This is something I've been wondering for some time: I notice that some op-amps are not specified to be unity gain stable. Couldn't you just set the gain to maybe 5 or 10, and then run the output to a matching 5 or 10 resistive voltage divider for a "virtual" unity gain effect?
Seems too simple. I'm probably missing something.
Seems too simple. I'm probably missing something.
You could, but of course you'll increase the effective output impedence by doing this, making systems like active filters troublesome to design. In fact, most systems/stages requiring a unity gain are going to be severely restricted by having this increased output impedence, the fix for which is a unity gain buffer following it (making it totally irrelevant!) 
I should ask specifically what it is you are trying to do, what this unity gain application is?
I should ask specifically what it is you are trying to do, what this unity gain application is?
Well, you'll get a bit of additional noise due to the excess gain, and the voltage divider (just as with a regular attenuator) will result in the chip being driven from a bit higher source impedance which will increase distortion a bit.
Don't know that I'd lose much sleep over it though.
Why not use a unity gain stable chip like the OPA-549?
se
Don't know that I'd lose much sleep over it though.
Why not use a unity gain stable chip like the OPA-549?
se
Yes, but why bother?
It is not the amp design itself which has a specific stability, but how much internal compensation is applied.
I am not sure there are any op-amps available which do not have a version which is unity gain stable.
So use that one.
Nothing specific. I started thinking about it partly because I have some older, high-power, non-unity gain stable opamps in my "junkbox".
Thanks guys for the responses. You brought up some points I hadn't considered.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.