LM5021: Doesn't startup properly - Unable to get more than minimum ON time for the MOS

D5 cannot be removed as it is needed to bias the U14.

Please provide the transformer specs, such as primary inductance, turns ratios, etc.

Is R86 value 0.15-ohm?

Please provide U14 VCC waveform together with other waveforms such as MOSFET gate, and CS pin

Hello Hong Huang,

Yes R86 is 150mohms. I've tested with values around 100mohms to 500mohms. Currently it's 125mohms.

Here is the link to the transformer : https://cdn.sos.sk/productdata/9c/40/329447df/7491195112.pdf

I know D5 is needed but I've removed it to simplify the design and finding the issue.

I've changed CSS for 100nF and VOUT=24V with no load.

When I connect a 50ohms load, here is what appens (Vout waveform and VCC waveform).

Vout : We can see that the output voltage can't rise to 24V. The MOSFET turn on time is about 100ns at 267kHz so that's not a surprise.

VCC pin waveform during PWM mode.

Hello Hong,

Sorry I didn't manage to make the captures you requested, I've accidentally touched two pins with the oscilloscope and boom !

I'll repair the board and do the captures.

Anyway, I was looking at the Rsense waveform and I think there is something really wrong here :

If you take the 41uH (@ 100kHz) of the transformer primary inductance, take the DC impedance of (transformer's primary, MOS and Rsense - around 200mohms) and trace the transient waveform for 48V -> 2.3A after 2us. On my waveform, it goes crazy high in less than 0.03us. It's like I got a short circuit instead of the transformer's primary.

There is something wrong here don't you think ?

Best regards,

Michaël

Yes, there is something wrong if you see that narrow pulses. There are several possibilities.

One is to confirm if the transformer primary inductance is 41uH - a simple way is to remove the transformer from your board and use an inductance measuring meter to measure.

Another is to confirm all transformer polarities are correct by testing its polarities so to verify its all physical connections of polarities are correct.

Hello Hong,

Thanks for asking !

I bought an LCR meter and received it yesterday.

Measurements on the transformer (not mounted on the board) @ 100kHz :

Primary : 41uH (as specified in the datasheet)

Secondary : 6.3uH each

So pinout and caracteristics seems OK. I'm thinking layout is not good, I'll start redesining the board and submit schematic and PCB to you. Do you have other ideas ?

Best regards,

Michaël

I would like to measure three waveforms, MOSFET T2,  Vgs, Vds, and the transformer secondary winding output (transformer pin 6 to pin 10). These waveforms can show if there is polarity issue by real test.

But if you are sure the transformer polarities correct, then you can skip these.

Yes, you can send the sch and layout for me to have a look.