LMZ14202H: Audible noise from the LMZ1420xH

Hello Emanuel,

The noise you are experiencing can be caused by many reasons, for example, PCB Layout, capacitor selection or the internal inductor itself. There is a high probability that the sound is coming from either the input or output capacitors. Do you have excessive ripple on VOUT? That would be a good starting point to determine it's origin. I also recommend you follow the recommended Layout section in the datasheet.

I've provided below 3 helpful resources on reducing audible techniques:

1. 
2. SLUA821
3. 

Regards,
Alejandro

Sorry for not replying, we had a very busy period after an unfortunate COVID lockdown.

The noise is clearly from the LMZ. We use the LMZ1420xH in all 1,2, and 3A  variants in almost 100 different devices with different PCB.

The frequency is set to about 280KHz to keep the peak in the slope part of the CE102 conducted emissions limits.

There are no large MLCC on the board, the output capacitor C13 is tantalum 47uF.
The other capacitors are C7,C9=1uF, C6,C8=10uF, C10=47uF electrolytic, C23=4.7nF and C14=1uF

The arrangement as shown is following some advise received in another thread, dealing with high frequency emitted EMI
see 

On the present board we followed the instructions and changed the PCB layout, added the small C23 cap near the input and switched the locations of C13 and C14. as in the older designs the tantalum was close to the output.

We had this problem a few years ago on another board and it seems that it is related to the production batch.

Emanuel, that's unfortunate to hear. I hope all gets better.

Did you have any further issues with the conducted/radiated EMI noise from the previous post?
It's difficult to conclude if it is coming from the device. I still believe that the noise is due to the piezoelectric effect. Check out this article that was mentioned in the comments of my 1st recommended resources.

Regards,
Alejandro

Hi Alejandro,

We made a new revision of this specific board to solve some issues in the prototype, among other things we changed the EMI filters layout and added the low value capacitor close to the input.

The board wasn't tested in the EMC lab yet so we don't know if that issue was solved, but this audio noise was immediately notified.

We can check if the reason is piezoelectric effect by removing capacitors, practically most of them are only EMI filters.

I also believe that the sound should be attenuated (or increased) when touching the caps, but this is not happening.

There is however one overlooked factor that I mentioned in my first message:

The LMZ powers several LED strings, each string's current controlled by a TL4242-Q1. The intensity is set by PWM from a microcontroller. 

When set at 100% full intensity there is no noise. This answers the ripple suspicion, as the current in this mode is at maximum, where the highest ripple is expected.

The noise is present when the PWM is at any lower value.

One more thing, the lamp has 2 modes of operation, one with a load of 1.4A (4 LED strings) and the other with 0.35A (one LED string).

The intensity of the noise increases with the load.
We tried to add an additional 47uF tantalum at the output, no change.

One thing is clear, the noise is caused by the PWM, but comes from the LMZ

Emanuel,

Thanks for the detailed information. To make sure I understood correctly, regardless of the mode, at 100% it is creating audible noise, correct?

Can you provide waveforms of the input and output? Do we see any oscillation happening?

Regards,
Alejandro

No, at 100% PWM (no modulation) there is no noise

My apologies Emanuel, that was a typo on my end. I meant to have a "not" in the sentence.

Understood. Is it possible for you to provide any waveforms of the input and output voltage? We would like to see if any oscillation is happening at either ends.

Thanks,
Alejandro