솔루션 찾기 기술 지원

SMPS noise sound

Posted by: ni.amir on

Hello

I used TNY280 in 12V-2.2A power supply. In light loads for example in 70mA, it makes annoying noise sound in treble range (high frequency noise sound ).

But in high load the noise sound changes from high frequency to lower freq. that is not very annoying.

How can I remove or decrease this noise? Should I change transformer construction method or feedback or  ...?

With Best Regards

 

 

댓글

Submitted by PI-Radiant on 06/10/2021

Hello Amir

 Thank you for considering our product for your application.

Could you please clarify if the transformer is varnished?

Would it be possible to share with us the Primary drain to source voltage waveform which would help us figure out the switching frequency for some of the test conditions where the noise is felt?

If possible, could you please share more information regarding the design such as magnetizing, leakage inductance?

Could you please let us know if the design was made as per suggestions in PIXls spreadsheet?

 

Regards

PI-Radiant

Hello

I used a Two-component polyester adhesive for varnishing. We immerse the transformers in the polyester liquid three times and in the end we put them in oven for 30 minutes for baking.  Could you please introduce best way and materials for dip varnishing to reduce audible noise of SMPS ??

I attached light load and full load state of drain-source voltage waveforms. In no load or light load or even in full load conditions IC skips the switching cycles and the frequency is decreased. In full load there are no skipping cycle very much but the audible noise is listened.

Also I changed the transformer structure from 1 layer primary to sandwich mode (two layers primary) and the noise decreased but it is not zero.

I checked the snubber capacitors for audible noise problem but they do not contribute in audible noise.

Unfortunately I didn't use the PIXLs but primary inductance is 1230 uH, leakage inductance is about 20 uH and maximum

flux density in the core is 2500 Gauss.

Thank you for helping

 

  

Submitted by PI-Radiant on 06/17/2021

Hello Amir 

 Based on the information you had provided, we believe that the control loop might not be stable. The cathode of the voltage regulator might be reacting considerably fast which might cause the switching pulses to bunch instead of skipping pulses at appropriate intervals as it is expected to. In this case, increasing the value of capacitor connected between cathode of the voltage regulator to ground might help.

We also believe that the flux density for this design might be considerably high which causes the core halves to resonate leading to audible noise. In this case, adding glue between the core halves and/or reducing the flux density might help.

In the previous comment, you had mentioned that you have attached the drain-source voltage waveforms. However, we are not really able to see them. Could you please try this again? Also, could you please share the schematic and input voltage range, which might help us analyze the issue better.

We suggest that you enter the input, output conditions, transformer parameters and other details onto the PI Expert Spreadsheet and evaluate how good the design is and warnings, if any that might be the cause of audible noise.

 

Regards

PI-Radiant

Hello

Thank you for your suggestions.

We use zener diode in feedback loop and I don't see stability problem. I attached the schematic of circuit and drain-source waveforms.

The voltage range is 170VAC - 240VAC but I tested in 220VAC. The transformer core is EC2820 (PC40).

 

첨부 파일 파일 크기
full load (1).png 28.06 KB
full load (2).png 28.06 KB
half load (1).png 21.96 KB
half load (2).png 21.82 KB
no load.png 21.11 KB
PI Expert Online_0.pdf 198.58 KB
Submitted by PI-Radiant on 06/21/2021

Hello Amir

 Thank you for sharing the waveforms. We are able to see the waveforms at Full load, Half load, no load and PI Expert spreadsheet, but not the schematic. Could you please share this with us, if possible?

 

Regards

PI-Radiant

Submitted by PI-Radiant on 06/23/2021

Hello Amir

 As suggested earlier, kindly try adding glue between the core halves.

 

Regards

PI-Radiant

Hello PI-Radiant

I have already used glue for core halves. I think the audible noise is generated from the windings themselves. I don't know exactly.

Thank you for taking the time