UCC28C43: ucc28c43

Part Number: UCC28C43

folks 

here is an issue I cannot head my  head around.

You can see in prior mails what circuit I have for UCC28c43 model.

All  app notes on this subject I have used would not lead to correct values for compensation nor of slope compensation. In fact, using some will not yield useful values at all.

Clearly, the circuit is designed for CCM current mode control of a step down stage with input 350 output 270V @ 10 amp.

From some of the app notes, I had gathered that the inductor had to be 625 uH for the  cycle by cycle model to show steady state.

Recently I used SLUP 340 , one out of many, to place fz1, fz2, fp1 & fp2 as indicated.

The transient model  yields decent results. It even works great with Typ III compensation.

Although I admit , I never could get clear cut sense of what slope comp value will be.

But the bode plot I used  absolutely  gave meaningless results. 

I sensed that the small signal model was somehow missing something.

Of course the real circuit did horribly.

Now, lo & behold, someone ran a Bode plot using his/her theory & it looks acceptable. The cycle by cycle in discrete parts also runs well.

But with completely different values.

This person said "vendor models are  not trust worthy.".

While I only exclusively use  TINA-TI for my work.

So, we need to sort this out.

See below picture of what this person used for Bode model.  Gain value is suspect. I also do not know the origin of this calculations. I get a sense of why he used a VCCS. The model does not contain the main inductor. But a leakage of the CMC. 

When I ran the TINA-TI model, it gave acceptable results with much lower inductor. 

So questions

- is this Bode model correct? Then how about the CCM PCM control models in the app notes? How are they so off? I used exact same as in SLVA 301.

- How about the Transient model in TINA-TI? ... is that trust worthy for design?

Any clarification that enhances usefulness of the TI models will be greatly appreciated. Would you know where the Bode model below relationships came from ? Are these trust worthy?

4 Replies