TPS63070: High Pitch Noise

Hi  Daniel,

Thanks for reaching out, what kind of load? pulsed one with audible frequency? 

Another kind reminder, it is not good to connect Pad and vias by thin trace, you can just pour over to decrease the parasitic inductor.

Regards

Tao

Dear Tao, 

Thank you for your kind response. The load I connected is static and it just a few SMD LEDs, and it's total consumption is about 60mA. It's a audible frequency and  the noise increases when I have my finger placed on the aera marked on the attached picture.





What can be the done to eliminate the high pitch audible noise and what might be the cause behind it?

Additionally, I would appreciate your kind and professional feedback on the overall layout.

Thank you!

Hi Daniel,

i think you can observe an instable switching, right?

Not sure what happened, seems it is caused by layout issue because the Vout sense is not stable enough.

Regards

Tao

Could you please elaborate a bit? What I know is I am having a good and accurate voltage output. 

Hi  Daniel,

I think you can observe a low frequency ripple at output although the RMS voltage is stable and accurate.

Just a suspect, noise coupling into FB node which makes switching instable.

Regards

Tao

Hi Tao,

I am going to insert some waveform I have taken from VOUT, INDUCTOR, AUX, and FB pins. 

DC Analysis
-------------------
OUTPUT

INDUCTOR (Left Side)

INDUCTOR (Right)

AUX

AC Analysis
-------------------

OUT

AUX

IND

FB

By looking at the waveform, what do you think? Do you have any suggestions and tips for improvement?

Thank you! 
Best regards,
Daniel 

The issues at hand involve not only the  persistent high-pitched sound but also the concerning probability of high-frequency/low-frequency noise causing disruptions in our digital communication protocols, specifically I2C, SPI, and UART.

As our project is approaching its deadline, your guidance in identifying the root cause and providing effective solutions would be invaluable. Your immediate attention to this matter is highly appreciated.

Thank you for your time and assistance.

Daniel

Hi  Daniel,

Sorry for delay.

According your figure of output in AC analysis, device work in PFM mode and the switching frequency is close to audable one. 

FPWM mode is acceptable? if so, could you help try with R12 mounted with a zero ohm resister? Anyway, it is worth to replace R12 with a zero ohm resister to find out the audable noise is caused by PFM or not.

Regards

Tao

Hello Tao, 

I've already did that. The good thing, the noise got minimized comparatively, however it still very much exists. And the noise seems to amplify at a range of 6.9 - 7.0V when it's the loudest. Even someone from a few feet away can hear the sound. I'm going to attach few snapshots of the output and inductor here, 

OUTPUT (BUCK/BOOST)
1. 

2.

3.

Inductor
1. 




Peak Noise at ~7V (Readings)
----------------------------------------

OUT (Doesn't seem stable at all): 
1.


Inductor
2. 



Moreover, I've tried to connect the FB Ground node to the control ground with a jumper, that didn't help either. The noise remained as it was.

1. 



At this stage what would be the right thing to do and what seems to be the reasons behind it? Moreover, how do we get a noise-free and audible high pitch free output?

Thank you! 
Best,
Daniel 

Hi  Daniel,

Please help to do following check.

• make sure load is a constant one
• check if there is any low frequency at input, according  the snapshot of Vaux you got, i am afraid there is low frequency noise at input node. But not sure what is root cause of this LF noise, may be it is caused by unstable switching, or just because source is not clear enough.
• Please help to capture the waveform of switching node in one snapshot, then we can check if there is a mode transition during work. please set the oscilloscope at 100us/div or higher.

Regards

Tao

Dear Tao,

Thank you for your message.

1. The load utilized remained constant throughout our testing.

2. I've already attempted transitioning the power input from the PSU to a battery as you suggested, but regrettably, the noise persisted.

Given these observations, I'm open to your guidance on how best to proceed in resolving this issue.

Best regards,
Daniel 

Hello Daniel,

This week, the expert is out of the office. Please await his answer then.

In the meantime, you might check if the noise is gone when you exchange the ceramic capacitors by either physically smaller ones or some of them by some electrolytic one for test purposes only. Ceramic capacitors especially in the bigger packages are known for creating audible noise.

Best regards,
Brigitte

Hi  Daniel,

Tao Zhang said:

Please help to capture the waveform of switching node in one snapshot, then we can check if there is a mode transition during work. please set the oscilloscope at 100us/div or higher.

How about this item, could you helo to capture waveform of both switching node in one snapshot? how many channels in your oscilloscope, one only?

Also, it is better your can help to capture output voltage, set-up of oscilloscope. 20ms/div in horizontal.  DC value with offset. 100mV/div in vertical.

Reagrds

Tao