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.

TPS23861: Some IC did not work anymore after a few days in the field part2

Part Number: TPS23861
Other Parts Discussed in Thread: TPS7A4001,

hi 

we already talked about that problem here:

but we have some devices with older layout/schematic where we use exactly the same 3.3V LDO as you do. the TPS7A4001.

And with this layout the TPS23861 will also broke some times.

VPWR is in this case delivered by a meanwell EPS-45S-48 and the step-up U1 is not used.

  • Hi User,

    The schematics are different in the other E2E post you reference above. There should be a 10k Pull up resistor from RESET to VDD instead of a short. This is why the power up sequence solution worked in the last E2E thread. For this design, the power up sequence is correct but the schematic shows VDD and VPWR are connected together which should not be the case. RESET should stay LOW during startup. 

    Since there is an internal pull in the IC already, can you try cutting the RESET trace to VDD on the PCB? Note you still have to replace the TPS23861 IC before trying power up again since it could be damaged already. Thanks!

  • Hi Darwin

    For this design, the power up sequence is correct but the schematic shows VDD and VPWR are connected together which should not be the case.

    Where is it are the VDD and VPWR connected together? can you pls draw it in the schematics?

    Since there is an internal pull in the IC already, can you try cutting the RESET trace to VDD on the PCB? Note you still have to replace the TPS23861 IC before trying power up again since it could be damaged already.

    I will cut the RESET Pin from the IC during solder in...

    I'm a bit disappointed / surprised, there are 3 different designs from TI and all have the RESET pin directly to VDD!? Under "reference design" I understand something else ...

    Thank you for helping me!

    I will post here in a few minutes my design for the new order, so it would be nice if you could have an eye on it :-)

    BR Dominik

  • Hi User,

    Very sorry, I typed this incorrectly. I meant to say RESET and VDD should not be connected together shown on U3 in the schematic above. Can you take a waveform of RESET and VPWR during power up? I'd like to make sure the start up sequence is being met. Thanks!

  • This is form the layout where VDD is connected to RESET direct.

    Yello is VDD/REST

    Blue is VPWR

  • Thanks for the waveform user. Do you get the same waveform for your other board? The reason I ask is because it looks like VDD ramps up before VPWR goes above UVLO. Also, when I look at the divider on the LDO, it should enable at ~40V but the waveform shows 3.3V rail fully turning on before this.

  • Hi

    i went now with R39 down to 22k and also add a 100n cap.

    now it looks like that:

    what you  think?

    Also, when I look at the divider on the LDO, it should enable at ~40V but the waveform shows 3.3V rail fully turning on before this.

    Yes thats what i though too but in reality its not the case.

  • Hi User, does your working solution (with pull resistor on RESET) also have the RESET pin going up like shown below in red? It looks like when VPWR is 10V, RESET/VDD goes to >1V during startup. This might be an issue since RESET needs to be <0.9V during VPWR ramp up.

  • sorry i cant see your picture?!

  • Sorry it does not look like it loaded. Please see below. Thanks!

  • Blue RESET Pin (pullup with 10k to VDD)

    Red VDD

    Green VPWR

    Yes i see what you mean, this small peak... but why?

  • This seems it might be the LDO. It could be an inrush issue while the output cap tries to charge up. Can you increase the 2.2uF input cap to a high ceramic input cap like 22uF? Just to prove it out. Thanks!

  • Hi

    I did you modification with 22uF but still the same.

    but i did an other test with 48V not from step (U1) up but from AC/DC 48V supply. There it looks good.

    therefore the problem should come from the edge/signal of the step up during start.

    No my idea for next week is to add on point (1) a N Channel Mosfet.

    with a under over voltage circuit. the idea is to "turn on" if 48V is stable and with a short delay. 

  • Thanks for checking! Glad you were able to narrow down to the boost converter. When the waveform is available with the new circuit, please share with me and I can review. Thanks!