RE: Protecting Inputs from Damage— EOS

Bruce Trump — Mon, 04 Feb 2013 16:57:11 GMT

David-- Very interesting comments. POTS (plain old telephone service) lines must rank among the most challenging EOS situations regularly dealt with in the industry. Your solution sounds great as you are unlikely to find an LDO that would meet the requirement for sinking impulses of this magnitude. Good job! -- Bruce

RE: Protecting Inputs from Damage— EOS

David Smith70335 — Mon, 04 Feb 2013 16:39:04 GMT

Bruce, this is an interesting article. My company designs telecommunication test gear, and some models connect to POTS lines where voltage transient protection for op amp "line drivers" can be a problem. Invariably we use R-D (resistor-diode) protection where the backdrive current is often shunted to either system ground or to the local op amp power supply. The supply may normally output 25-50mA but to handle a current backdrive of 6A? I resorted to rolling my own regulator using a fast, precision op amp with a discrete push-pull output stage. The asymmetric output stage was designed to source up to 100mA continuously while the current sink capability was 7A peak. Due to AC coupling of the op amps to the line that we are protecting, the current pulse would die out in 25mS-45mS, but during the pulse event we were able to keep the 5V rail from moving more than 20mV so none of the other devices using this distributed rail ever get close to an overvoltage situation. I biased the output sink transistor so that it would idle under 1mA (Class A-B) so that the supply would remain relatively efficient but would respond cleanly and quickly when called upon to switch from source to sink and back to source.