Does anyone have experience protecting <1nA current measurement circuits? Specifically a transimpedance amplifier input from transient overcurrent events.
Looking at ESD devices, diodes, zeners etc I don't see anything with specified leakage below 1nA.
On the other hand all these datasheets are specifying leakage near their breakdown voltage whereas the transimpedance topology means that during normal operation voltage will be ~0V.
And for example the circuit originally had standard back to back diodes that would simply forward bias and clamp to ~0.7V and this worked fine until at least the nA territory.
In many cases, pA input stages won't use other protection elements than the IC substrate/protection diodes, rated for 5 or 10 mA continuous and much higher transient current. Another option would be zero biased low leakage diodes, but you rarely get guaranteed specifications for the leakage at mV voltage.
Discrete nA or pA input stages often used JFETs as low leakage diodes, but I guess modern low leakage silicon diodes have similar specification.
Here is a Jfet which specifies gate current absolute max of 10mA. Do we think this would conduct 100mA or perhaps 1A for a very short transient or will this path 'pinch off' at a lower limit?