OP Amp offset

[BeBuLamar]

I build a simple op amp using the TL82 chip and use only 1 of the amp. I configure it as a simple non inverting amp with the gain of 11. It works well except the ouput has some DC offset that is the output is shifted slightly to the negative. How do I compensate for this offset? I know some of the op amp chip has pins for this but this one doesn't.

[Kim Christensen]

Post a schematic of the circuit and explain what you're using it for. There are ways of doing it depending on the design and what you are trying to do.

[Vovk_Z]

It seems to me 'Art of Electronix' bible gives good clear examples of how to work with unused opamps and how to work with an offset voltage.

[shapirus]

I build a simple op amp using the TL82 chip and use only 1 of the amp. I configure it as a simple non inverting amp with the gain of 11. It works well except the ouput has some DC offset that is the output is shifted slightly to the negative. How do I compensate for this offset? I know some of the op amp chip has pins for this but this one doesn't.

https://www.ti.com/lit/an/snla140d/snla140d.pdf

page 12.

For dealing with unused op amps, see the respective datasheet. But in the general case, you don't want to either: 1) leave the inputs floating; 2) tie them to either of the power rails. A good way to deal with them is to wire them as voltage followers with the non-inverting inputs connected to the middle point (or thereabouts) between the power rails.

[David Hess]

How do I compensate for this offset?

For the non-inverting configuration, a small offset can be added, with another divider, to the grounded side of the divider.  Sometimes this added offset is buffered with a second operational amplifier configured as a voltage follower.

[dobsonr741]

Post a schematic of the circuit and explain what you're using it for.

Yes, please do post the schematics. And tell what's the slight offset voltage you are not happy with.

[Terry Bites]

A TL082 has a worst case offset of about 15mV (Original TI datasheet)
With a gain of 11 that gives a possible 160mV but more typically 50mV.
That's measured with the + input grounded.
It may be above or below 0V.
If its bigger than that you have a circuit or a measurement error.
Where did you buy the TL084, is it a genuine part?

see e2e.ti.com/blogs_/archives/b/thesignal/posts/where-are-the-trim-pins
It shows the simple offset correction trims that can be applied.

[donlisms]

If you think of the offset as a small voltage source in series with one of the inputs, just before the input, you can compensate by introducing the opposite voltage at the inputs, in whatever way you like.

One method, if it's an inverting amp, is to replace the ground connection at the + input with a small adjustable voltage. This is probably easier to adjust if you put the trimmer between two larger resistors that soak up most of the supply voltage, leaving only a little dropped across the pot.

For a single supply circuit, it changes a bit because of the issues with combining dividers and the way all the currents through the resistors interact; you don't want two dividers talking directly to each other. If you've already got a divider at + you may be able to just put a trimmer in the middle.  Simple. 

It's also possible to sum a small offset into the inverting input. Check "summing amplifiers" if you aren't already familiar.  Here you also have to think about combined impedances and how the current will flow; a simple voltage divider feeding a summing resistor becomes part of the feedback network of the amplifier, so the voltages will get wonky. This is a good place to use a buffer: from the adjustable voltage divider, to the buffer amp with low impedance output so it can't get pushed around, through the summing (scaling) resistor, to the summing node of the amp. This isolates the adjustment divider from the feedback network and gives you simple control of the voltage.  And here, you can use a big summing resistor to scale the divider voltage way down, so it's easier to adjust. And of course the three-part divider trick.

It isn't always necessary to fix it at the first amp's input(s); you can fix it at the next stage or somewhere else in your circuit.

And depending on what you're doing, if your signal is AC, you might be able to just use AC coupling to block the offset.   

[John B]

If you're only using a single opamp out of the dual package anyways, considering using the TL081 or TL071 which has offset adjustment pins.

[MathWizard]

In general, if they make single op-amp's in a dual-sized chip, do they design the op-amp's output drive to be more powerful ?

[Kim Christensen]

In general, if they make single op-amp's in a dual-sized chip, do they design the op-amp's output drive to be more powerful ?

Might be able to dissipate a bit more power as a single op-amp per package, but otherwise the individual amplifiers' performance are identical.
ie: each amplifier in the TL081, TL082, and TL084 have pretty much the same specs.

[David Hess]

In general, if they make single op-amp's in a dual-sized chip, do they design the op-amp's output drive to be more powerful ?

No, it will have the same current limit.

What does change is that the power limit of each operational amplifier in a dual will be slightly lower, but this is seldom a problem.  For instance both the single and the dual LT1006/LT1013 have the same junction-to-ambient thermal resistance of 130C/W in an 8 pin DIP, but that means the single could in theory dissipate twice as much power as each amplifier in the dual.

[BeBuLamar]

I hope you can see the schematic.
With zero input the output is about -80mV. With 1V sine wave the output is offsett by about 100mV.

[Kim Christensen]

If the circuit is meant to only amplify AC signals, then simply putting a capacitor in series with R1 (The 10K) will reduce the offset by a factor of 11.

But, if you want it to work all the way down to DC, you could do something like this:
(Circuit could use some tweaks, but that's the general idea)