When using one of the two op-amps in a dual op-amp DIP (TLV2372), which is the best way to terminate inputs & output? Inputs tied together and then to GND or Vdd, and what of the output, tie that to GND or Vdd?

 

Tie the two unused inputs to Vss (likely the most negative voltage), and let the output float.

 

Don't tie the output to GND or Vdd. That could cause excessive current flow.
You could tie the output to the inverting input, and tie the non-inverting input to either rail. One rail might be better than the other in terms of standby current, depending on opamp topology (I haven't checked the TLV2372 spec).

 

Don't tie the output to GND or Vdd. That could cause excessive current flow.
You could tie the output to the inverting input, and tie the non-inverting input to either rail. One rail might be better than the other in terms of standby current, depending on opamp topology (I haven't checked the TLV2372 spec).

I tie the non-inverting to the center of a voltage divider consisting of two 100k resistors between the rails on an OP amp powered by a single sided supply. This was recommended practices on the LT datasheet.

 

Just use pull down resistors on both in/out

 

I would tie the output to the inverting input and ground the non-inverting input. That way the amp output isn't saturated, which could conceivably have some adverse effect on the other amp, and it requires no added parts.

 

I agree with crutschow. Configure it as a voltage follower and ground the + input.

 

Your responses intrigued me, so I did more research. Seems my method is preferred for single sided supplies, and Crutschow's method for split rail supplies. I googled "how to terminate unused OP amp"; one article is:
http://www.electronicproducts.com/A...rs/Properly_terminating_an_unused_op_amp.aspx

PS It was Maxim's page that I first referenced this circuit from.

 

After reading the TI app note referenced by djs, the advice should read: wire it as a voltage follower with inverting input tied to output, tie the non-inverting input to a (bypassed) dc voltage source greater than Vss and less than Vdd which is within the Common-Mode Range.

 

After reading the TI app note referenced by djs, the advice should read: wire it as a voltage follower with inverting input tied to output, tie the non-inverting input to a (bypassed) dc voltage source greater than Vss and less than Vdd which is within the Common-Mode Range.

I found the same here.
http://e2e.ti.com/blogs_/archives/b...spx?DCMP=analog_signalchain_mr&HQS=thesignal1

 

Yes, I could have Googled (I did a little) and looked-up details but having dialog with all of you is much more fun. Thanks for all the helpful feedback (no pun intended). You're all terrific! I'm looking forward to our next exchange. Now it's time to go play with a cranky Agilent B1500A. Wish me luck! ...I may end-up needing a bigger hammer.

 

Now it's time to go play with a cranky Agilent B1500A. Wish me luck! ...I may end-up needing a bigger hammer.

Hey Techni, relax!

The worst you could do is to get angry with any piece of gear...but if worst comes to worst I wonder if you could convince him to lend his hammer to you

 

I would tie the output to the inverting input and ground the non-inverting input. That way the amp output isn't saturated, which could conceivably have some adverse effect on the other amp, and it requires no added parts.

As recommended by Walter Jung in one of his books.