Searching \ for '[EE]: forward converter design' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=forward+converter
Search entire site for: 'forward converter design'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: forward converter design'
2001\09\17@165311 by Tom Messenger

flavicon
face
My apologies to everyone for changing the subject line; I don't have the
post here at work about the request from someone asking for forward
converter design methods/equations. I do have some answers however.

First, get a book. Here's mine:
SWITCHING AND LINEAR POWER SUPPLY, POWER CONVERTER DESIGN,
ABRAHAM I. PRESSMAN
Hayden Book Co.
ISBN 0-8104-5847-0
1977

Somewhat dated now, (very little mention of IC's), but an excellent
starter. If you design smps circuits for a living, it's ok to have this in
your library. If you only design smps when forced to, you will get a lot
out of it.

Forward converter design starts with transformer core selection.  Get
magnetic core catalogs from your vendors. Find a core that will handle your
power levels.

Begin with a conservative design- maybe around flux densities of 1500 peak.
Core size is inversely proportional to frequency; smaller cores work at
higher frequencies; larger ones are required the lower the frequency.

Thus higer freqs allow smaller xfmrs but also have greater switching losses
- you decide!

Calculate required number of primary turns to standoff the applied voltage:


             (Edc-1)*10e8
   Np = ---------------------
              4*f*Ae*Bmax

Where:
Np = number of primary turns
Edc = applied input voltage (the -1 is to account for switch losses)
10e8 = 100,000,000
f = operating freq in hertz
Ae = core area in square centimeters
Bmax is peak flux excursion in gauss

Ae will be found in the mag core data books/sheets.


Secondary winding.
For any desired secondary voltage, Vs, its turns, Ns, are:

          Np*Vs
Ns = ----------------
      Edc - Vce(sat)

This doesnot take in account the secondary diode losses, etc.

There is much much more to the whole thing but this should allow you to
choose a set of cores, wind a transformer and begin testing it. I usually
start out by running tests on the first transformer with a signal generator.

Also, you need two more things in your mental junkbox to get to the finish
line here. Your xfmr will have some amount of leakage inductance which will
act like a flyback or boost converter when the switch shuts off.  This is
flux that goes into the xfmr but does not manage to find it's way out via
the secondary winding.  It will make a large voltage spike appear at the
switching transistor.  It is usually handled with a 'snubber circuit'. The
usual snubber is a half wave rectifier (fast) that has an RC load to absorb
the leakage energy.

The other thing needed is a reset winding. If you operate at greater than
50% duty cycle, the core will become saturated with flux and will cease to
look like a transformer.  The reset winding is designed to conduct when the
switch is off to get rid of the core flux.

Finally, I'm one of the people who only do this when forced to; ie.,
definitely not an expert. If anyone sees errors in this writeup, by all
means please correct it so no one is made to stray from the path. Thanks.

Best regards,
Tom M.

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spam_OUTlistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...