UCC28251: Switching frequency problem

Sergei Ivanov

Prodigy 160 points

Part Number: UCC28251
Other Parts Discussed in Thread: UCC28250

Hello!

I have a problem with switching frequency after power-on. In my circuit Rt = 75KOhm (Fsw = 100KHz), Csoftstart = 100nF.

When I "power-on -> power-off -> power-on immediately" -> Fsw = 50kHz.

When I "power-on -> power-off -> wait for 15s -> power-on" -> Fsw = 100kHz.

When I "power-on -> power-off -> discharge Csoftstart immediately -> power-on immediately" -> Fsw = 100kHz.

What happens with Fsw after fast power cycling?

Thanks.

over 6 years ago

0 David Huang33 over 6 years ago

TI__Genius 9140 points

Hi Sergei,

You can have a look at the page 7 of the datasheet. I think maybe you forget to set the Rsp in your circuit.

Best regards,

Wenbin

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

Both SP and PS are connected to VREF.

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

In my circuit EN pin is connected to VREF. I have noticed that when I power cycle the circuit the Csoftstart does not discharge. And when UCC28251 in started while Csoftstart is still charged the Fsw is wrong (50KHz instead of 100Khz). How does Csoftstart influence Fsw?

0 David Huang33 over 6 years ago in reply to Sergei Ivanov

TI__Genius 9140 points

Hi Sergei,

SP connect to the VREF mean there is no dead time between the SR mos and the Primary mos.

I suggest you try add the Rsp.
according to your question "How does Csoftstart influence Fsw?", I will ask our IC designer. Once I have the answer, I will reply to you.

Best regards,
Wenbin,

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

Hello!
Is there any information about Csoftstart and Fsw?

0 David Huang33 over 6 years ago in reply to Sergei Ivanov

TI__Genius 9140 points

Sorry， I still get the feedback.

0 David Huang33 over 6 years ago in reply to David Huang33

TI__Genius 9140 points

sorry， I still do not get the feedback.

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

Hello!
How soon your IC designer can answer?
Our development can't wait. We are thinking to reject UCC28251 because of undefined behavior.

0 Sergei Ivanov over 6 years ago in reply to Sergei Ivanov

Prodigy 160 points

Hello!
I have also determined that Csoftstart is not discharged after powering off the IC although on page 24 it is said that SS pin is pulled to ground when UVLO condition happens.

0 David Huang33 over 6 years ago in reply to Sergei Ivanov

TI__Genius 9140 points

Hello Sergei,
can you help scan the waveform for me? including the uvlo process and the ss cap do not discharge with the 50Khz frequency.
since it is not a new IC, I do not have the evm to do the test.so we have to present as much a possible to our IC designer , or they will unable to find the cause. many thanks.

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

Hello! Here are some pics.

Circuit:

Normal Vramp (red, 1V/div):

Vramp after power-cycle (red, 1V/div):

Vcc(yellow 5V/div) Css (red, 1V/div):

0 David Huang33 over 6 years ago in reply to Sergei Ivanov

TI__Genius 9140 points

hello !
you just need to capture the waveform that is 50khz and 100khz pwm with both the Vss and VCC voltage in the waveform. Please make sure to be clear. thanks.

0 Sergei Ivanov over 6 years ago in reply to David Huang33

Prodigy 160 points

Hello!
You can look at tha waveforms of Vramp. Vramp after power-cycle is decreased. Vcc is 15V and supplied from stable power source.

0 Peter Meaney over 6 years ago in reply to David Huang33

TI__Mastermind 31580 points

Hello Sergi

Are you still working on this issue or have you found a solution to your issue?

Regards

Peter

0 Sergei Ivanov over 6 years ago in reply to Peter Meaney

Prodigy 160 points

Hello!
Still working...No solution.

0 Peter Meaney over 6 years ago in reply to Sergei Ivanov

TI__Mastermind 31580 points

Hi Sergei,

I need to order a test board internally and we can test this on our bench as well. In the meantime can you record the Vcc waveform when the IC shows the issue and by this i mean from the time you turn-off the Vcc to the time to reapply the Vcc. From the scope picture above the Vcc appears to decay slowly, the time div is 2s/div?

Thanks

Peter

0 Sergei Ivanov over 6 years ago in reply to Peter Meaney

Prodigy 160 points

Hello!

On the last picture I manually decrease VCC (yellow) from 12V down to 3V. But when VCC is below 4.3V (UVLO according to datasheet) Css is not discharged fully. When VCC = 3V the voltage on Css is 1.9V. And if I power up now (while Css is not discharged) the frequency will be lower than expected (waveform 2, Vramp after power-cycle).

I don't know how the Vcc waveform will help you. But what I have discovered more is that when feedback circuit is open and Vcc is power-cycled the working frequency is OK (however Css is not discharged). But when feedback circuit is closed and Css is not discharged before powering-on the working frequency will be wrong. So the working frequency depends on some combination of voltage on Css pin and voltage on EA+ pin.

0 Sonal Singh over 6 years ago

TI__Expert 6080 points

Hello Sergei,

I was trying to replicate this issue but I do not see FSW change during VDD ON/OFF. Attached are the waveforms I see and the SW is ~100Khz throughout. Channel2: OUTA Channel3: VDD

Can you please share the waveforms where you experience this behavior.

Kind Regards,

Sonal

0 Sergei Ivanov over 6 years ago in reply to Sonal Singh

Prodigy 160 points

Hello!

Please show me your circuit. If Csoftstart is discharged with some external circuit than everything will be ok.

0 Sonal Singh over 6 years ago in reply to Sergei Ivanov

TI__Expert 6080 points

Hello Sergei,

I used the EVM for UCC28250 and there is no external circuit to discharge the CSS. Can you show me the rest of your schematic so we have a better understanding if something else is keeping the CSS charged?

Regards,

Sonal

0 Sergei Ivanov over 6 years ago in reply to Sonal Singh

Prodigy 160 points

The rest of a schematic has nothing interesting. OutA and OutB are connected to half-bridge driver (IR2113S). Feedback signal is connected to optocoupler open collector.

0 Sonal Singh over 6 years ago in reply to Sergei Ivanov

TI__Expert 6080 points

Can I ask you to replace the controller with a fresh one. I suspect since I am not seeing the same issue with my controller, you might have damaged some internal circuit in your attemp to start the controller. Do you have another controller or can you order new samples?

Regards,
Sonal

0 Sergei Ivanov over 6 years ago in reply to Sonal Singh

Prodigy 160 points

I have already tried to change the controller. Both work in the same way.

Power management

Power management forum

UCC28251: Switching frequency problem