Quick question about diff op amp?

Discussion in 'General Electronics Chat' started by rougie, Aug 31, 2013.

  1. rougie

    Thread Starter Active Member

    Dec 11, 2006

    If we use a typical LM324 op amp with rails of +/- 15VDC, is it normal that:

    A) When the non inverting input is +5Vdc and the inverting input is +3Vdc, the output results in the rails minus approximately 1V. Output measured is +13.7 Vdc?

    B) Also the opposite happens when the non inverting input is +3Vdc and the inverting input is +5VDC, the output results in the negative rail plus approximately 1V. Output measured is -13.7 Vdc?

    I think this is normal behavior but not sure, can someone please confirm this.

    Last edited: Aug 31, 2013
  2. pwdixon


    Oct 11, 2012
    Most op amps do not output the full power supply voltage. The datasheet will tell you what to expect but it's usually the power rail less 1V to 2V.
  3. Jony130

    AAC Fanatic!

    Feb 17, 2009
  4. rougie

    Thread Starter Active Member

    Dec 11, 2006
    thanks guys
  5. WBahn


    Mar 31, 2012
    You need to start looking at datasheets. It will take some effort to start being comfortable pulling out the information you need and, frankly, there will always be sections of most datasheets that you ignore largely because they deal with issues that haven't bit you... at least yet.

    With the Wonderful World of Google, all you usually have to do is just type in the part number and you will get a link to the manufacturer's datasheet.


    You are looking for the spec on Output Voltage Swing. You see that with a 30V supply and 10kΩ loading, the LM324 typically gets to within 2V of the supply but is only guaranteed to get within 3V.

    On the otherhand, it is supposed to get very close to the lower supply claiming to get below at least 100mV when powered between 0V and 5V with a 10kΩ load. But I think that is misleading. I can't find if or where they give their test circuit, but I think that 10kΩ load is to ground. So it is completely unloaded. If the output is comparably loaded, you can probably expect to get to within roughly the same 2V to 3V of the negative rail.
  6. rougie

    Thread Starter Active Member

    Dec 11, 2006

    I appreciate your tutorial the part where you show:

    I never had thought of simply using just the feedback resistor... but seems pretty simple....

  7. rougie

    Thread Starter Active Member

    Dec 11, 2006

    Yes I am looking at it as we speak... I have the LM324 from ST... I guess their the same...

  8. rougie

    Thread Starter Active Member

    Dec 11, 2006

    When I initially posted, I was intrigued as to why the output would almost equal the rails when assigning two different voltages directly to the op amp's inputs, but okay, I seem to understand now, as these op amps have a huge gain by themselves and even if one applies a 1 volt difference, the output will hit the roof almost attaining the rail voltage.

    And now I also understand that even if the inputs are both grounded or both held at say for example 5Vdc, the amplifier is internally imperfect and so there will always be a small difference in voltage causing the output to hit the roof again lol!

    Interesting little critters these op amps aren't they.

    As for the voltage swing, bof... I will discuss in accordance to your spec, as the ST one may differ a little.

    The way I see it, at Voltage output High (VOH) at 30Vdc with a 10Kohm load, they guarantee 27 Vdc or typically 28Vdc in respect to a 30Vdc supply rail. In contrast, when looking for the maximum Voltage output low (VOL) in respect to the negative rail, with a 10Kohm load and a 5 Vdc positive supply rail you may come as close to the low rail by 5mv typical to maximum 20mv. Yup, I don't get it either... weird... Maybe I should try building a circuitand see what really gives.... but not today...
    Its labor day weekend and I'm at-a here :D

    Anyways, I get the point, I experimented by simply lowering the rails in one of my experiments, and the lower I went the more my output was attenuated by a volt or so beneath its respective rail.

    Thanks guys
  9. wayneh


    Sep 9, 2010
    This applies to comparators as well, such as LM339. If you attempt to compare against ground on one of the inputs, you cannot take for granted where the output will be. A bit of testing can resolve the issue, but I generally use some reference voltage other than ground so that I can be certain of the result.

    What I've said may make no sense to you unless you've experienced the problem for yourself. The take-home message is to not assume you know what an IC is doing under "borderline" conditions. Study the datasheet and/or experiment.

    Assume ignorance, and you will often be right!
    rougie likes this.