1
\$\begingroup\$

I have a 4-channel TI TPD4EUSB30 TVS/ESD diode array that I've been using successfully on differential ethernet traces. Our design changed and the positive and negative pins have swapped positions which means I would have to flip the traces if I am to observe the indicated polarity on the TVS. However, I am not convinced that polarity needs to be observed because all channels are simply connected between two diodes as shown on the device schematic:

TI TPD4EUSB30

A similar Nexperia PUSB3F96 doesn't label the channels with any sort of polarity indication:

Nexperia PUSB3F96

Is there any reason I can't simply ignore the polarity on the TI device and avoid having to swap the positions of my differential traces? Nowhere in the datasheet are specifications given that differ between the D+ and D- pins. Is there any particular reason TI have labeled the channels with + and -?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The only problem I could possibly see is something about one pin being connected to the die substrate or something, but that feels unlikely to be a problem for this, it's just the only potential problem I can imagine. \$\endgroup\$
    – Hearth
    May 28, 2022 at 0:10

1 Answer 1

1
\$\begingroup\$

Probably just to suggest a typical layout. I don't see any differences between channels.

A similar protip: transformers don't care about P/N, RX/TX etc. They're identical and symmetrical, as long as you make the same changes on both sides (i.e. swap primary and secondary polarity). (Although that isn't even much of an issue, with almost all PHYs supporting Auto MDIX.) So layouts can often be simplified this way.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.