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.

LM5117-Q1: Load Sharing

Part Number: LM5117-Q1
Other Parts Discussed in Thread: LM5117, , LM5119

Hello,

For our application, we have a PCB with two 100W synchronous buck converters designed with the LM5117 (we chose this based on the recommendation in TI Webench).

We would like the option to jumper the outputs of the two bucks to create a single 200W supply. Is this possible? Does it present stability problems or other issues I should be aware of?

Thank you,

Dan

  • While I do not support LM5117-Q1 directly, I can make some comments.

    Generally speaking, you cannot just tie the two outputs together. Usually the load currents will not be balanced. One converter will tend to run towards current limit. There are possible work arounds for this. I have seen applications such as:

    www.ti.com/.../slva389

    In this case, the two converters are paralleled by synchronizing the clocks and tying the COMP, VSENSE, EN and SS pins together. I have to say that this is really more of an academic exercise. In practical terms, the design was very finicky especially with regard to layout.

    Some other ICs are designed for parallel operation and are stackable and interleaved. An example would be www.ti.com/.../TPS40180

    You could also consider an external current share controller such as:
    www.ti.com/.../UCC29002

    Hope this helps to give you some ideas.
  • John is correct; you need a controller that has been design for parallel operation, a controller you may want to look at is the LM5119. The LM5119 has a master/slave mode, where channel 2 error amplifier can be disabling; this places it in a high impedance state. You can then connect COMP1 to COMP2, this will allow current sharing between the two channels.
    Terry
  • Thank you both for your responses. I am taking a look at the LM5119. At a glance, it looks like it makes much more sense than the LM5117 for this application. I will have to play with it for a bit.

    Thanks,
    Dan
  • So I have actually attempted to tie the outputs of the LM5117 circuit together, and it appears to work quite nicely. I am still not entirely comfortable with the circuit, but I would like to know if there are some specific test cases in mind that could break it?

    Thanks,

    Dan