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.

LM393: Offset Voltage

Part Number: LM393
Other Parts Discussed in Thread: , , LMV762

I have a design that requires low input offset voltage.  I have used the LM393 with success in the past for this design for many years.  On my latest batch of boards the boards tested and the final product tested fine during production.  After an hour running at the customer they started to fail.  I have found that the offset voltage is greater then the spec allows.  The typical is 2mv, max is 5mv and absolute worst is suppose to be 9mv at 125C.  I am measuring 10mv to 12mv at room temp on the sample I have.  I could move to the LM393A as the specs are better but how do I know they are going to be in spec when the ones I have now are not.  I could see how these would have passed your initial inspection as they passed ours initial inspection.  I have at least 750 of these suspect parts populated on boards.  I can send you whatever other information you need to make this right.

  • Few more details. LM393 is powered at 3.3v. Single supply. Negative input is connected between two 10K resistors to make a voltage divider of the 3.3v supply. This makes the negative input at about the center of the 3.3v supply or 1.65 volts. Positive input is connected to 15 turn 10k pot. This side of the pot is also connected to 510K resistor which is then tied to the 3.3v. The other side of pot is connected to the 10K voltage divider. This makes a precise voltage difference between the positive input and negative input that is always positive. This emulates my actual circuit but is on a demo board so I can isolate the device.
  • Hello Cortland,

    Can you provide the full part number that you are using? There are two versions; the LM393 and the LM393-N, the latter being the former National Semiconductor version. They are different "under the hood" and supported by different groups.

    It is also helpful to see a schematic - including input stages and what the output is driving....and supply bypassing.

    Are you looking at the output with a scope or DMM? Do you have hysteresis added in the actual circuit?

    If you need low offset, consider the LMV762, with a worst-case offset of 1mV at 125°C and the same pinout. Note that it is faster and has a push-pull output (no need for output pull-up).

    Regards,

  • Cortland,

    The VIO on LM393 is specified at common mode at GND. VIO at other valid common mode voltages will be nearly the same as GND input VIO.

    LM393 common mode upper range is VCC - 1.5V at 25C and VCC - 2V at cold temperature which is 0C.

    At mid rail, common mode is VCC - 1.65V which could be a problem especially is 3.3V VCC is running on low side. Or the device is running at 0C.

    When inputs are different voltage, the device see the lowest input voltage as the common mode.

    As you stated it, positive input is at least 3.23V and negative input is 1.65V (assuming an exact 3.3V supply)

    This is really close to the upper range for 25C. 

    How are you verifying VIO? Can you explain in detail?

  • I think you miss understood the explanation. I hope to have more time to right up a better explain this week. Been busy replacing bad parts.