• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 16 subscribers
  • Views 4370 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TS3L301: Using an analog switch to switch Gigabit Ethernet between two devices and thus between the magnetics

Chris Fiege14
Prodigy 30 points
Part Number: TS3L301

Hi,

I am working on a design that is supposed to be able to disconnect an Gigabit Ethernet (1000 Base-T) Ethernet link between a Switch and a Device under Test.
Usually the test-switch and the DUT will be powered from the same power outlet - thus I don't worry about ground-shift on the mains side.
And usually I assume short cable length < 10m.

The obvious idea was to use RF-rated relays for this job. But those usually have a huge package, are cost-intensive and (depending on type) may need quite a lot of power to operate.

I am currently thinking about using the TS3L301 or TS3L501 to solve this problem.

My setup is currently:

  • The switch is powered with a locally generated 3V3-supply. This supply can be referenced to the GND of the DUT or to the GND of a test-server that should be on the same power outlet.
  • To reference the potential-free Ethernet-Link to my local GND I use a series combination of an inductor (22uH) and a resistor (1kR) to archive a DC-coupling to VCC/2. This gives me 330R for DC and >10k @ 200MHz
  • I accept the impedance mismatch between the twisted pair cable and the switch and do not do any impedance matching.
  • I will not use this design with PoE since it will destroy the switch.

Has anyone tried this?
Or has an idea of why it will not work?  For example I am currently unsure how high the differential level on the Ethernet-Lines really are.

I have attached my current test-board schematic.

Kind Regards
Chris

  • 0
    TI__Expert 3180 points

    Hello Chris,

    If you are using this device to isolate between two devices, then yes this should work as a method of switching between two devices.

    Please see the E2E post below converning magnetics with regards to your application:

    Could you provide us with more details as to why you need something like this for isolation?

    Thank you,

    Louie

  • 0 in reply to Louie
    Prodigy 30 points

    Hello Louie,
    thank you for your reply.

    I am working on device that allows us to automate testing of embedded devices. Thus it has features like USB-disconnect, power switches, potential free contacts. One more feature should be that we want to disconnect an ethernet link between the ethernet switch and the device under test.

    I don't need a galvanic isolation between the devices. The main purpose is to simulate a disconnected cable.

    Regarding your post I understand that I will need to protect the switch from voltages above Vcc and below GND.
    But I am still not sure about the magnetics: Why do I need those?

    Regards,
    Chris

  • 0 in reply to Chris Fiege14
    TI__Expert 3180 points

    Hello Chris,

    There are multiple reasons why you would need magnetics, but the two main ones are for protection and because the standard for the application requires you to have it. Please see the link below for for information:

    I hope this helps.

    Thank you,

    Louie 

  • 0 in reply to Louie
    Prodigy 30 points

    Hello Loie,
    I understand that the standard requires me to have magnetics on both ends of an ethernet-link. In my application this would still be the case.

    I am just unsure if the logic levels in between the magnetics are suitable for the TS3L301 analog switch.

    Thank you,
    Chris

  • 0 in reply to Chris Fiege14
    TI__Expert 3180 points
    Hello Chris,

    TI cannot guarantee any logic levels not specified in the "recommended" portion of our data sheet.

    If your logic levels going in to the device do fit within the VIL and VIH ranges, then our solution should work for you application.

    Thank you,
    Louie