• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 259 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS745-Q1: Only use PG function block

Haury Zhou
Expert 6946 points
Part Number: TPS745-Q1

Hi team,

Could you help to check whether the below design is feasible?

I want to use the PG feature for another DC-DC w/o PG, and plan to float the output of TPS745-Q1.

BRs,

Haury

  • 0
    TI__Guru 50025 points

    Hi Haury,

    That's an interesting idea. It would almost work, but I think what will get you is that the DC/DC and LDO will turn on with different timing characteristics, and whichever one turns on first will be the one that gets the TPS745-Q1 output to its set point (I assume that you would configure the TPS745-Q1 device with the same output voltage as the DC/DC). If you can ensure that the TPS745-Q1 is enabled when the DC/DC gets to its set point while also not getting to its set point before the DC/DC, it might work. What I'm not sure of is that if the DC/DC output is slightly above the LDO set point such that the LDO thinks its output voltage is too high and so closes its channel, I'm not sure if it will still keep its PG asserted. Furthermore, connecting the output of power supplies like this will result in one of them basically turning its output off since one of them will inevitably have a higher set point, so if the TPS745-Q1 has a higher set point, the DC/DC would be bypassed and the TPS745-Q1 will supply the load. I think there's enough unknowns to recommend using a comparator or supervisor to do this instead. Or you can use an EVM to test this concept with the TPS745-Q1 to make sure it behaves the way you expect. 

    Regards,

    Nick

  • 0 in reply to Nick Butts
    TI__Expert 6946 points

    Hi Nick,

    Thanks for your support!

    Nick Butts said:
    so if the TPS745-Q1 has a higher set point, the DC/DC would be bypassed and the TPS745-Q1 will supply the load.

    As shown in the diagram, the output of TPS745-Q1 is float. So the situation would not happen.

    Nick Butts said:
    What I'm not sure of is that if the DC/DC output is slightly above the LDO set point such that the LDO thinks its output voltage is too high and so closes its channel, I'm not sure if it will still keep its PG asserted.

    Yes. Thanks for your opinion. I agree with this point should be paid attention.

    Nick Butts said:
    I assume that you would configure the TPS745-Q1 device with the same output voltage as the DC/DC). If you can ensure that the TPS745-Q1 is enabled when the DC/DC gets to its set point while also not getting to its set point before the DC/DC, it might work.

    I don't understand this power-up sequence. Is it necessary that I should enable TPS745-Q1 after DC-DC reaching to the set output? What would happen if I enable LDO and DC-DC in the same time as the the diagram shown which means LDO is enabled before DCDC reaching to the set output?

    Could you help to do a test to make sure this function for TPS745-Q1?My customer is considering use TPS745-Q1 in this application.

    Thanks.

    BRs,

    Haury

  • +1 in reply to Haury Zhou
    TI__Guru 50025 points

    Hi Haury,

    Haury Zhou said:
    As shown in the diagram, the output of TPS745-Q1 is float. So the situation would not happen.

    In the diagram it looks like the TPS745-Q1 output will be tied to the DC/DC output. How else will the TPS745-Q1 PG signal be related to the DC/DC output? Would they just rely on the startup time being similar enough that they can assume that if the TPS745-Q1 output is good the DC/DC output must also be good? 

    Haury Zhou said:
    I don't understand this power-up sequence. Is it necessary that I should enable TPS745-Q1 after DC-DC reaching to the set output? What would happen if I enable LDO and DC-DC in the same time as the the diagram shown which means LDO is enabled before DCDC reaching to the set output?

    What I meant is that if the TPS745-Q1 is enabled and gets to its target before the DC/DC gets to its target, then the PG signal comes early. If the DC/DC gets to its target before the TPS745-Q1 is enabled, the PG signal won't assert. 

    I could do a simple test to see if the PG will stay asserted when there is a voltage applied that is slightly higher than the set voltage, although you mentioned that the TPS745-Q1 output will be floating and not connected to the DC/DC output so I don't know if this kind of test is relevant. I need some clarification on how the PG will be related to the DC/DC output, if at all, to be able to test. If they want to test with the full DC/DC then they should order the EVMs and set up a test themselves. 

    Regards,

    Nick