UCC2819A: About abnormal noise (sound)

Hi Masazumi-san,

Thanks for your interest in UCC2819A. The noise is likely coming from the transformer. In the NG case, you can see some very narrow on times of the FET and there is an abrupt change in the FET on time from cycle to cycle. I would suggest looking at MOUT and CAOUT to make sure noise is not coupling into these signals and corrupting the switching logic.

Best Regards,
Ben Lough

Hello.

thank you for your answer.

Waveforms for the following PINs were taken.
IC-5PIN（MOUT)
IC-3PIN（CAOUT)
IC-14PIN（CT)
Vgs

Please see the waveform of MOUT.
When comparing the waveforms of OK_IC and NG_IC, MOUT_PIN seems to have more noise on the NG_IC side.
Is this the cause of the malfunction?

Also, we changed the IC only using the same board (all parts etc. are the same).
Why does noise appear in this way?

Thank you.

UCC2819A0904.pdf

Hi Masazumi-san,

Please check VAOUT (pin 7) during startup. If the output of the error amplifier is falling below 0.33V, the controller may go from low duty cycle to full duty cycle.

Best Regards,
Ben Lough

Hello.

thank you for your answer.

When looking at the operation of VAOUTPIN, IC input PIN,

Due to malfunction of NG_IC, switching noise (surge voltage of rising and falling of inverter Vds) is larger than OK_IC because inverter of PFC circuit of power supply was performing hard switching.

(Even in the NG_IC waveform of the attached document, when looking at the Vds waveform

I think that you understand. )

So switching noise is on 5 pin 13 pin of NG_IC. In case

I think that it is not malfunction due to noise.

How?

Thank you.

UCC2819A-3.pdf

Hi Masazumi-san,

How many ICs do you see this kind of behavior?

Best Regards,
Ben Lough

Hi.

Ben-san

I delayed to reply.

Other ICs with similar phenomena have appeared.
We use UCC2819A also for old products and new products, but we have about 30 items in total.
It has occurred since around 2008.
I am investigating the cause. (For revision)

Please see the attached document for the question.

Thank you.1513.UCC2819A-3.pdf

Hi Masazumi-san,

Try reducing the bandwidth of CAOUT. The design might be only marginally stable.

Best Regards,
Ben Lough

Hello.

Thanks for the advice.

I would like to do it after tomorrow.

Also, I changed the monitor pin and conditions and took the waveform.

It is in the attached document.

Please advise if there are other causes.

Thank you.

UCC2819A-4.pdf