Op-amp input impedance when unpowered

AX tech editor
Joined 2002
Paid Member
If you separate the switches from the switching control, use DPDT small signal relays in R-2R, using one section for each stereo channel. That would be really compact, constant input Z, put the relay close to the signal circuitry and control from wherever you want.
That's what I used in the autoranger (OMRON relays G6K-2P-DC5) and it works extremely well.

Jan
 
Yes, something using relays should be the ultimate switcher, very linear and all. Even with limited lifetimes and the unequal switching transitions of the LSB and MSB relays, they're likely to outlive most of the equipment in the signal chain.

Are you aware of any popular developments in the continuous domain? I happened to find a Thailand paper showing a simple opamp-only multiplier based on the identity (a+b)^2-(a-b)^2=4ab. Huge THDs though!
 
I gather that there are only a few methods for proper multiplication in the continuous domain and none of these, in my opinion, seem to be worth the pain. It looks like I would do a modified R-2R ladder attenuator, if I have to do one someday. For now, I have settled for a fixed ratio.

Thank you all, especially Mr. Didden, for your valuable inputs and tips.
 
Not sure how wise it is to put volts on an op amp input with no power.
It could latch up and cause problems.

You'll probably damage/destroy the opamp front-end if you inject a strong signal when unpowered - this is one reason series resistors are often seen on inputs even though it increases noise - it helps protect the front-end pair.

Audio opamps usually have back-to-back diodes across the inputs to reduce the risk of zener-breakdown in the front-end (which knackers noise-performance permanently).


Opamps are remarkably robust when the inputs and outputs are within the supply rails - take them outside and all bets are
off - a good reason for protection schottkys to the rails.