UCC3895: 3KW phase shift schematic confirm

Hi Tommy,  an expert will get back to you tomorrow. Thanks. ——Teng

Hello Tommy

This is an interesting and unusual arrangement and I think that it can be made to work. I have no direct experience of this arrangement so I think it would be worth your while modelling it using TINA-TI (free to download at http://www.ti.com/tool/TINA-TI )  You can use some of the existing reference designs at http://www.ti.com/product/UCC3895/toolssoftware as a starting point for your simulation schematic.

I have the following comments on your schematic -

You don't have a shim inductor in series with the PSFB transformer primary windings. This is sometimes needed if the leakage inductance of the transformer is not high enough to force ZVS transitions. If your transformers have enough leakage inductance then you circuit is ok, if not then you must either accept the increased switching losses or add a shim inductor - please have a look at the schematic for the UCC28950 EVM at http://www.ti.com/lit/ug/sluu421a/sluu421a.pdf for an example of how this can be done.

The series connection of the secondaries will force equal current sharing in the primary. IF you can ensure that each primary is identical then the single controller will drive them correctly. You should make the layouts for the three primary circuits as similar as possible.

The current transformer will not work as shown because it sees a bidirectional current - you should use a full wave rectifier on its output so that you get a CS signal on both positive and negative going parts of the switching cycle.

The CT will not give you any protection against shoot through currents on the primary.

The CT will not give you any protection against power stage transformer saturation.

You would need a CT on the input rail to each of the three primary side power stages and use some additional circuits - perhaps a comparator - so that a severe over current event on the primary activates the 2V Current Sense Comparator or the 2.5V Over Current Comparator in the UCC3895.

The optocoupler in the feedback path will limit your control bandwidth - and opto-coupler CTR also degrades over time - this is not necessarily a problem for you I just wanted to note it.

Regards
Colin