How / can you connect Ethernet devices together without magnetics? How do you wire it up? What sorts of practical limitations are there on doing this?

To clarify, I'm not trying to connect two computers here. I'm thinking specifically of creating a small network of microcontrollers on a single board. Lets say the total length of a network segment is no longer than 6 inches. This would be for 10BaseT or at most 100BaseT.

  • 1
    \$\begingroup\$ With a cable, as the magnetics are already in the complaint devices. You can make a non-compliant capacitively coupled device which sort of works, but not well. \$\endgroup\$ – Chris Stratton Dec 19 '12 at 19:47
  • 1
    \$\begingroup\$ Curiously, if you're staying on board.. why Ethernet over something like SPI? You can still reach very high data rates and you avoid having to fiddle with Ethernet. \$\endgroup\$ – Toby Lawrence Dec 19 '12 at 23:46
  • \$\begingroup\$ @TobyLawrence do you mean Ethernet over SPI (which sounds like a huge challenge) or IP over SPI (which sounds like overkill). \$\endgroup\$ – jippie Dec 20 '12 at 21:18
  • \$\begingroup\$ @jippie I meant SPI in lieu of Ethernet. Of course, it's not a mesh network.. but I don't know if the OP actually needs a mesh or it just happened to be the way he worded his question. \$\endgroup\$ – Toby Lawrence Dec 20 '12 at 21:19
  • 1
    \$\begingroup\$ In a follow up, I found this article that might be of interest: micrel.com/_PDF/Ethernet/app-notes/an-120.pdf \$\endgroup\$ – vicatcu Jul 8 '13 at 13:57

Keep in mind that the type of ethernet that uses magnetics requires the transformer coupling in the spec. The transformers do some nice things for you, like completely decoupling common mode signals, providing significant voltage isolation, and in some cases projecting a different impedance to the electronics than what is natively on the cable.

If your whole device is floating or the connection is short with a common ground, you can present the correct impedance and voltage levels, and can ignore common mode signals, then you can in theory eliminate the transformers.

This won't work with some PHYs, since they require center tapped windings on their side. It's not uncommon for the PHY to drive the transmit pair by having the center tapped tied to power with only low side switches on each end. That effectively allows for bipolar drive with single ended switches, and is a legitimate architecture since a transformer is supposed to be there anyway. Look carefully at ethernet transformers, and you will see many offerings with center taps partly for this reason.

If you're careful, you can put just one transformer between two devices if they are physically close. Each end is still transformer coupled, so common mode rejection is there and the center tap drive trick still works. I actually did this on a small system once where both ethernet devices were contained in the same physical unit.

  • \$\begingroup\$ I read this article recently: youritronics.com/magnetic-less-ethernet would you be willing to weigh on in it with respect to your answer? \$\endgroup\$ – vicatcu Jan 7 '13 at 17:58
  • 1
    \$\begingroup\$ @vicatcu: It doesn't look like anything new. All this shows that if you have the right PHYs with the same ground on the same board, then sometimes it can work. Again, some PHYs expect to drive center tapped windings, and not all PHYs work with 1:1 transformers. In those cases, direct coupling may not work. Over long distances and with different grounds, it is definitely a bad idea. \$\endgroup\$ – Olin Lathrop Jan 7 '13 at 19:03
  • \$\begingroup\$ Thanks for the feedback @Olin, my interest in this question was over a short distance - maybe a couple of feet. \$\endgroup\$ – vicatcu Jan 7 '13 at 20:12

As Olin said, it can be done with capacitors in certain situations. The most interesting one is Ethernet on a backplane.

  • \$\begingroup\$ That is indeed the sort of situation I was thinking of... \$\endgroup\$ – vicatcu Dec 19 '12 at 20:57
  • 3
    \$\begingroup\$ There is a standard for Ethernet on a backplane, it is IEEE 802.3ap. Near as I can tell, it still requires a transformer. \$\endgroup\$ – user3624 Dec 19 '12 at 21:53
  • \$\begingroup\$ No, it's not. 4 capacitors will suffice (in a so-called SERDES, aka PICMG, aka 1000base-kx mode). \$\endgroup\$ – oakad Nov 12 '13 at 4:17

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.