run of truly lame 2nd anode supplies with terrible failure rates.  Partly it was cost pressures, but partly it was because they changed designs and hadn't gone through the extensive trial and error learning curve.



OTOH, if you're just hacking out a flyback for a bug zapper....



>
>Together with Daniel Uhrenholt, I am in the process of making a CCPS, 
>based on SLR topology, and I`m stuck on designing the output transformer.
>
>Hope someone helps :-)
>
>heers, Finn Hammer
>
>_______________________________________________
>Tesla mailing list
>Tesla@xxxxxxxxxxxxxx
>http://www.pupman.com/mailman/listinfo/tesla

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla