Noisy and spiky output when input voltage is below 95Vac

Posted by: bbmilo on 02/18/2013

Hi,

This is a 12V 2.8A design using TOP267EG. We have passed EMC for the CE market and also some other certifications for safety. Unfortunately we found that the power supply has problem starting up at full load when the input voltage is <95Vac. Our design requirement for minimum startup voltage is 85Vac.

It's found that the output voltage is very noisy, high ripple voltage of ~1V with lots of high spikes at Vin below 95Vac. The lower the Vin the noisier the output voltage.

Can you suggest how can I solve this problem, hopefully without too much change to the circuit?

I would like to send you the schematic and pcb directly to your email address, if required.

Thank you.

Comments

Can you upload your schematic?

thanks.

Hi PI-Cochrane,

Schematic added. Thanks.

I am wondering how the ripple was tested, PLS check the ripple again using the way specified at RDK242. RDK242 is a 12V30A design which is very close to your power level and I believe it is helpful to you. PLS find the link as below:http://www.powerint.com/sites/default/files/PDFFiles/rdr242.pdf

If the problem is really there, PLS check the following list:

1.open D5 to verify if the issue is caused by the soft finish circuit. If it is, you have to modify C11, or try the solution using in RDK242 ( short R6 and remove C11).

2. Use phase boost circuit to parallel with R8, you can start from 100ohm and 100nF.

3. modify c13 from 10uF to 47uf, and you can try 33nF for C12 too.

regards

If the problem is really there, PLS check the following list:

1.open D5 to verify if the issue is caused by the soft finish circuit. If it is, you have to modify C11, or try the solution using in RDK242 ( short R6 and remove C11).

2. Use phase boost circuit to parallel with R8, you can start from 100ohm and 100nF.

3. modify c13 from 10uF to 47uf, and you can try 33nF for C12 too.

regards

Hi Pi-Cochrane,

I've tried all of the above. They didn't affect the voltage spikes pattern at all.

The voltage spikes appear only when the input voltage is below 100Vac (95Vac on some boards). Above that I can't see any spikes at all.

What else could have caused that? I have also tried increasing the bulk capacitor which didn't help either.

Thanks.

The issue maybe caused by the CCM operation at low line, sine the noise is always the worst case at CCM. PLS try the following possible solutions:

1. 2.2Nf Y cap between secondary ground to primary input plus.

2.secondary snubber:R1=50OHM ,c10=220pF.

Regards

Hi PI Cochrane,

Solution 2 above takes out all the lower frequency ripples and some of the higher frequency spikes.

Then the 2.2nF cap takes out even more of the higher frequency.

The power supply operates down to 78 V input at full load now without too big change to the circuit.

It's awsome!! Many thanks.