Simulating pulse transformers in LTSPICE.

[John Nagle]

I'm playing around with a flyback DC-DC converter design, and
I'd like to know if I'm modeling a stock part correctly.

The part is a Pulse Engineering PA1546NL, which is a small
transformer.

Data sheet:

http://ww2.pulseeng.com/products/datasheet.aspx?Datasheet_Id=935

I want to use the winding between pins 4 and 5 as the primary (L1),
1 and 8 as the secondary (L2), and not use the 7-2 feedback winding
(not modeled).

The SPICE model I'm using for the transformer is:

L1 Vcc N003 17µ Rser=2.1
L2 0 N001 247µ Rser=0.7
K1 L1 L2 0.95

The data sheet gives the inductance of L2 as 247uH.
The data sheet doesn't give the inductance for L1, but it gives
the turns ratio as 3.82 : 1. So 247uH / (3.82^2) = 16.9uH, right?
DC resistances are from the data sheet.

I have no idea what the efficiency value, K, should be for one of these
little things. I'm using 0.95 as a guess. Big iron-core transformers
run around 0.98 or so, but I don't know about these little things.
This matters; for the circuit I'm trying, for values much below 0.95, the
switcher won't oscillate.

John Nagle

 

[John Nagle]

"K" isn't "efficiency", it's _coupling_. Grab your manual to get the
equations... manipulate to get leakage inductance. 0.95 is likely way
to low.

...Jim Thompson

OK. The leakage inductance from the data sheet is 8uH (measured with the
other coils shorted) vs. 247uH with the other coils open. So

L1 = (1-K^2) * L2
or
K = sqrt(1 - (L1/L2))
K = 0.9837

Is that right?

John Nagle