PIExpert obtains Vbias=70V for a Design with Universal Input, LNK363, E-Shield transformer, Feedback Bias Winding

Posted by: cabedikes on 05/07/2009

Hi,
I have designed a SMPS, using PIExpert, with following requirements :
- Universal Input
- 10V/250mA and 20V/40mA Outputs
- LinkSwitch Family
- E-Shield transformer
- Feedback by Bias Winding
Applying this values, PIExpert calculates Vbias=70.3V I have not any troubles on using this excessively (I think) high voltage on my design. But, why is so high Vbias? I'd like to understand the reason. Does the E-Shield construction of the transformer condition the number of turns of Feedback Bias Winding?

Thank you very much for your assistance!

Comments

LNK363 is LinkSwitch-XT family. PI Expert design for your spec returns about 27 V on the bias winidng if you use Bias winding feedback.
I guess I am a,little confusd about what exactly your problem is, as I cant reproduce it. Would you elaborate or send me a PM.

What does PM mean?

I send you the PIExpert report, converted to .pdf file, of my design. Here you can see VB = 70.3V ('Bias Variables' window)

If it helps, I could send you a .doc file with the 'Print Screen' of PIExpert Specifications forms, Design forms, Input Stage forms ...
The size of this file is approx. 1700Kbytes

Thank you very much

I cannot view your attachment. Can you try to attach it again or just send me a PM?

Oops sorry. PM is private message. Just click my alias (Spock) and then select the "Contact tab" and type a private message and attach the file.

I have tried a private message, but I haven't saw any tabs for your account. I send my the attachment to you again. I hope that it functions this time.

My problem is the following :
PIExpert calculates for my Design (with E-Shield transformer) Vbias about 70V. I think PIExpert is wrong because of PIExpert calculates, for a similar design, without E-Shield transformer, Vbias about 27V.

I'm interested in components with lower working voltage because of they are more cheap.

Is PIExpert wrong?

Thank you

Ok I can open this file. I can site several reasons why this is happening
1) for bias winding feedback - to provide some shielding the bias winding is wound as the first winding in the transformer.
2) Your problem arises because you have only 3 W design and you are using a core that is capable of delivering 13 W. As a result the primary winding only has 1 layer. This causes a large bias winding voltage. If you have peimary winding in 2 or 3 layers (which is what is normal for designs at this power level) you will see a lower bias wionidng voltage.
There are 2 solutions
1) Ideally, use a smaller core like EE13 or EF12.6
2) Duplicate the same design in PIXls and manually change the bias winding turns to a lower number (such that you see about 15 V on bias winding) This will disable the shielding effects of the bias winding but will work for you.

I understand

Thank you very much!