UCC2897A: ucc2897

robin gangopadhya18

Mastermind 6296 points

Part Number: UCC2897A
Other Parts Discussed in Thread: PMP31092, UCC2808A-2, LM25037

Tool/software:

Need a fast resolution: can we do a step up converter using ACTIVE CLAMP FORWARD controllers?

Input: 22 to 32VC DC

or can have it boosted to 200V using another TI REF Design module such as AC Flyback- which has isolation.

Wont really want to have 2 isolation transformers.!

Output: 100v

power : 300 W

Since there is a transformer, is there anything topologically impossible with 1:N step up conversion? Apart from the PMOS, will be using all GaN devices.

I tried to use "Power Stage" designer, it just would not take these inputs and outputs!

Will appreciate a fast resolution of this..-

6 months ago

0 Hong Huang 6 months ago

TI__Guru 74115 points

Hi,

If you design the transformer turns ratio Np/Ns < 1/5, it looks you can get 100V output from 22V input, so 1:N (N > 1) is ok. If you do not use P-channel MOSFET, but N-channel, then you need to put the clamp MOSFET on high side.

You can use the below application note to make design if Power Stage Designer Tool does not work.

www.ti.com/.../slua535

0 robin gangopadhya18 6 months ago in reply to Hong Huang

Mastermind 6296 points

Hello Hong:

A must thank you for taking your time and putting in this effort to review our requirements and suggest a viable solution.

I will look into UCC2897...if they have a TINATI model that will be super cool.

LMK

0 Hong Huang 6 months ago in reply to robin gangopadhya18

TI__Guru 74115 points

Hi,

Yes, UCC2897A has TINA TI models.

UCC2897A data sheet, product information and support | TI.com

0 robin gangopadhya18 6 months ago in reply to Hong Huang

Mastermind 6296 points

Hello Hong:

I found 2891 although I will check out 2897.

Very interesting.

I can run it at 800 kHz because of the GaN devices. The PMOS can be there - simpler. Maybe bigger in size.

The rough situation is the xmfr primary side peak current!

A whopping 43 amps!...with 800 kHz, imagine Lizt wires needed...

But still it is worth looking into.

thnx

0 robin gangopadhya18 6 months ago in reply to Hong Huang

Mastermind 6296 points

UIGH!

SLUC662 spreadsheet has something gooffy..

With 800 kHz switching, it puts out 94 uH inductor!, 100V 30% ripple

Imagine litz wire bulk...LOL

Some typo there line 37

Let me explore this further though.

Will get an EVALKIT if it looks doable.

appreciate...

0 robin gangopadhya18 6 months ago in reply to Hong Huang

Mastermind 6296 points

SOME PLACE SOME TIME AGO YOU HAD SUGGESTED 2897 A BETTER CHOICE FOR SIMILAR APP..

I ran the xl...

shock: the xmfr pri peak current is 109 AMP.

it is not practical. For 500 kHz switching magnetics..

See attached.

r5751.calculations forward converter.xlsx

Ignore the secondary "red " area...We will be using GaN...if this topology is possible.

109 AMP!

No way....

0 robin gangopadhya18 6 months ago

Mastermind 6296 points

Hello Hong:

Seems PMP31092 ref design-select rev 8 to get 200 v range output- with a step down buck is a better practical scheme.

I wrote a detailed msg...got lost!

I am writing the short repeat observation here...comment pl.

0 Hong Huang 6 months ago in reply to robin gangopadhya18

TI__Guru 74115 points

Hi,

Push-pull is better for 22V to 32V input and 300W power. UCC2808A-2, or LM25037 is ok to use.

Power management

Power management forum

UCC2897A: ucc2897