Looking for general advice on ESD protection on board inputs that are carrying small signals, for example, tens of millivolts at very low current. I suspect that hanging diodes off it might compromise signal integrity due to diode leakage. Advice?
\$\begingroup\$
\$\endgroup\$
3
asked Aug 11, 2015 at 9:44
-
\$\begingroup\$ If you're worried about leakage, it doesn't really matter what the voltage is, it matters what's the source impedance. \$\endgroup\$– The PhotonCommented Aug 11, 2015 at 9:47
-
\$\begingroup\$ The source impedance is a piezo transducer, but I am driving into <10K so it is sufficient for that. \$\endgroup\$– Dirk BruereCommented Aug 11, 2015 at 10:07
-
\$\begingroup\$ There are a lot of low leakage devices available. Many were designed for high speed serial interfaces, but they should be good for a piezo interface as well. \$\endgroup\$– Peter SmithCommented Aug 11, 2015 at 10:37
Add a comment
|
1 Answer
\$\begingroup\$
\$\endgroup\$
2
Diode clamps (eg. back-to-back diodes to ground after some series resistance such as 100 ohms) are probably your best bet if your input impedance is < 10K. Ordinary 1N4148 diodes or a BAV99 dual will do. If the leakage at high temperature is deemed to be excessive, then just hang two or three in series. Here's what the typical leakage looks like at 0/25/50/75°C for 2 pairs in series back-to-back (4 diodes) with input voltage varying from 0mV to 50mV. 120nA makes your input appear as 9.77K rather than 10K, probably not a big deal.
answered Aug 11, 2015 at 10:53
-
\$\begingroup\$ Spehro, if there is a metal enclosure, connected to digital ground somewhere pretty far frombthe device itself, is it better to clamp to GND or to thus enclosure? \$\endgroup\$– user76844Commented May 2, 2016 at 10:50
-
\$\begingroup\$ If there is lots of series resistance to limit the current, then GND is generally better. \$\endgroup\$ Commented May 2, 2016 at 15:46