UCC28951: Different pulse width on OUTA-B-C-D

Shaofeng Luo

Part Number: UCC28951
Other Parts Discussed in Thread: UCC28950

Hi expert team

Our customer is developing 5kW DCDC with UCC28951, voltage mode, but we encountered some tricky issues as below:

Yellow-OUTA, Green-OUTB, Blue-C, Red-OUTD

We found that the pulse width are different, for fig 1, Yellow-OUTA pulse width is 5.7us, but for fig 3, Green-OUTB is shorter than OUTA.

And Red-OUTD is also shorter than 5.7us. And this mismatch will continue for several cycles.

This issue can be easily caught and it occurs in light and heavy load. From our side, at least A-B are fixed 50-50 duty cycle, but in this case, each OUT channel is different on pulse width.

Any insights on it? Thanks!

over 2 years ago

0 Mike O' over 2 years ago

TI__Guru* 82536 points

Hello,

I have received your inquiry and it is under review and I will get back to you shortly.

Regards,

0 Mike O' over 2 years ago in reply to Mike O'

TI__Guru* 82536 points

Hello,

The UCC28950 and UCC28951 both experience this behavior under certain conditions. The following link they reviewed it and discussed this issue in detail.

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/224390/strange-h-bridge-driver-signal-for-ucc28950

If this information help resolved your issue. If this resolved your issue, please hit resolved at the end of the thread.

Regards,

0 Shaofeng Luo over 2 years ago in reply to Mike O'

TI__Intellectual 1630 points

Mike

It's not the same with my case, please help to take a look on my scopes and help analyze the root cause. Thanks.

0 Mike O' over 2 years ago in reply to Shaofeng Luo

TI__Guru* 82536 points

Hello,

Let me review this some more.

Regards,

0 Mike O' over 2 years ago in reply to Mike O'

TI__Guru* 82536 points

Hello,

The UCC28951's outputs A and B are 180 degrees out of phase with each other and synchronized to a master clock. Output C and D are phase shifted to control the converters duty cycle. The phase shift if is controlled with the voltage amplifier.

If the converter all of sudden demands 0 duty cycle. Out C and D will be immediately phase shifted 180 degress to demand duty cycle. This will make out C and D duty cycle look twice long, when it is really not. The outputs are just being phase shifted. I believe this is the behavior you observing.

Regards,

0 Shaofeng Luo over 2 years ago in reply to Mike O'

TI__Intellectual 1630 points

Hi Mike

Thanks for your reply, yes, I know this issue of C&D. But our case here is different with that, we notice that the pulse width of OUTA and OUTB are different. And we found that it appears when UCC28951 enters normal mode from burst mode, and the switching frequency dithers slightly comparing with the setting one. Is it normal on UCC28951? We think that the switching frequency should be fixed when we set the RT resistor.

Expect for your insight, thanks!

0 Mike O' over 2 years ago in reply to Shaofeng Luo

TI__Guru* 82536 points

Hello,

I am reviewing your inquiry and will get back to you shortly.

Regards,

0 Mike O' over 2 years ago in reply to Mike O'

TI__Guru* 82536 points

Hello,

What are the measured on times for OUT A, OUT B, OUT C and OUT D?

Regards,

Power management

Power management forum

UCC28951: Different pulse width on OUTA-B-C-D