This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC2897A: Issues with low voltage at 30A

Part Number: UCC2897A
Other Parts Discussed in Thread: UCC24612

Team,

I have a customer designing with UCC2897A to build an isolated 2.7V 30A converter working on the bench relatively quickly. He started with the EVM, which outputs 3.3V@30A. The intention was to modify the EVM to function in constant current mode, and then switch to constant voltage when 2.7V was attained at the output. They had no problems designing the crossover circuit.

They then realized that their load might drag the output voltage below 2V at times, and realized that they hadn't considered the gate drive requirements of the synchronous rectifier MOSFETs. (Below 2V, there might be insufficient gate drive for these SR's and since this design features self-driven SR's.)  

Their first thought was to swap the SR MOSFETs for low-gate threshold devices, and they tried this, but the converter failed after swapping out the fets, and they then realized that there was probably a simultaneous conduction path (short circuit) on the output.  

 Question(s):

Do you think it is possible to build a small piggyback board to modify the 2897a EVM's for operation down to 1.5V???  Their thoughts are to design a direct drive circuit as follows.  (Two options.):

A.

  • Utilize waveforms from the 2897 (either the clock/oscillator or output pins).
  • Build a piggyback circuit using a MOS driver chip to drive pulse transformers in order to achieve isolated gate drive signals for the SR MOSFET's.
  • Cut the gate pin traces from the board layout, isolating these pins from the original circuit.
  • Feed the SR's from the pulse transformers on my piggyback circuit with jumper wires.
  • This should acheive operation all the way down to near zero volts on the output, correct?

OR...

B.

The 2897a EVM's were not the right choice from the beginning, and thus it would be prudent / best to scrap the boards altogether and start over with a board that already incorporates direct driven SR's.

  • Hi Carolus,
    Thank you for the detailed question about your issue and your description of the extensive testing that was previously done on the EVM.
    As you discovered, it is not an easy issue to change the EVM output voltage from 3.3V to a constant current 30A and constant voltage 1.5V (CC/CV) output .
    These are the design issues involved:
    (1) In CC mode the output voltage can be zero.
    The EVM uses Vout to generate a P_BIAS signal to power the UCC2897A. This means you need some other method to generate P_BIAS.
    (2) Synchronous rectifiers on the secondary will probably not have enough voltage to turn on .
    This is true whether you use self driven or SR controller driven like the UCC24612 family of parts.
    This is what I suggest as ideas for your design:
    *Use the OUT signal of the UCC2897A to drive the forward synchronous rectifier Q3,Q4. This needs a gate driver transformer for isolation
    *Use the inverted AUX signal to drive the freewheeling rectifier Q5.Q7. This requires another gate drive transformer.
    * You need two feedback loops. A constant current feedback and a constant voltage feedback.
    These two feedbacks are diode ORed to drive U2 opto coupler.
    * You will need to power the UCC2897A from a dedicated bias supply .

    The whole project is beyond the scope of this forum but these ideas may be of help.
    Your option A is a good route to investigate
    I think the UCC2897A with synchronous rectification is a good controller to use and it is a good design challenge.
    Regards
    John
  • Hi Carolus,
    There has been no new thread on this post in over 1 week so I will close the post.
    Please open a new thread if you have further issues.
    Regards
    John