strange sound from transformer when load increased

[jegandren]

Dear All,

Attached is a voltage mode control SMPS,

I tested the circuit for 0A to 3 A load at +25V output no problem.

But as i increase the load more that 3A i notice there is a strange sound coming from the EE55 trannsformer.

Is this means that the transformer is running into saturation.

may be if i increase the value of bootstrap capacitor from 330uF/200V to 470uF/200v , would the sound from the EE55 transformer eleminate.

[zilog]

I suspect you are running the transformer into saturation. What Bmax are you aiming for during normal operation? Maybe you have left too little margin to allow for the feedback loop to swing when the input voltage drops due to output load.

[Eva]

Control loop instability or saturation of the output inductors or both.

Compensating the loop with a single pole (100nF capacitor at COMP pin) after the optocoupler is not a good idea, it should be compensated before the opto and a pole-zero is usually required.

Taking a look at the output of the error amplifier with oscilloscope may help in finding optimum compensation values.

[jegandren]

Dear zilog,

Attached is the transformer that im using.

[zilog]

Now after some more inspection, I see that you are only taking feedback from one of the rails. Why not take feedback between +25 and -25 instead of only sensing one of the rails?

Regarding Eva's suggestion, you can use a TL431 to do compensation before the OC, I have attached an article explaining how below.

http://wintermute.csbnet.se/~zilog/f...31_article.pdf

[jegandren]

I took feedback from +25 V to 0V because, i have connected +25V only to the load and for the -25V i did not connect to any load.

[jegandren]

I have tested the SMPS voltage mode control untill 7.2A
( 180W ), and both the mosfet's stay cool and the switching rectifier at secondary part stay warm at 7.2A.

As I said there is some kind of noise or hum sound coming from the switching transformer is now eliminated, the switching transformer is silent now. I notice the hum sound when i tap the oscilloscope probe to both the mosfet's to see the switching pulse from drain to source, when the probe is removed from the mosfet there is no hum sound coming from the switching transformer.

Btw can anyone advise me on what measurements that i should consider for a SMPS.

[Fob]

Increase input caps. I thing they are insuficient.

[jegandren]

Originally posted by zilog
Now after some more inspection, I see that you are only taking feedback from one of the rails. Why not take feedback between +25 and -25 instead of only sensing one of the rails?

Regarding Eva's suggestion, you can use a TL431 to do compensation before the OC, I have attached an article explaining how below.

http://wintermute.csbnet.se/~zilog/f...31_article.pdf

Hi,

I have done the compensation using TL431 before the optocoupler using this article http://wintermute.csbnet.se/~zilog/f...31_article.pdf

but i noticed that when i do the compensation using Cz,Rz and Cp as increase the load the output voltage drops say from 25.0V to 24.5 V .....

but when the Cz,Rz and Cp are removed then as the load is increased i still get the 25.0V.

can anyone advise on how exactly to do the compensation using Cz, Cp and Rz.