2
\$\begingroup\$

I'm not very familiar with proper ESD protection and wondering about the correct implementation of a TVS diode to protect an input of an isolated circuit.

In the case of an ESD event into the protected input the current is shunted to ground through the TVS diode - but I would like to avoid the buildup of a huge potential difference between earth and circuit ground. As a method of slow discharge I have often seen a ~1MOhm resistor to earth, and i believe I recall that sometimes a small capacitor in parallel is used to provide a low impedance path for fast transients like ESD? If that is correct, what type of Capacitor would be used? I guess it should be Y rated, but most Y Caps that I can find are limited to something around 300V, probably not enough for the ESD surge?

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$

2 Answers 2

2
\$\begingroup\$

Just consider the likely flow of the ESD current back to earth without the added capacitor: -

enter image description here

The capacitor in red by the transformer will be the inter-winding capacitance of the transformer. It'll be maybe 100 pF and is a good route for surge currents.

Adding C2 does create an extra path to earth but, given that your circuit is floating then it's not really that big of a deal. However, if your circuit connects to other equipment that may be earthed, there may be a different route to that shown above. In other words, you have to consider the big picture. If you are happy that the circuit you posted does that then I would say that C2 is not needed.

\$\endgroup\$
4
  • \$\begingroup\$ You're right, the circuit could potentially be suddenly connected to earth at any time, thats why I want to avoid a large potential difference \$\endgroup\$
    – Miomio
    Sep 3, 2021 at 13:24
  • \$\begingroup\$ Well, its the added path to earth via your circuitry that you need to work on. I can't guess what it is. If it's a hard path to earth then the current flow I've shown may be miles off. It all depends on the details you give. \$\endgroup\$
    – Andy aka
    Sep 3, 2021 at 14:15
  • \$\begingroup\$ Nothing fancy, just a simple bench lock-in amplifier. The input and circuit ground could be connected to practically anything, be it on purpose or accidentally \$\endgroup\$
    – Miomio
    Sep 4, 2021 at 5:42
  • \$\begingroup\$ @Miomio I think you don't understand what I'm saying. The problem lies in the details of the actual circuits and interfaces. I am not able to offer anything more than this. You need to work on this bit and fully detail the paths to earth from the various points in your circuit. The devil might be in the detail but I am not privy to that detail. \$\endgroup\$
    – Andy aka
    Sep 4, 2021 at 8:27
2
\$\begingroup\$

Just an opinion, but not a solution.

When I bring a product to a test lab, in order to get approved for a standard, I do not want to have any doubt if the TVS works, discharges ESD the most effective way.

ESD has to be discharged to wherever the ESD voltae is referenced. And, that reference is likely the Earth/Chassis ground.

R1 & C2 are to dissipate "developed charge" or "voltage by parasitic coupling/leakage" on the secondary side, but not very ESD effective (can be damaging).

I would place 2 of TVS, one between high-side and Earth, another between low-side and Earth.

\$\endgroup\$
1
  • \$\begingroup\$ TVS diodes can be used for all on board ESD protection. \$\endgroup\$ Sep 16, 2021 at 4:54

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.