This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM311: Voltage develops at input pin when comparator is disconnected

Part Number: LM311
Other Parts Discussed in Thread: TL331, LMV7235, LM393, TLV7021, LMV331, LM2903, OPA2340

I have attached a schematic where I use the 311 comparator. We removed the hysteresis feedback C5 and R7. Pin 3 is set to a voltage value of 0.2V with the voltage divider. We also removed the pull down resistance R3 to keep the circuit simpler.

The J6 input voltage should just be zero unless there is an actual input, but in practice we see that if J6 is disconnected, we measure a voltage of about 0.06V across the terminal. We used another comparator (single sided type) and the there was a different voltage but it was still present. In effect the voltage at J6 creates a bias in the voltage signal that was to be fed into J6 so that makes the setup not work.

The values are actual measured prototype values and not simulation values. The circuit seems simple enough that we probably dont need to simulate it, and I recreated all the values on a separate proto board with just the LM311 and no other components.

I am wondering if I am doing something wrong to get this bias voltage or if this is normal and I should be thinking of another fix.

  • Hello Alvin,

    The LM311 is a bipolar PNP input stage. There is bias current flowing OUT of the input pins from V+ towards ground. So any resistance in series with the input will develop a voltage across it.

    60mV/9k = 6uA

    The "typical" bias current should be around 100nA. But you are violating the input common mode range when at 0V - so the input stage is not in a proper operating condition and the internal tail current is flowing out the inputs.

    For a 5V supply, the LM311 valid input range is 0.5V to 3.5V.

    To operate at 0V input, you need a device that includes ground. There are several devices that can meet this requirement. Is there a particular reason you are using the LM311? The LM311 has a unique pinout, so there are no drop-in, pin to pin replacements.

    I would suggest you try the TLV7021, LMV7235, TL331, LMV331, LM393, . All of these can operate from 0V (the first two are Rail to Rail input).

  • Hi Paul, 

    I actually see this 60mV in my proto circuit without the 9k resistor. We observe the same problem even when supplying the inputs with a negative voltage. It doesn't seem to work well when it is open ended. I did change the power supply to be double sided, but it still doesn't seem to be working well near ground. 

    We are able to change the comparator, no real reason to stick with the LM311. I did have the LM393 but I think I was trying out the LM2903 as an alternative. I will give the LM393 a try and see what happens.

    As a last resort, do you think putting a voltage follower in front of the input works? I tried it with an OPA2340 and that seems to work well, it isolates the input from the comparator but I'm not sure if there are any drawbacks to that. 



  • Hi Alvin,

    You cannot "float" an input pin. They must have a path to ground.

    The "LM" devices are bipolar and will have measurable bias currents.

    If you need "0V" at the input when it is unconnected, then I would recommend a CMOS input device, such as the TLV7021.

    The bias currents are in the pA range, and will generate tiny voltages across the input resistance. You can also reduce the size of the input series resistance (R5).

    You can try the LM393, but I recommend trying the TLV7021.

  • Hi Paul,

    Thanks. I will give the TLV7021 a try, we have the LM393 on hand to see how it works too. 


  • Thanks Alvin

    Let us know if you need any further assistance.

  • We have not received a response on this thread in a while, so I will be closing the thread.  If you need further assistance, feel free to to reply or open a new thread.