What am I supposed to put in magnetization inductance?

Hi,

I want to simulate this push-pull DC-DC converter in simulink:

I am modifying the magnetization inductance (Lm) for a value that is appropriate for my specs: Vin = 137.14 V (it comes from a previous stage),Vout = 350 V, D = 0.13, Iin = 7 A (approx.). I see that even if I set it in SI units, it says (pu) for the magnetization inductance. It confuses me because I was trying to have 1 mH or 10 mH to avoid an overvoltage and I don't know in which units it's actually. I was also trying with higher values of Lm but I have that peak at the beginning. In fact I have seen 2 situations.

A) Overvoltage, then stabilizes but the value was a few volts far from the reference. This is the case when magnetization inductance is 0.0.

B) Overvoltage, then starts dropping to 155 V.

Can someone tell me what are the expected values in the parameters of the transformer for such specs and in which units is calculating Lm when I set it in SI units?

Also, the input current has this aspect:

Thanks.

Carlos

Answer 1

[sneha]

Hello,

In your push-pull DC-DC converter model, the magnetizing inductance (Lm) value is shown in per-unit (pu) because Simulink normalizes it using the transformer’s base voltage, base current, and frequency.

That means even if you enter Lm in Henrys (H), Simulink converts and displays it as a relative (pu) value, it doesn’t change the actual inductance used in simulation.

Helpful MathWorks Documentation:

Per-Unit System (Simscape Electrical)

https://www.mathworks.com/help/physmod/sps/powersys/ref/units.html

explains how Simulink converts SI values to per-unit.

Transformer Block Reference

https://www.mathworks.com/help/simscape/ref/idealtransformer.html

shows parameter units, base values, and how magnetizing inductance (Lm) is defined.

Push-Pull Converter Example

https://www.mathworks.com/matlabcentral/fileexchange/94220-simulation-of-push-pull-converter

demonstrates proper transformer parameterization and flux balancing.

DC-DC Converter Examples

https://www.mathworks.com/help/sps/ref/dcdcconverter.html

“Avoiding DC bias in push-pull” section explains flux reset and why startup spikes occur.

[Carlos]

Hi,

Sorry but this was not helpful at all. I am using a multi-winding transformer from the black blocks (not the blue ones like the ideal transformer you referenced). The link you stated about how to avoid DC bias in push-pull doesn't say anything about that. And I still don't know if I have to enter the magnetization inductance in SI or pu and which is an appropriate value (or leave it blank).

Not only those problems I am having. It works more or less OK (with the startup spike and a 5.something Amps ripple in the input) but only in open-loop. When I was trying it in close-loop, the duty cycle bounces between 0.05 and 0.5 a few times before it gets saturated in one of those values. I fixed some limits because the push-pull can't go beyond 0.5 and the lower limit is for practical purposes, but duty cycle is not calculated to be 0.5, but a value between those limits. In fact, if I make automatic tuning of the PI in frequency response, it calculates very high gains, sometimes negative. Could you help me with this please?

Let me show you how the diagram looks like:

Thanks.

Carlos