TOPSwitch Multiple Outputs Poor Cross Regulation

Hello juinbeh,

You need to improve cross regulation. There are a few examples you can check here:
http://www.powerint.com/en/design-support/reference-designs/design-examples
Just select one close to your need, for example:
http://www.powerint.com/sites/default/files/PDFFiles/der31.pdf

Cheers,
Adi

Hi Adi,

Thanks for your reply. I looked at several examples for multiple outputs but I couldn't find any with floating configuration. der-31 is for stacked configuration and so is der-218 and didn't discuss anything about cross regulation.
Out of all the examples I also couldn't find one that both outputs with similar power requirement as what I require (55W from both outputs). Most of them has one high power output which is used as the monitored output and the others would be comparatively very low power.

Please help.

Regards,
Juin

Hello juinbeh,

One quick test for you is to use coupled output filter inductors. For regulatory compliance use triple insulated wire if necessary.
The second test is to add adjustable preloads using zenner diodes and resistors: for example, 13V diodes with 50R.
A third choice is to double the number of feedback loops and wire the optocouplers in parallel on the primary side.
And a fourth option is to add a low voltage dropout linear regulator, at least for the output without feedback.

Cheers,
PI_Crusher

Hello Crusher,

Can you please explain more on the 3rd choice?
I don't see how the 4th option is going to work. When it's loaded and the other output not loaded, the output without feedback reduced to ~10V. Adding a LDO regulator is just going to make it turn off all together right?

However I'm going to give option 2 a try now. Thanks.

Regards,
Juin

Hello juinbeh,

With second choice you just have to double the feedback loop, one for each output. On the primary side you have to connect the optocouplers together with some weighting resistors since your special isolated outputs does not allow you to implement this technique on the secondary side.
You identified correctly what change you need using an LDO. This gives you perfect regulation on both outputs with the obvious requirement to increase the rectified voltage on the one output with extra regulation and no feedback, at least 3.5v according to your measurements. Also you have to remove/reduce some output preloads and you need no other cross regulation technique.

Cheers,
PI_Crusher

Hello Crusher,

Can you please sketch a schematic on the doubling the feedback loop idea? or give a DER that uses this idea and the LDO idea.

Thanks,
Juin

Hello Crusher,

Had a good thought of some of your suggestions.
I can't see how the zener diode option would work. By having a series resistor of that value (10s of ohms), there won't be any output voltage remains with higher load when the current drawn is high!
The doubling the feedback loop idea... seems to me that it would only introduce a new problem where while trying to generate enough energy for Output2, Output1 would reach the desire voltage which is now also fedback and the controller IC would then stop cranking up the voltage. As a result Output2 would have voltage lower than the desired value. Is it not?
It seems that it's almost the nature of the flyback SMPS that it's quite hard to have two isolated output with similar power requirement to have good cross regulation.
So may be using the LDO is the way to go. I'll give it a try. In the mean time let me know if you think of better solution which would give better efficiency.

Cheers,
Juin

Hello again Crusher,

Looking forward to see your comment on the little conclusion I had above.

By the way, what value of coupled inductor should I use if I were to try that option?

Thanks,
Juin