An Op-Amp wired as a comparator can change many wave form (including sine wave ) into square wave but I asked about correct feedback resistor and input resistor values of an op-amp here and we came to the conclusion that these values should be small to avoid capacitance effects.

In an Op-Amp in comparator design, the feedback resistor is around tens of mega ohms and this causes many parasitic effects as well as low pass filter effect that diminishes the output.

In practice, I used AD826 as a comparator and I can see it makes a clean square wave in frequencies less than 1MHz but distortions start in higher frequencies until the output is nothing except a background noise in 5MHz.

What is the best design for an Op-Amp as a comparator in higher frequencies? I am using this simple design (pin numbers are not related to AD826):

enter image description here

  • 5
    \$\begingroup\$ Without being snide, you don't use opamps as comparators at high frequency. You use comparators. There optimized for such function. That's why they're there. \$\endgroup\$ – Scott Seidman Oct 10 '13 at 13:43
  • \$\begingroup\$ Thanks!I didn't know that. I thought comparators are kinds of op-amps. Do you know a good part number for up to 50MHz? \$\endgroup\$ – Aug Oct 10 '13 at 13:49
  • \$\begingroup\$ analog.com/static/imported-files/data_sheets/… I haven't used it, but it looks very fast. You may need to convert the Emitter coupled logic. max9691 is also very fast. \$\endgroup\$ – Scott Seidman Oct 10 '13 at 15:31
  • \$\begingroup\$ @ScottSeidman Thanks! Please put this comment as answer that I can mark it as the correct answer. \$\endgroup\$ – Aug Oct 10 '13 at 20:25
  • \$\begingroup\$ No need, Aug, you got pretty much the same advice below. Pass the rep to a newbie! \$\endgroup\$ – Scott Seidman Oct 10 '13 at 21:01

Use a MAX999 comparator.

It has rise and fall times of 2.3ns. A 50MHz input has a period of 20ns so this would be "OK" for the MAX999 but don't expect miracle-fast waveform shapes at this frequency.

enter image description here

  • 1
    \$\begingroup\$ You don't explain why the comparator is better than an opamp. \$\endgroup\$ – Johan.A Oct 15 '13 at 14:49
  • \$\begingroup\$ @Johan.A - Clearly, in the question, the OP knows what a comparator is but is unaware of the capabilities of the MAX999. \$\endgroup\$ – Andy aka Oct 15 '13 at 14:58
  • 1
    \$\begingroup\$ I don't agree. He specifically asks about an opamp as comparator. He doesn't ask about any comparator. \$\endgroup\$ – Johan.A Oct 15 '13 at 15:19
  • \$\begingroup\$ @Johan.A - You say I don't explain why the comparator is better than an opamp, then you say the OP doesn't ask about comparators - these statements are not consistent. The application is best suited to a comparator and my answer reminds the OP of this by showing a comparator that easily does what he wants. The OP accepted my answer proving that my observations were correct. If, you feel that my answer hasn't touched on sufficient detail then I'm very happy for you to post an answer and I'll ask the OP to reconsider his acceptance of my answer. \$\endgroup\$ – Andy aka Oct 15 '13 at 17:56
  • \$\begingroup\$ Like I said: he asks about using an opamp as comparator. This has issues, like pebbles indicates. Your answer doesn't touch that. Instead you offer the MAX999, which is not an opamp. I don't think my previous comments are contradictory. But for me OP doesn't have to reconsider the accept: if he's happy with your answer that's fine. Peace. \$\endgroup\$ – Johan.A Oct 15 '13 at 19:26

You can't do that. Opamps are frequency-compensated to keep them stable (read: avoid oscillating). That limits the gain-bandwidth product, so that for the high gain needed as comparator the bandwidth will always be low.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.