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.

TPS22971: load switch reverse current in sequencing application (TPS22971)

Part Number: TPS22971
Other Parts Discussed in Thread: TPS82085, TPS22917

Hello,

I am considering the TPS22971 for an application where I want to use load switches instead of separate regulators to sequence power supplies that use a common voltage. For example, one regulator supplies a common 1.8V rail and the load switches sequence power to several other 1.8V rails. The load currents are low enough (combined with the TPS22971 low Rdson) that there should be no problem meeting the minimum voltage specs for the sequenced rails.

I am concerned about reverse current damaging the load switch. This part has thermal shutdown but not reverse current protection. I assume that the thermal shutdown would not react fast enough to prevent damage if there is large enough reverse current. I am hoping that I can tailor my application to avoid large reverse current.

In my application I can independently control the enables for the load switches (using a microcontroller). In a controlled powerdown scenario I can allow time for the load-switched rails to discharge before disabling the common buck regulator. However if the main power source is removed suddenly then the common regulator output might discharge faster than a given load-switched rail. In that scenario the load switch enable will be negated fairly quickly but I am not sure if this will prevent reverse current.

I can try to minimize the capacitance on the load-switched rails in order to reduce the reverse current duration, but some rails might still need up to 100uF.

Should I change to a load switch with reverse current protection for this type of application, or is my concern not justified? I should mention that board space and power efficiency are primary drivers for part selection. I was using multiple TPS82085 devices in the initial design but started looking at using load switches to save space and power.

Thanks - Jason 

  • Hi Jason,

    TPS22971 has no reverse current protection so any reverse current through the device is possible. The body diode of the device will be biased in such a way that allows for reverse current to flow through it even when the device is turned off if the output is held at a high voltage and the input is 0V. But this is more a concern during DC reverse current operation that in a transient event.

    Assuming both load switches are on, when the common regulator output discharges, the output of the load switch will discharge as well as the rails are connected through the switch. I don't think a large reverse current spike is possible as the nodes are connected to one another. If your system is sensitive to any reverse current, then I believe you will have to use a device with RCP integrated like TPS22917.

    I would suggest you try out your system on two TPS22971 EVMs first as a proof of concept. If the behavior is acceptable here, then I think your system should be fine. Thank you for choosing TI Power Switches for your designs.

  • Thank you for the fast response. I will test with an EVM as you suggest.

    Jason

  • One other point - this part has a quick output discharge feature which might help avoid reverse current if the switch is disabled in time.

    The devices with RCP tend to have Rdson that is too high for my application, so my fallback would be to use regulators instead of switches.

    Jason