TPS61020: 40kHz internal operation

Dice-K

Part Number: TPS61020

Tool/software:

Hi support team,

A customer reported that the audio signal of their finished product contained tone noise around 40kHz.
After investigating, we found that the power output from the "TPS61020" contained noise components of this frequency when observed with FFT.
The oscillation period of the TPS61020 is 600kHz, but when we placed a probe on the oscillation part (coil) and measured it with FFT, a large 40kHz tone (and 80kHz, which is thought to be a harmonic of it) was confirmed.
The PS pin is set to high and FPWM is used.

So, my questions are:
Are there any parts of the TPS61020's internal operation that operate periodically around 40kHz (1/15 of 600kHz)?
If there are, is there a way to prevent this from happening or to push it to a higher frequency?

I have attached the switching waveform of the coil.

There is some ringing, but it appears to be oscillating without any problems.
(However, there is a change point in the middle of the rise, and this appears to be fluctuating periodically.)

Regards,
Dice-K

8 days ago

0 Travis Wang 8 days ago

TI__Expert 5505 points

Hi Dice,

The expert for this device is out of office. Please wait till tomorrow.

Best Regards,

Travis

0 Nini Zhong 5 days ago in reply to Travis Wang

TI__Expert 7975 points

Hi Dice,

None of the TPS61020's internal operation that operate periodically around 40kHz. As you could see from the switching waveform, the switching frequency is 625kHz. What you said about the SW ring, it is normal behavior of the boost and is in MHz level.

0 Dice-K 5 days ago in reply to Nini Zhong

Mastermind 6827 points

Hi Nini,
Thank you for your reply.

If you have an EVM and FFT analyzer, would it be possible for you to check for the presence or absence of 40 kHz transmission?

Best regards,
Dice-K

0 Nini Zhong 4 days ago in reply to Dice-K

TI__Expert 7975 points

Hi Dice-K,

I have to say that at present I don't have a EVM at hand. But according to the full validation in the past , none of the 40khz oscillation is noticed . By the way, what kind of coil is used, is it an inductor? Could you help give me a whole picture of the schematic or parts exploited?

0 Dice-K 4 days ago in reply to Nini Zhong

Mastermind 6827 points

Hi Nini,
Thank you for your support.

I sent the schematic and other information via private message.

Do you have any opinions on the cause of the 40kHz?

Best regards,
Dice-K

0 Nini Zhong 4 days ago in reply to Dice-K

TI__Expert 7975 points

Up to now , I haven't found any clue for the 40kHz oscillation, maybe you could check other components 's parameters to see if the capacitor or inductor cause the oscillation。

0 Dice-K 3 days ago in reply to Nini Zhong

Mastermind 6827 points

Hi Nini,
Thank you for your reply.

As I wrote in the private message, there is no significant change when I change the inductor from 33μH to 10μH, so I don't think it's inductor dependent.

Even if the TPS61020 is a 40 kHz source, is there no way to suppress it?
Is the only option left to cover it with a noise shield?

0 Nini Zhong 3 days ago in reply to Dice-K

TI__Expert 7975 points

Hi Dice-K,

could you use a smaller inductor , as is recommended in datasheet ? Also could you try external EMI filter to attenuate the oscilaation?

0 Dice-K 2 days ago in reply to Nini Zhong

Mastermind 6827 points

Hi Nini,
Thank you for your support.

The effects of noise can be suppressed by installing LC filters on the input and output sides of the DCDC,
but it cannot be completely eliminated.
As reported earlier, noise is generated even when the DCDC is operated in complete isolation,
so we believe there is also a path for noise to get into the audio circuit via GND.

You said that 40kHz noise was not confirmed in past tests, but have you confirmed it with FFT measurement?
If not, would it be difficult to re-examine it?

Best regards,
Dice-K

0 Nini Zhong 2 days ago in reply to Dice-K

TI__Expert 7975 points

Hi Dice-K,

I could check if we could do the FFT about the Vout waveform in the lab. I could upate later to you.

0 Dice-K 1 day ago in reply to Nini Zhong

Mastermind 6827 points

Hi Nini,
Thank you for your help.

As I mentioned in my private message, if they probe the SW pin and do an FFT measurement they can see a component around 40kHz.

Best regards,
Dice-K

0 Nini Zhong 1 day ago in reply to Dice-K

TI__Expert 7975 points

Hi Dice-K,

I will update when this is done.

Power management

Power management forum

TPS61020: 40kHz internal operation