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.

TPS25947: TPS25947xx bidirectional current limiting

Part Number: TPS25947

Hello,

we want to implement direction dependant current limiting. For example a battery with max. 1.5 A charging and 4 A discharging. Additionally a signal should allow to disable one direction, whilst still allowing the other.
Is this possible by using two TPS25947 in anti-parallel configuration?
My understanding is that the IC can limit up to 5.5 A, but the 62368-1 certification is only applicable up to 3.96 A continuous. What does "Output Current Limit: 0.575 to 4.84A" mean?

Sincerely
JK

  • Hi JK,

    As per IEC62368-1 standard it will only mention current limit upto 5A. So certification mentions that device can limit current from range 0.575 to 4.84A. 

    I think you want to use like below which is fine. You can replace LM73100 with TPS25947.

    Regards

    Kunal Goel

  • Thank you,

    you are right, a current limiter fulfilling 62368-1 (G.9.1) must not allow currents above 5 A. So for higher currents we have to figure something out.

    Under technical documentation one can find the "TPS25947x CB Test Certificate IEC 62368-1", which gives the following information on page 3:
    Additional Ratings:
    ...
    TPS259460, TPS259461, TPS259470, TPS259472, and TPS259474:
    Input Voltage: 2.7 to 23Vdc
    Output Continuous Rating: 0.425 to 3.96A
    Output Current Limit: 0.575 to 4.84A
    Maximum Operational Temperature: 105°C

    Could you shed some light on this please.
    Does this mean that these ICs are certified according to 62368-1 only up to design currents of 3.96 A?
    Whilst higher current are technically possible, the eFuses would not be considered as 62368-1 compliant?

    The schematic is close to what we want to do, however

    • the connections on the left (Vin and Vout) would also be tied together
    • by default both eFuses would be enabled at the same time. One or both are deactivated under special conditions.

    Is this possible?


    Sincerely
    J K

  • Hi JK,

    Device can support guaranteed continuous current of 5.5A. For IEC62368 certification we have to choose max current limit setting below 5A that is tested in our production . For this device it is 4.84A. For this 4.84A case minimum current limit can be 3.96A which is same as continuous current device can allow at this current limit setting. So certificate mention 3.96A. 

    Agree with you comment on schematic.

    Regards

    Kunal Goel

  • Thank you,

    i'm currently trying to simulate aforementiened setup (two antiparallel TPS25947). It's close to the schematic, but

    • the connections on the left (Vin and Vout) would also be tied together
    • by default both eFuses would be enabled at the same time. One or both are deactivated under special conditions.

    The model file does not appear in PSpice, but i could download it from the website. Unfortunately the simulation is very slow and has convergence problems.
    Can this setup work or is it prone to instability?

  • Hi JK,

    Did you try running single device simulation first? Simulation can have convergence issues as it is a complex device and we are putting two in schematic. We need to adjust simulation parameters. 

    Regards

    Kunal Goel

  • Yes,
    single part simulation with a constant load takes less than two minutes. The auto-converge feature is enabled and the output log advised to set
    PTRANABSTOL = 1e-5 (default 1e-7)
    PTRANVNTOL = 1e-4 (default 1e-6)
    Even after 3 hours the simulation of the anti-parallel configuration didn't finish. We also tried adding 1 Meg pulldowns to the capacitors (ILIM and DVDT pins). In other situations this helped the convergence (reduce gmin). It's quite difficult to make use of PSpice that way.
    Was it a problem on our side or is the TPS25947 really not among the pre-installed models?

  • Hi JK,

    Need to check with modelling expert. I will recommend testing on EVM.

    Regards

    Kunal Goel