1
\$\begingroup\$

Generally TVS diode curves are shown as

Uni and Bi directional TVS Diode curve (Source https://www.ti.com/document-viewer/lit/html/SSZT784)

With the left most being a uni directional and right being bi directional TVS diode.

TVS diodes have a

  • Working or Reverse stand off voltage and this should be less than or equal to your voltage of operation (VRWM)
  • Breakdown voltage where if the voltage across your load meets this voltage, the TVS diode will begin to conduct in reverse bias mode (VBR)
  • Clamping voltage where if the voltage at the source continues to increase (e.g. due to a momentary surge) then voltage across your load will remain at this clamped voltage and the TVS diode will conduct more current through it in order to do so (VC)

TVS diode voltages (Source TVS diode specs - clamping voltage lower than breakdown voltage?)

However when looking for a TVS diode to protect a 3.3V Ethernet data line on the Ethernet PHY side of the magnetics. I have come across some TVS diodes with a clamping voltage less than the Breakdown voltage.

Reading this stack exchange question and answer TVS diode specs - clamping voltage lower than breakdown voltage? I understand this 'snapback' behaviour (as shown below) but why?

Snapback vs regular TVS diode curve (Source https://blog.mbedded.ninja/electronics/components/diodes/tvs-diodes/)

For example this TVS diode https://www.mouser.co.uk/datasheet/2/1057/PS1703_D32-3498786.pdf on page 2 it says

  • V reverse stand off = 3.3V (MAX)
  • V breakdown = 11V
  • V clamp = 3.4V

The device I want to protect (Ethernet PHY) has an absolute max voltage on the digital input pins of 3.6V so a V clamp of 3.4V is perfect but the breakdown is 11V.

Is this TVS diode still okay to use for my purposes? Is it reasonable to assume that the 11V breakdown 'knee' voltage would be only for a moment and then the TVS would operate in clamping mode and although the device I wish to protect says it has an Absolute maximum of 3.6V, this is 3.6V continuous and would survive an 11V momentary spike?

In my mind, ideally I would find a TVS with

  • V reverse stand off = 3.3V
  • V breakdown - 3.4-3.5V
  • V clamp 3.5-3.6V

But I cannot seem to find one that meets this? Am I missing something or this does not exist hence the reason for the 'snapback' TVS diode? If this is not the case, why would you need a snapback TVS diode?

\$\endgroup\$
2
  • \$\begingroup\$ How much pF load and how much surge current do you expect? \$\endgroup\$
    – D.A.S.
    Commented Nov 27 at 18:18
  • \$\begingroup\$ Maybe you could be more lucky with a TVS varistor? \$\endgroup\$
    – user317139
    Commented Nov 27 at 23:38

1 Answer 1

1
\$\begingroup\$

A TVS diode with snapback behaviour has different semiconductor structure than a "regular" TVS.

Regular TVS uses a P-N junction or actually many of those.

Snapback TVS is structured almost like a thyristor or SCR. So at certain breakdown voltage it becomes conductive and the voltage drops. It can actually drop quite far in "deep snapback" TVS. Then it maintains the holding voltage for the surge duration. Of course you cannot use it as a thyristor due to its power rating.

So obviously the snapback TVS is better suited for low voltage system protection as it maintains lower voltage after the overvoltage event triggers it.

Unfortunately I cannot say if your "device" will survive a momentary 11V spike as I do not know what the device characteristics are and what are you trying to protect it from.

\$\endgroup\$
2
  • \$\begingroup\$ How would I determine if the device can survive a momentary 11V spike? The datasheet does not mention anything about its Transmission Line Pulse (TLP). \$\endgroup\$
    – MRB
    Commented Nov 28 at 10:30
  • \$\begingroup\$ It is probably more important to know what transient are you trying to protect the device from. In most cases a TVS certified for this purpose, like TPD1E1B04 by TI would do, but it is not one size fits all. Keep in mind you also need a small resistor between your device and the TVS. The TVS alone will not clamp the voltage enough. You need a path of greater impedance when the TVS starts conducting, so the current flowing to your device is negligible. The best way to be sure is to test it thoroughly, but i get that you might not have the time or resources. Hope this helps. \$\endgroup\$
    – SRomie
    Commented Nov 28 at 23:57

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.