• State Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 498 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.

Original question:

TPS25947EVM: Thermal performance

TPS25947: Start up issue

Vincent Chen64988
Expert 4480 points
Part Number: TPS25947

Hi Team,

My customer is evaluating TPS25947 EVM, since R&D will use TPS259474 so he tested the functions on CH4. When I visited customer and we found a strange phenomenon during soft-start period and need comments for the issue we found.

The EVM setting is dVdt pin=3300pF, ITIMER=220pF, ILM=1.65Kohm(measured threshold is about 2.048A after soft-start), PGTH=5.192V(R29=499Kom, R36=150Kohm),  electronic load is set at CR mode. Input voltage is about 5.56V. We didn't change input/output capacitance.

The question is why the PG keeps at low level after sort-start even Vout voltage is large than PGTH(5.192V).

1. Scenario 1, CR is set at 2.862ohm.

After Vout cap. is fully charged, Iout raised to 1.926A and Vout was about 5.55V that was large than PGTH. But why PG pin was still at low level?

When we disable electronic load, Iout reduced to 0A and PG signal can raise to high level. Enable electronic load again, PG was still at high level. Iout raised to 1..926A again. Please help to commend is it a normal behavior? 

2. Scenario 2, CR is set at 2.067ohm.(the Iout current should be about 2.69A)

After Vout cap. is fully charged, Iout raised and limited at 2.076A and Vout just raised to 5.42V but it still large than PGTH. But why PG pin was still at low level?

When we disabled electronic load, Iout reduced to 0A and PG signal can raise to high level. Enable electronic load again, PG was at high level just for a while meanwhile Iout raised to 2.69A. TPS259474 entered into circuit breaker after ITIMER was expired

The PG behavior here is similar to scenario 1 during soft-start period.

3. If above behavior is expected. Why the clamped current is 1.926A in scenario 1 but not 2.048A? Why there is a about 100mA error? 

Thanks a lot.

Vincent Chen

  • 0
    TI__Guru* 79181 points

    Hi Vincent ,

    Thanks for reaching out! I will get back to you by tomorrow.

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Expert 4480 points

    Hi Kunal,

    OK, thanks very much for your support.

    BR,

    Vincent Chen

  • 0 in reply to Kunal Goel
    TI__Guru* 79181 points

    Hi Vincent,

    It is unexpected. Lets try to debug step by step:

    1. Please try to start device with no load connected at output and then observe PG

    2. In the waveforms you shared it looks that eload is slowly turning on while VOUT is still rising from zero. Looks like it is not turned on after VOUT cap is fully charged . So first let output cap fully charged and then turn on e load. Is eload automatically turning on ,based on some threshold or manually you are enabling or disabling it?

    3. What is the setting of jumper J17? 

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Expert 4480 points

    Hi Kunal,

    Please find my reply below.

    1. We had tested it. With no load(eload is off) the Vout can be established to close to Vin voltage and PG raised to high level. I didn't capture the waveform but you can refer to below one.

    The eload was set at CR mode and sink current was about 1.806A which is under current limited threshold. PG signal raised to high level normally.

      

    2. We enabled eload manually before power supply was turned on. Due to we configure eload to CR mode, that is why you found the Iout raised slowly and followed Vout's raising slew rate proportionally.

    3. Jumper J17 setting is we put a jumper on pin2 and pin 3, PG pin was pulled up to U6's Vout.

    Thanks a lot.

    Vincent Chen

  • 0 in reply to Vincent Chen64988
    TI__Guru* 79181 points

    Hi Vincent ,

    When final load current is below a particular value during startup then PG is asserted. But when final  load current during startup crosses a  threshold value then PG is not asserted? Is the threshold value around current limit? What I mean is that in the above waveform when current is around 1.8 then PG is asserted. But when we apply higher current such as 1.92 then PG is not asserted?

    Can you try with pure resistive load and not eload? This will help us know if eload is causing any issue

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Expert 4480 points

    Hi Kunal,

    When final load current is below a particular value during startup then PG is asserted. But when final  load current during startup crosses a  threshold value then PG is not asserted? 

    --> Yes.

    Is the threshold value around current limit?

    --> Almost same to current limit but the threshold current is under limited current 100mA. We expect actual threshold current should be equal to limited current.

    Customer will do the experiment again with real resistor, I will get back you the result.

    Thanks a lot.

    Vincent Chen

     

  • 0 in reply to Vincent Chen64988
    TI__Guru* 79181 points

    Hi Vincent,

    There can be some variation in current limit threshold due to PVT. As you can see we have mentioned in datasheet the variation:

    Lets see the results with resistive load.

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Expert 4480 points

    Hi Kunal,

    Yes, there is variation in current limit threshold but I think it is not in this case.

    Please find below waveform, we tried to increase load current step by step, then we can find the threshold of current limit. It is about 2.048A.

    Besides, power on efuse without eload enabled then enable load to sink 1.926A the PG signal still maintain at high level.

    As to the experiment of changing to pure resistive load, this issue is still there. Please find my explanation below.

    Changing the current limit threshold to 2.304A(measured value).

    1. Set eload to sink 2.195A and enabling eload before turn on TPS259474. PG signal cannot raise to high level after soft-start, sink current keeps at 2.195A. TPS259474 not shutdown.

    2. Change load to a real resistor, the sink current is about 2.203A. the PG signal cannot raise to high level too.

    I think are you be able to setup same environment to replicate this issue? it would be helpful to understand this issue quickly.

    Thanks a lot.

    Vincent Chen

  • +1 in reply to Vincent Chen64988
    TI__Guru* 79181 points

    Hi Vincent,

    I was not able to access lab past week due to covid restrictions. I will try to test in lab next week. What I understand is that:

    1. When load is enabled before turn on and when current is greater than current limit then after VOUT is fully charged ,PG does not go up and (current>current limit) passes through device as it is without any limiting?

    2. Results are same with resistive load and eload in CR mode?

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Expert 4480 points

    Hi Kunal,

    1. There are two conditions and we can take the first picture above as an example. a) The current limit threshold is 2.048A, if the current is above 1.926A(about threshold current -100mA) and under the current limit threshold. PG does not go up. b) When load current larger than current limit threshold, say 3A for example, the eFuse will limit current to threshold current and Vout drop.

    2. Yes. Customer tested it with both resistive load and eload in CR mode. He can find completely the same result. PG won't go up when load current is about or above limit current-100mA.

    Thanks a lot.

    Vincent Chen

  • 0 in reply to Vincent Chen64988
    TI__Guru* 79181 points

    Hi Vincent,

    Thanks for explanation. I will  get back to you asap.

    Regards

    Kunal Goel

  • 0 in reply to Kunal Goel
    TI__Guru* 79181 points

    Hi Vincent,

    Can you accept my friend request? We can follow up on private messaging.

    Regards

    Kunal Goel

  • 0 in reply to Vincent Chen64988
    TI__Mastermind 39550 points

    Hi Vincent,

    As Kunal is out of office, I'm taking care this thread. We will discuss the issue through private messaging and hope to solve as soon as possible.