Sine to square wave conversion using LT1016 comparator

Discussion in 'General Electronics Chat' started by bt316, Oct 9, 2015.

  1. bt316

    Thread Starter New Member

    Aug 17, 2015
    4
    0
    Hi All,

    I am trying to convert a 500khz sine wave signal to square wave using the LT1016 comparator. I have just built a basic comparator circuit with the inverting grounded(pin3) and applying a 2v sinusoidal signal to the non-inverting(pin2). The latch and ground pins are grounded (pins 5 and 6). I am using it with a single +5v supply with the negative supply (pin 4) grounded too. However, in the output(pin I am seeing a large offset of ~2.6v. I am not ure what is wrong with the circuit. Can any1 help please?
     
  2. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,234
    I suspect the sine wave goes below ground, which means you also need a negitive supply.
    The alternitive is to AC couple it and bias the comparator to 1/2 the supply voltage.
     
  3. OBW0549

    Well-Known Member

    Mar 2, 2015
    1,308
    883
    You say you're using the LT1016, but the data sheet you posted is for the LT1116. They are not the same. Which are you actually using? Either way, your input is extending outside the allowable input voltage range and will give you unpredictable results; that needs to be fixed.
     
  4. bt316

    Thread Starter New Member

    Aug 17, 2015
    4
    0
    Thanks alot that seems to work!
     
  5. bt316

    Thread Starter New Member

    Aug 17, 2015
    4
    0
    Sorry my bad! I attached the wrong one! I am using the LT1016!
    Thanks for the reply!
     
  6. OBW0549

    Well-Known Member

    Mar 2, 2015
    1,308
    883
    For an LT1016 operated on a single 5V supply, the input voltage range is from 1.25 volts to 3.5 volts. This applies to both inputs, at all times. Go outside this range on either input, and your comparator output becomes "who the heck knows?"
     
    atferrari likes this.
  7. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,234
    I think @OBW0549 is right. You would be safer to reduce the input voltage slightly. Having said that I think it still works as a zero crossing detector since the signal is back inside the common mode range where it crosses the threshold.
     
    OBW0549 likes this.
  8. OBW0549

    Well-Known Member

    Mar 2, 2015
    1,308
    883
    Yeah, but what about when the signal is outside the common mode range? Opamps and comparators have been known to do some pretty bizarre things sometimes (like momentary phase inversion, for example) when you go outside the limits. Best to stay inside them.
     
  9. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,234
    I think it's ok.

    upload_2015-10-10_7-37-35.png
     
  10. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    ronv, the op is using a single +5v supply, so the input range is 1.25v to 3.5v as has already been stated. Go outside that range, and the output gets dicey. Reducing the input voltage is not a correct option.
     
  11. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,234
    I still don't see the problem with either option. He has a two volt peak to peak sine wave. If he biases it to 2.375 he doesn't go outside the common mode range. If he were to reduce the amplitude to say 1.5 volts he would have more room.
    Even if he goes outside the common mode range but stays above the low, and below the high side supply the spec says there is no damage or latch up.
    Where did I go wrong, go wrong?:)
     
    OBW0549 likes this.
  12. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
  13. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,234
    I see the problem when the input is above the supply. Is that what your talking about?
     
  14. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    It also talks about running it on a single 5V supply, and what the common mode range is...

    Lots of info about bypassing, layout, etc...
     
  15. OBW0549

    Well-Known Member

    Mar 2, 2015
    1,308
    883
    You're right: I didn't see the "1.25V above the negative supply" part and was only going by the "+/- 11V" value in the table for +/- 15V operation.
    AN72 is about the LT1394 comparator, and doesn't discuss the LT1016 except in the introduction. AN13 discusses the LT1016 but doesn't have much to say about common mode range other than the need to keep signal voltages inside it to avoid spurious outputs.

    In the end, I think ronv is right: if the TS biases his inputs up to a 2.375 volt center voltage (instead of ground, as he has it now) and doesn't exceed 2 volts peak-to-peak on his non-inverting input, he should be OK. 1.5 volts peak-to-peak would be better, as it would give him a bit more safety margin.
     
Loading...