1
\$\begingroup\$

How does one surge protect a differential amplifier circuit? Assuming the absolut maximum stresses on the amp IC input pins is Gnd-0.5V -> Vcc.

Clearly TVS alone are insufficient to clamp to Vcc. What is the usual solution for this? Say for a 1kA residual lightning protection assumption.

\$\endgroup\$
  • \$\begingroup\$ You could probably simulate this in LTSpice. \$\endgroup\$ – Andy aka Aug 11 '14 at 14:03
  • \$\begingroup\$ What are the details of the pulse you want to handle? And how much series resistance can you tolerate? Series impedance, TVS, more series impedance, another clamp. Active impedances are attractive in some situations (low noise amplifiers). \$\endgroup\$ – Spehro Pefhany Aug 11 '14 at 19:01
3
\$\begingroup\$

For reference, here is information for real-world protection units, meeting sets of safety regulation specification, as used in industrial plant settings, says, sensors and remote control wiring running hundreds meters on outdoor paths, subject to lightning strike.

enter image description here

As original poster suspects, cascaded protection is typically used inside these unit.

You may design your own unit, based on below general information, or buy off-the-shelf.

Gas discharge tube, with different rating specification, is used as the front, to take majority of the energy of the incoming lightning induced surge pulse. This example unit is rated 10kA for 8/20 surge timing profile.

The next is inductor to attenuate fast rising pulse. The current rating is 250mA in this example and can be used for sensor and power supply protection. Some units have lower current rating and limited to sensor use (like 20mA loop sensor/transmitter or near zero current like thermal couple).

Very fast semi conductor devices are used at the final stage to clamp the residual surge to the rated 6V, as in this example. Different unit has different voltage rating, choice depends on the expected operating voltage of the signal line (for thermal couple, near zero. a bit higher for power supply or 20mA transmitter)

For info at a randomly chosen example meeting safety regulation specifications

For info on test specification on pulse shape, like 8/20us, pulse shape info

enter image description here

\$\endgroup\$
0
\$\begingroup\$

You normally clamp the two inputs to their maximum allowable common-mode range using diodes and some sort of series impedance (resistance) to limit the fault current.

If you have more stringent limits on the allowed differential voltage input, you might apply additional clamping directly between the two inputs. This might take the form of back-to-back silicon diodes for low-level signals, or zener diodes or other technologies for higher-level signals.

If you need particularly low parasitic capacitance, JFETs can be configured to do the clamping, and they generally have much lower capacitance than general-purpose diodes.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the answer. Schottky diode clamping would be ok, for normal operation EMI, but I fear they will not survive a lightning surge, e.g. 1kV 10/350 pulse. TVS diodes would work, but their forward voltage is too high to clamp to satisfy the limits. Combining both would work for normal operation (due to the Schottkys) but I fear during a lightning surge, the Schottkys will burn out, again leaving the input pins exposed due to the too high forward voltage of the remaining TVS. I suspect that some sort of cascaded protection might be a solution but am unclear on the details. \$\endgroup\$ – ARF Aug 11 '14 at 17:42
  • \$\begingroup\$ I'd give it a try ... carefully ... \$\endgroup\$ – Spoon Aug 11 '14 at 20:23

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.