a general question regarding op amps and negative feedback

[yuvalkesi]

Hi,
I have a rather general question regarding op amps, which I couldn't find an answer to...
Lets say I have an op amp with a negative feedback of some kind, +V=+3v, and -V=-5v.
If there's some kind of problem with the design, the op amp output will output +3 or -5v. That's what I don't understand.
I've encountered these kind of problems in a few of boards I took care of, and tried to figure out why this is happening. I was told this is a known issue, concerning with the feedback, maybe?
Need your help!
Thx.

 

[leo_o2]

I guess one of reason is positive feedback. If negative feedback becomes positive feedback, it leads to power rail ouput.

 

[FvM]

Unconnected (floating) OP inputs or a "broken" feedback loop are the most likely explanations.

 

[LvW]

I guess one of reason is positive feedback. If negative feedback becomes positive feedback, it leads to power rail ouput.

Just to explain, what leo_o2 means is the following:
When you intend to realize negative feedback (for low frequencies) this feedback can turn into positive feedback with a loop gain > 0 dB (causing instabilty or saturation) if the opamp is not unity-gain compensated.

Besides this, it's not clear to me what you mean with: "If there's some kind of problem with the design"
Another possibility could, for example, be a feedback path that blocks dc (containing a capacitor). In this case, also the opamp goes into saturation due to an unstable operating point.

 

[yuvalkesi]

Thank you all for your replies.
I will be glad if you could explain, more thoroughly, why does "it can turn into positive feedback with a loop gain > 0 dB (causing instabilty or saturation) if the opamp is not unity-gain compensated"?
If there is a neg.feedback connected, how can it turn into positive?
and FvM,
if the feedbackl is broken, why does it go to V- or V+?

regarding " some kind of problem with the design", I mean, that few boards I've seen had this error, and I tried to figure out why.
Thanks again.

 

[LvW]

I will be glad if you could explain, more thoroughly, why does "it can turn into positive feedback with a loop gain > 0 dB (causing instabilty or saturation) if the opamp is not unity-gain compensated"?
If there is a neg.feedback connected, how can it turn into positive?

When the output signal of the opamp exhibits a phase shift of -180 deg, this phase shift - together with the signal inversion at the inv. input - results in feedback with 360 deg (=0 deg), equivalent to pos. feedback.
For your understanding: There will be always frequencies at the output with a phase shift of -180 deg, however, if the gain magnitude is already below unity this will cause no problem.
Study Nyquist's stability criterion!

 

[yuvalkesi]

And how does unity gain affect stability? can you explain?

---------- Post added at 16:27 ---------- Previous post was at 16:23 ----------

I found this link:
https://focus.ti.com/lit/an/slyt087/slyt087.pdf
So I'll try to figure it out.
can you give a short explanation?
Thanks.

 

[FvM]

if the feedbackl is broken, why does it go to V- or V+?

Because a few uV differential input voltage are sufficient to drive the ouput into saturation. If no feedback path is in effect, it's rather unlikely that you hit the small input voltage band that results in a non-saturated output.

 

[alexan_e]

An opamp tries to keep both inputs in a zero difference (same) by design, if any of them is higher then the output changes to try and equalize the two inputs, if there is no feedback then the output goes to a voltage level very close to the supply rail because this is the best effort it can do to equalize the inputs.

Alex

 

[LvW]

And how does unity gain affect stability? can you explain?

---------- Post added at 16:27 ---------- Previous post was at 16:23 ----------

I found this link:
https://focus.ti.com/lit/an/slyt087/slyt087.pdf
So I'll try to figure it out.
can you give a short explanation?
Thanks.

roymesi, before explaining special procedures to enlarge stability properties, you should try to become familiar with stability in general. That means: Why resp. under which conditions can an amplifier become unstable at all?
Therefore, consult any textbook dealing with opamps and feedback