UCC28950: Designing 960W Phase Shifted Full Bridge with ZVS

Hi Colin,

Thank you so much for your response. To answer your question first - I chose 140KHz because, as you have guessed correctly, I wanted to keep fundamental frequency below 150KHz. However, my magnetics supplier had suggested me 200KHz as well, in a quest to reduce the transformer heating because with 200KHz, the number of turns could be lesser and the wire size could be bigger to reduce the copper losses. I am still undecided which one to use. What do you suggest ?

On the shim inductor : Can the leakage inductance of the main transformer be considered as shim inductor ? I could do away with an additional shim inductor in that case.

Unfortunately, my magnetics supplier is not very well versed with this PSFB topology (He is an expert in LLC) hence I am seeking TI's guidance here. I too don't have hands on experience on this topology.

Regards,

Sanjay

Hello Sanjay

I'd suggest that you use the UCC28951 instead of the UCC28950. The UCC28951 is a newer drop in replacement device that works better when operating simultaneously at Dmax and in current limit.

The easy question first: yes, you can certainly use the leakage inductance of the transformer and if the leakage inductance is large enough you will be able to eliminate the shim inductance. One of the features of the LLC is that it can use transformers with large leakage inducatnce to form part of the resonant circuit. the PSFB is also relatively intolerant of leakage inductance and I'd expect that your magnetics manufacturer can design a transformer for your application with a well controlled amount of leakage.

Most of the applications using the UCC28951 run at 100kHz although some customers are running it at 200kHz so you should not have any problems with the controller if you decide to use 200kHz. In fact we have at least one reference design using this part running at 300kHz (PMP20657).

A higher switching frequency will allow you to use a smaller transformer core. Core losses per unit volume will increase but will be offset by a smaller core volume. Copper losses due to proximity losses and skin effect will increase as the switching frequency increases too. These can be offset by using litz wire and other techniques.

I'd suggest that unless you have a requirement to minimise the design size that you keep to a 100kHz switching frequency. Note of course that even if you manage to significantly reduce the core size, you may not get much of a reduction in the overall product size because the switching MOSFETs, control circuit, heatsinking and so on won't get smaller.

Please let me know if you need any further information.

Hi Colin,

Thank you for your suggestions. In fact, I have already ordered samples of UCC28950. Let me fire the first proto using it. May be later on I can changeover to UCC28951 !

One quick question on the switching frequency - When you say that switching frequency is 100KHz, it means that Mosfets in the bridge are switching at 50KHz, right ?

Moreover, I tried using the excel calculator. I am not very sure why there is "measured leakage inductance" before it is actually calculated ? I mean, how can one measure leakage before the transformer is actually built ? 

The excel calculator no doubt helps to determine the electrical parameters like inductance value of inductors but what I am specifically looking for is to determine/ select the correct core.

Regards,

Sanjay

Hi Sanjay

First, when I say 100kHz switching frequency I mean the switching frequency at the MOSFET bridge and this is the frequency calculated by the equations in the data sheet (eq10 and eq11). The output inductor will see a switching rate of twice this or 200kHz.

The use of the term 'Measured Transformer Primary Leakage Inductance' is a bit misleading. You should put the expected leakage inductance here, or put in the amount of leakage inductance required so that the 'Calculated Shim Inductance' result becomes 0. Once you have a physical transformer in your hand you should measure it to see if it meets the designed value. Here is a link to a document that gives some more details about how the ZVS behaviour is affected by the inductances./cfs-file/__key/communityserver-discussions-components-files/196/6355.ZVS_5F00_SubSet.docx

Transformer design is not trivial and I'm afraid I won't be able to help you very much. However you might want to look at the document at

https://www.ti.com/seclit/ml/slup205/slup205.pdf

and the TI Magnetics Design Handbook by Lloyd Dixon at https://www.ti.com/seclit/ml/slup132/slup132.pdf

(the section on fractional turns should be treated with caution but you won't need fractional turns in any case)

If you are looking for a text book then I'd recommend 'Transformers and Inductors for Power Electronics: Theory, Design and Applications' by W.G. Hurley, W.H. Wölfle: Published by Wiley ISBN: 978-1-119-95057-8

Regards
Colin

Dear Colin,

Thank you for your reply and apologies for not responding immediately as I was not in office.

We have begun drawing schematic diagram of the project which I wish to have reviewed by you and your team. However, I am not sure how do I share the schematic with you ? I would not be able to share it on the forum because of our company policy.

The schematic drawing is likely to get completed by this weekend. Going forward, I would also like to share the PCB layout when it will be done for your expert comments.

Please advise how do you want me to share the files ? I would prefer sharing on email, if you say so. In that case, please advise your mail id.

Thanks and regards,

Sanjay

Hello Sanjay

I'm glad to hear that you are making progress. Yes, I'd be happy to review your schematic, please send them to me directly at colingillmor@ti.com

I'll also be happy to review your PCB too and you can send them to the same address - I can read a wide variety of native files (Altium, PADS etc) or Gerber files.

Regards

Colin