I know that gigabit ethernet needs to use all four pairs of wires. The idea is that POTS is such low frequency that I believe it should be possible to separate it from the gigabit signal using a filter.

Is there any other technical problem that makes it difficult? e.g. The effect of a filter on gigabit signal timing (I'm just making this up. I don't actually know much about how ethernet signaling works.)

  • Did any answer help you? if so, you should accept the answer so that the question doesn't keep popping up forever, looking for an answer. Alternatively, you could provide and accept your own answer.
    – Ron Maupin
    Aug 7, 2017 at 22:11

4 Answers 4


Encode the telephone signal, assign it to a VLAN, send it across with the rest of the data, as data, pick off the VLAN, decode the signal. Any other route is madness. (If doing POTS at all. VoIP obviously solves it at a higher level.)


Ethernet over twisted pair uses baseband signalling(as in 1000BASE-T). The baseband starts at (close to) 0 hertz, so frequency-wise there's no band to accomodate POTS which also starts at 0 Hz.

However, there's a new 1000BASE-T1 standard (for automotive) that uses just a single pair. Maybe you can find a media converter somewhere.

  • It's baseband in the sense that the gap below the signal is small compared to the bandwidth of the signal but there is still a gap. I'm strugging to find exact details but it seems that they only bother to specify the performance of the transformers down to 1MHz so if we take that as our guideline there would be plenty of room for POTs. Feb 23, 2018 at 20:52

This is expressly forbidden by the standards; you are not allowed to mix applications in the same eight-conductor sheath, even on different pairs, but you can mix applications on a 25-pair sheath, as long as the applications are electrically compatible.

You would never get such mixed use to pass the required tests. You will create impedance mismatches and asynchronous echoes from splitting off the applications, and it would cause problem with the various measurements (e.g. frequency, insertion loss, NEXT, PSNEXT, FEXT, ELFEXT, PSELFEXT, return loss, propagation delay, delay skew, balance, longitudinal conversion transfer loss, etc.). POTS also uses a high voltage to indicate RING, and this would be detrimental to network equipment.

Why, in a world of VoIP, would you want to remain with POTS?


Filters get ugly, insertion losses, phase distortion and all manner of badness. I'm not saying it couldn't be made to work but designing the filters would be non-trivial.

However there is an alternative. We can borrow an idea from the POE folks and use transformers to run the POTs like between two pairs while running the Ethernet signals within the pairs. I will admit I haven't actually tested this but I would be extremely surprised if it did not work.

  • I think this would actually work, but you'd have to build it yourself.
    – Zac67
    Feb 23, 2018 at 19:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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