[EE] Dual Power Input Circuit

[Josh Koffman]

Hi all,

I am working on a project that will connect via USB to two computers.
In regular operation either or both of the ports will have power
coming through them. If only one is powered, regardless of which one
it is, the entire circuit needs to use the same power.

My current plan is to send the incoming +5V line from each port
through a selector/combiner. I could do this with a diode or, but I
don't really want that voltage drop. I have found the TI TPS2115A
(https://www.digikey.ca/en/products/detail/texas-instruments/TPS2115APWR/652785
) which I believe would also work.

I'm curious if there's another solution using discrete MOSFETs (or
anything really) that might be a better option.

Thanks!

Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
        -Douglas Adams
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Brent Brown-2]

On 14 Jun 2020 at 15:41, Josh Koffman wrote:

> Hi all,
>
> I am working on a project that will connect via USB to two computers.
> In regular operation either or both of the ports will have power
> coming through them. If only one is powered, regardless of which one
> it is, the entire circuit needs to use the same power.
>
> My current plan is to send the incoming +5V line from each port
> through a selector/combiner. I could do this with a diode or, but I
> don't really want that voltage drop. I have found the TI TPS2115A
> (https://www.digikey.ca/en/products/detail/texas-instruments/TPS2115APWR/652785
> ) which I believe would also work.
>
> I'm curious if there's another solution using discrete MOSFETs (or
> anything really) that might be a better option.

Hi Josh,

The functional block diagram in the TI datasheet shows how they chose do it... In
essence 2 x internal N-Ch MOSFETs. First off it requires a charge pump to get high
side drive. Then you start considering protection requirements (ULVO, thermal,
reverse conduction, cross conduction) and pretty soon the supporting circuitry and
control logic gets quite complex. Hence the specialised chip... hard to do better with
discrete components.

I had wondered if a simple circuit with 2 x P-Ch MOSFET's might work, thinking an
extension of the 1 x P-Ch MOSFET circuit sometimes used for reverse battery
protection with minimum voltage drop. Seems there would be issues with both on at
the same time... in which case you may as well just have a wire connecting the 2 x
USB's together (and I'm sure you want something better than that).

Two x Schottky diodes is hard to beat if you can tolerate some voltage drop.


--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Josh Koffman]

On Sun, Jun 14, 2020 at 5:23 PM Brent Brown <[hidden email]> wrote:

> The functional block diagram in the TI datasheet shows how they chose do it... In
> essence 2 x internal N-Ch MOSFETs. First off it requires a charge pump to get high
> side drive. Then you start considering protection requirements (ULVO, thermal,
> reverse conduction, cross conduction) and pretty soon the supporting circuitry and
> control logic gets quite complex. Hence the specialised chip... hard to do better with
> discrete components.
>
> I had wondered if a simple circuit with 2 x P-Ch MOSFET's might work, thinking an
> extension of the 1 x P-Ch MOSFET circuit sometimes used for reverse battery
> protection with minimum voltage drop. Seems there would be issues with both on at
> the same time... in which case you may as well just have a wire connecting the 2 x
> USB's together (and I'm sure you want something better than that).
>
> Two x Schottky diodes is hard to beat if you can tolerate some voltage drop.

Hi Brent,

I agree about the chip, it seems to be an easy to implement way of
managing this. I admit, part of me just didn't want to have to create
a new part in Eagle!

I'd prefer to avoid connecting the two USBs together. Seems like it's
a recipe for damaging a computer someday.

Agreed on the diodes. It's where I first went, and might not be too
bad. I do have a 7 segment driver (HT16K33) and some WS2812 LEDs on
the 5V rail though, so I'd have to do some tests to make sure I am not
too low for them.

Thanks!

Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
        -Douglas Adams
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist