# LM324n weird output when used as buffer

I am using an LM324n quad-core op-amp. I have connected it between +24 V and GND. One of its op amps is used as a buffer. When I pass the buffer a negative voltage, I get it inverted at the output although I expected it to be zero. Is this weird response somehow logical or expected?

Unfortunately I had some issues with Windows lately and I have no other schematic but for this handwritten one. MOTOR VELOCITY input can become both positive and negative and that's where the issue is!

• Schematic? Picture of the implemented circuit?
– Null
Nov 11, 2014 at 18:31
• Kind of schematic added Nov 11, 2014 at 18:35
• This is real. I think Nat Semi documented its use as a full wave rectifier in one of their Floobydust articles ... anyone have that databook handy?
– user16324
Nov 11, 2014 at 18:37
• Olin and The Photon have the correct answers, but this is the first time I've heard of a quad op-amp described as quad core, LOL. Nov 11, 2014 at 19:09
• LOL that was a typo, but I'll leave it for the sake of the laugh!! Nov 11, 2014 at 19:33

This behavior is called phase inversion.

It happens in certain op-amps when you exceed the input common mode range limits of the op-amp (for example, when the input voltage is below the most-negative supply voltage).

It's explained at the circuit level in a TI presentation I found (warning: Powerpoint link), and also in an EE Times article.

Notice the input common mode minimum voltage for your op-amp is equal to the most negative supply voltage, so you should not use a 0 V VEE with this op-amp if you will have negative inputs. An easy solution in your case might be to use the op-amp with split supplies to handle the velocity input.

• But since the negative input voltage is out of spec I should avoid it anyway right? Nov 11, 2014 at 19:30
• In the two datasheets I checked (On Semi and TI), the specs are written assuming VEE = GND. If VEE = -15 V, input common mode range should go down near -15 V. The chip itself really has no idea what "0 V" is except what is attached to it's VEE pin. Nov 11, 2014 at 19:38

You need to actually read the datasheet:

Note that a negative input voltage is out of spec. When you violate a spec, there are no longer any guarantees what the part might do. So yes, there is nothing unexpected about it inverting the signal onto the output, or doing anything else that might appear strange. When you don't follow the rules, you can't expect it to follow the rules either.