Connect with us

LM339 comparator input levels

Discussion in 'Electronic Design' started by David Novak, Feb 28, 2007.

Scroll to continue with content
  1. David Novak

    David Novak Guest

    I find two input voltage specifications in the LM339 data sheet, Vidr
    (input differential range = 36V) and Vicmr (input common mode range,
    -0.3V to Vcc). I understand Vidr, but how do I interpret Vicmr? Is this
    stating the voltage range of the inputs if they move together at the
    same rate? In other words, the DC offset that effects both inputs equally?

    Since the Vidr specification does not refer to Vcc, it appears that I
    can set IN- = 1V, Vcc = 5V and let IN+ swing from 0V to 36V. Is this
    correct? I am worried about letting either input exceed Vcc.

    Thanks,
    David
     
  2. Jamie

    Jamie Guest

    yes, i think you have it.
    CMR = Common mode rejection.
    Simple test.
    Take both inputs driven by the the same source with the unit
    set up to have equal gain on both inputs.
    the Rejection ratio is that of the difference that changes as you
    increase the input.
    the ideal Op-amp would have 0 change on the output from low to high
    on both inputs.
    ----
    and yes, you can allow it to swing to 36 volts.

    and if the question ever comes up about offset voltages.
    that usually means if both inputs are set to 0 for example.
    in a +/- Dual rail supply, the offset is usually the maximum
    voltage that should appear at the output.
    but then again, i've seen batches of cheap op-amps actually
    perform far lower in the offset than specified in the specs.
    THe LN741 has offset inputs for bias to balance that problem, that
    was nice to have to tweak up the problem how ever, you had to maintain
    a steady temperature around the chips once set.
     
  3. You are correct. It is fine to exceed Vcc with either
    input. This device is fairly unusual in this respect.
     
  4. Eeyore

    Eeyore Guest

    No you *can't* !

    Don't exceed Vcc on either input.

    Graham
     
  5. Eeyore

    Eeyore Guest

    Taking a look at the data sheet you would appear to be right. In which case the
    common mode input votlage spec is slightly misleading.

    Graham
     
  6. Not really, Common mode voltage range covers the comparator
    effectively deciding which of the two inputs is more
    positive. Differential voltage range has nothing to do with
    comparison, except that it remains correct (no phase
    reversal) as long as at least one of the two inputs also
    remains inside the common mode voltage range.

    This ability of the lateral PNP input transistors of the
    LM339 and LM393 (dual) can come in very handy when the
    signals are not derived from the same supply as the one
    feeding the comparators (i.e. 24 volt field signals being
    converted to 5 volt logic levels).
     
  7. Jon

    Jon Guest

    Vicmr is the absolute maximum voltage range, beyond which the device
    will be damaged. The operational common mode range is given in the
    "Electrical Characteristics" section of the data sheet, and is the
    input common mode voltage range over which operation is guaranteed.
    This range is given in terms of Vcc. Unfortunately, the LM339 data
    sheet does not give the Common Mode rejection ratio (CMRR), so you can
    not determine how much the input offset voltage will change over this
    common mode input range.
     
  8. David Novak

    David Novak Guest

    Thanks for all the help! I received the following from an ON Semi FAE.

    "On page 6 of the data sheet it states (clarifies) that Vdif can indeed
    be greater than the supply, as long as neither input goes negative by
    more than .3 V.

    The common mode range is the input voltage where the part will function
    properly. This device has PNP inputs so once the input signal
    approaches the positive supply rail (within 1.5 - 2.0 V)the transistors
    turn off. So if you have a 5 V supply and the inputs are 0 V and 36 V,
    it will work ok since the Vdif is met and one input is within the common
    mode range. If the input signals are 5 and 36 V, Vdif is met but the
    common mode isn't so the output will be unknown. Neither of the input
    transistors are on so the output is not controlled. However, the part
    won't be damaged."

    I had completely missed this page in the data sheet. It is worth noting
    that the previous paragraph on the same page recommends input resistors
    of < 10K and a < 10mV positive feedback.

    David
     
  9. Jim Thompson

    Jim Thompson Guest

    Your last paragraph is correct.

    However the specification is confusing. Inputs CAN exceed VCC but,
    for proper comparator decision output, ONE input must be below VCC by
    about 1.5V. (The National data sheet shows this.)

    ...Jim Thompson
     
  10. Jamie

    Jamie Guest

    What an asshole, Go pound sand you puke head.
     
  11. Jim Thompson

    Jim Thompson Guest

    I knew Eeyore was ignorant, but I guess I mistakenly presumed he could
    at least read a data sheet?

    ...Jim Thompson
     
  12. joseph2k

    joseph2k Guest

    Study the datasheet. Graham is correct on this one. It really pays to note
    the test conditions for each paramater.
     
  13. Eeyore

    Eeyore Guest

    The datasheet is slightly confusing in this respect actually. Note 8 IIRC
    explains in more detail. It's really a question of being very precise about the
    measurement conditions.

    Graham
     
  14. Jim Thompson

    Jim Thompson Guest

    Except that Graham is NOT correct.

    "Study the data sheet"? Indeed!

    In the positive direction, as long as at least ONE input is below
    VCC-1.5 the comparator will output correct results. The other input
    can be well above VCC, but can not exceed 36V.

    ...Jim Thompson
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-