Of course that's what he means. The simple answer is that they're not perfect. It takes a small (usually microvolt or millivolt) DC "signal" between the + and - inputs to get the output to zero. It's spec'ed as the input offset voltage. Multiply this number by the open loop gain, and you get a big number which is typically greater than the supply volatge and that's where it "tries" to go. It stops at the rail.
"Latch-up" is a seaprate phenomenon, and that requires overdrive.
"Latch-up" is a seaprate phenomenon, and that requires overdrive.
Re: Most of the time...
That's not the op-amps fault.
SpeakerScott said:the go to one supply rail or the other because of a mistake in the input circuitry...and when you think about it, they are simply responding to the conditions at the input terminals.
Scott
That's not the op-amps fault.
TL072 and similar opamps do strange things when input nears negative rail voltage...
http://www.dself.dsl.pipex.com/ampins/webbop/072.htm
New TI datasheets say "Latch up free operation" though so it might only apply to old parts and/or parts from other manufacturers.
http://www.dself.dsl.pipex.com/ampins/webbop/072.htm
New TI datasheets say "Latch up free operation" though so it might only apply to old parts and/or parts from other manufacturers.
Here's a TL072 or similar.
Node B sits at 2Vbe:s + the drop over Re above negative rail. If an input is driven negative enough so that the gate junction of one of the JFET:s lets the input steal current from the current source then there won't be enough voltage at B to turn on the vas and the output will go to the positive rail.
For example, if IN+ is connected to negative rail, S will be 1Vbe above negative rail. B will be at about the same voltage, which isn't enough to turn on the VAS so the output will go positive instead of negative as expected.
The TL0xx series opamps input common mode range includes positive rail though!
Node B sits at 2Vbe:s + the drop over Re above negative rail. If an input is driven negative enough so that the gate junction of one of the JFET:s lets the input steal current from the current source then there won't be enough voltage at B to turn on the vas and the output will go to the positive rail.
For example, if IN+ is connected to negative rail, S will be 1Vbe above negative rail. B will be at about the same voltage, which isn't enough to turn on the VAS so the output will go positive instead of negative as expected.
The TL0xx series opamps input common mode range includes positive rail though!
Attachments
Re: Re: Most of the time...
Nope, and that was my point. I guess I should have been more clear. Most times I've had an op-amp go to one rail, it's because of a wiring or design mistake. Rarely have I had a bad part. Understand how op-amps work and circuit faults can be readily diagnosed.
Scott
wg_ski said:
That's not the op-amps fault.
Nope, and that was my point. I guess I should have been more clear. Most times I've had an op-amp go to one rail, it's because of a wiring or design mistake. Rarely have I had a bad part. Understand how op-amps work and circuit faults can be readily diagnosed.
Scott
AndrewT said:why do some invert the output when driven hard?
Simplified, because an inverting gain stage may (depending on the toplogy) become non-inverting when the active device(s) saturate. Therefore, negative feedback becomes positive feedback - and the output latches.
I had TL074 went to the rails before. It was wired as Unity Gain Buffers for a distribution amp as in one of their Application Notes. Initially on power up, no problems. Left it on overnight. Next morning, output was at one of the rails. Apparently, the output DC offset crept towards the rail as the opamp temperature increases.
Pardon my ignorance, but what exactly does a DC bias setting resistor do?megajocke said:If one AC-couples the input without a DC-bias setting resistor after it something like that could happen... Input bias current of the opamp is pretty low so it might take a while though
janneman said:Simple: they rail because the differential input voltage times the gain is higher that the supply voltage.
Now, the reason for that may be another story...
Nice try Jack!
Jan Didden
Jan,
Your response was what I was trying to say...but you said it way better.
Thanks,
Scott
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Why do opamps "rail" ?