Maker Pro
Maker Pro

forward converter, transformer reset winding

J

Julius Junghans

Jan 1, 1970
0
hello netlanders,

Im in the process to calculate a forward converter transformer, im hanging on the reset winding... in the smps book from A. Pressman theres this formula:

Lmg = Al * (n/1000)^2 = 2900nH * (4/1000)^2 = 46,6mH

Al = 2900nH Epcos Datasheet E42/21/15 - N27 core.
n = secondary turns

Im not sure if nH for Al is right here, if its right the next step is to calculate the rms current through that winding:

Img = 0,365 * ((Vdc(min) * ton(max)) / Lmg) = ((0,365 * 248V * 13,3us) / 46,6mH) = 71mA


Img = current of the reset winding
Vdc(min) = min rectified voltage
ton(max) = max ontime, f = 30kHZ, T = 33,3us


is this correct?
 
J

Julius Junghans

Jan 1, 1970
0
hello netlanders,

Im in the process to calculate a forward converter transformer, im hanging on the reset winding... in the smps book from A. Pressman theres this formula:

Lmg = Al * (n/1000)^2 = 2900nH * (4/1000)^2 = 46,6mH

used the secondary turns number instead of the primary, very bad idea.
 
T

Terry Given

Jan 1, 1970
0
Julius Junghans said:
hello netlanders,

Im in the process to calculate a forward converter transformer, im hanging
Lmg = Al * (n/1000)^2 = 2900nH * (4/1000)^2 = 46,6mH

Al = 2900nH Epcos Datasheet E42/21/15 - N27 core.
n = secondary turns

Im not sure if nH for Al is right here, if its right the next step is to
calculate the rms current through that winding:


A common practice is to specify Al in mH per (1000 turns) squared. This is
the inductance (in mH) that you get with 1000 Turns. It is numerically
equivalent to nH/turn^2. (4/1000) is re-scaling this silly Al value for the
correct number of turns.

I had a look thru my 2nd edn of Pressmans book, and couldnt see this
equation, but it appears the actual calculation is also wrong. If Al was
specified as 2900H/1kT^2 then 2900*(4/1000)^2 = 46.4mH. This would be an Al
of 2900uH/T^2 which is a very high number - perhaps for iron, but certainly
not for ferrite, where 2900nH/T^2 is about the right number.

Img = 0,365 * ((Vdc(min) * ton(max)) / Lmg) = ((0,365 * 248V * 13,3us) / 46,6mH) = 71mA


Img = current of the reset winding
Vdc(min) = min rectified voltage
ton(max) = max ontime, f = 30kHZ, T = 33,3us


is this correct?

this equation is numerically inconsistent too = ignoring us & mH,
0.365*248*13.3/46.6 = 25.835. If we throw in us & mH we get 25.835mA which
is nothing like 71mA.

one of my pet peeves about Pressman is his over-enthusiastic use of magic
numbers. Its bad enough he uses silly units, but then goes and makes up his
own equations by fixing things like D. It gets really hard to tell if he has
made any errors.

But what you probably want to do is this:

Given a forward converter transformer design - ie the core Ae, le, Ve, ue
and Al are known, along with Bmax, Np, Nsec, T, D, Vin, then:

Lmag = Al*Np^2 (for Al in nH/T^2 L is in nH)

Imag = Vin*D*T/Lmag - peak magnetising current, which is a triangle wave


IIRC Pressman likes to use RMS values, so has crazy numbers everywhere..


Cheers
Terry
 
J

Julius Junghans

Jan 1, 1970
0
hello netlanders,

Im in the process to calculate a forward converter transformer, im hanging on the reset winding... in the smps book from A. Pressman theres this formula:

Lmg = Al * (n/1000)^2 = 2900nH * (4/1000)^2 = 46,6mH

Al = 2900nH Epcos Datasheet E42/21/15 - N27 core.
n = secondary turns

n = primary turns
i love the math:)
 
Top