UCC28951: magnetizing Inductance

Hello,

Please see my comments below.

Vin MIN : 400VDC

VinmAx : 950VDC

Vin ( Typ ) : 600VDC

1. the question is primary magnetizing inductance of a psfb transformer should be selected considering Max input voltage or minimum Input Voltage ? Considering min duty cycle or MAx duty cycle ??

> Turns ratio and Lm min input voltage max duty cycle.

>Application note slua560 in the following link will give guidance on how to select these parameters.  It can be found at the following link.

https://www.ti.com/lit/pdf/slua560

As per my manual calculation,

considering Vin Min ( 400VDC ), max duty cycle ( 0.66) , Lmag is 1.6mH

considering Vin Max ( 950VDC ), min duty cycle (0.3) , Lmag is 8mH.

Additionally excel tool considers typical voltage value while calculating. I mean what is the correct value here such that the control operates in peak current mode control ? My system can be anywhere between 400VDC to 950VDC. 

> Excel tool use the minimum to calculate transformer Lm and Np/Ns.

2. additionally, datasheet says, " If the magnetizing current increases enough, it can swamp out the current sense signal across RCS and the converter will operate increasingly as if it were in voltage mode control rather than current mode." As per excel tool calculation, considering 400VDC, Ipp = 6.17A. I want to Understand how much is enough current here ? Please explain in plain simple terms.

> Application note SLUA560 gives recommendations on sizing the current sense resistor and slope compensation so this does not occur.

I have to share data to Transformer Manufacturer

SLUC222E - 63277 R05_12V.xlsx

3. Additionally I have a prototype board with transformer magnetizing inductance of 4mH( taken from excel tool ), when I try to operate at low load or No load @950VDC, this is the waveform. Is this the actual issue you're trying to address in datasheet , about shifting from peak current mode to voltage mode ? If not,  What exactly is this ? How to troubleshoot?

> Your waveform looks to be the voltage across the input of the transformer and the current through it.

The transformer primary peak current does shoot up higher than it should.  This occurs when the design goes into freewheeling.  It also looks like during this period that the output inductor is saturated as well.

From looking at this waveform it looks like you are using a current doubler.  Two inductors instead of one and when the design freewheels both SRs are on at the same time.  I would use the DCM comparator to turn them off before the output inductors go critical conduction.  Application note SLUA560 describes how to do this.  I believe this should remove your issue.

Regards,