UCC28950 Transformer Design

hi,

I am using UCC28950 in a full bridge converter (without ZVS). Here are the specs

Input 385 VDC (from PFC)
Out Put Voltage 28VDC
Output current 29A
Converter primary switching frequency is 100Kz

I am using RM14 C95 core. and following the schematic given in the datasheet and using the excel sheet calculator.

I can test the DC/Dc converter for 30 min. At the start transformer generates some sharp noise. Transformer temperature rise to 80C an i need to place a fan near it.

After 30 min secondary side synchronous FET get damaged.

I want to discus the transformer design. If the calculated primary magnetizing inductance (using UCC28950 Excel Design Tool_SLUC222C.xls) is 2mH, is it possible to make primary 7mH or even large? I am also using ansys PExprt tool to design transformer and it always give high primary inductance then " UCC28950 Excel Design Tool_SLUC222C.xls".

Pexprt gives 39 primary turns with 13mH magnetizing inductance (turn ratio is 9)). " UCC28950 Excel Design Tool_SLUC222C.xls" gives 1.94mH magnetizing inductance. Which design should i follow?

Second i have to place 20 ohm resistor (connected to the CT transformer) to run the converter. The design sheet gives 36 ohm.

Regards

Sharp noise? Temperature rise?

This appears suspiciously like the core is being operated with too much flux density. But this suspicion requires to be corroborated.

Do you have access to a current probe? You will require to see the no-load primary waveforms.

Edit: you may see a voltage replica of it across R7.

Thanks for reply.
I dont have current probs.

Intilay I was testing the DC/DC board at 200V RMS ( generated by auto transformer and rectifier) .There was no noise at 20 a output and also at 29 A. When input voltages are increased from 200 to 220 RMS stange noise us generated and after some timed it automatically vanished or become very low.

Temparure with fan rises up to 80C at full load .

Another thing , I can't short DCM pin to ground to force the contoleer into continuous mode. In must have to place some resistor divider here.

- - - Updated - - -

I have not installed shim inductor to achieve ZVS.

- - - Updated - - -

I have burned very few bridge side MOSFETs . Most of the time synchronous FEt are burned.

- - - Updated - - -

Can RM 14 (C95 )handle 800 watts?

If the calculated primary magnetizing inductance (using UCC28950 Excel Design Tool_SLUC222C.xls) is 2mH, is it possible to make primary 7mH or even large?

Click to expand...

Yes. 2mH is the minimum primary inductance.

hi,

I have placed few 1uF /500v capacitor in series with the primary winding. There is no noise from transformer at 21A of load. At 0 load, out put voltage are 28V which drops to 21V (and remain constant) at different loads 5, 10 15 and 21 A.

Second i can't run the ucc28950 in CCM mode even by shorting DCM pin to ground. even at 20 A load the Synchronous FETS are off.

Attached pic are Current transformer Rs signals.

Regards

have you implemented a soft start? if your layout is too "open" you will make a fair amount of RF noise that will get into the control circuits and cause grief...