12
\$\begingroup\$

enter image description here

This schematic appears in a recent question. The triangle may be an op-amp or a comparator, because they both have the same connections. (Why) isn't there a separate symbol for a comparator? The functions of op-amp and comparator are very different.

Or from the document @boardbite writes about in a comment:

enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ An interesting app note from TI related to the "similarities" and the differences (still no etymology though) \$\endgroup\$ – boardbite Sep 3 '12 at 14:18
  • \$\begingroup\$ Frankly, I'm glad there is no separate symbol. There is already too many of them. The app note above shows a very nice comparison - the insides of a op amp and a comparator are very similar. \$\endgroup\$ – Jonny B Good Sep 3 '12 at 16:54
  • 2
    \$\begingroup\$ @Jonny - If there are too many symbols it's because of duplicates for identical parts, like US and IEC resistors, several capacitor, current source and logic symbols, etc. For me an opamp and a comparator are different enough (do you really think they're very similar???) that the comparator at least deserves an indication. \$\endgroup\$ – stevenvh Sep 4 '12 at 10:49
  • \$\begingroup\$ @steven: I think they're simillar enough - and since they haven't changed for so many years, apparently the majority of people involved in the industry would agree, or likely do not care. If there was to be a proposed new symbol, it would be very hard to enforce its use. One organization will pick a new symbol, a different one will use their own, and suddenly we have a whole new set of unnecessary symbols. Just like in the case of voltage/current sources, analog/digital signal buses on block diagrams, etc. In principle, different elements should differ, but there should be only one. \$\endgroup\$ – Jonny B Good Sep 4 '12 at 12:53
  • 2
    \$\begingroup\$ @Jonny - Similar enough? Well, then replace the opamp in that amplifier by a comparator -- just for laughs. "suddenly we have a whole new set of unnecessary symbols". The idea is of course that everybody listens to me! :-) \$\endgroup\$ – stevenvh Sep 4 '12 at 12:59
10
\$\begingroup\$

I've never seen a different symbol for comparators, so I had to make up this one:

enter image description here

The hysteresis symbol refers to the hysteresis which is often built-in into the IC, or otherwise is almost always needed in the circuit. Does this make sense?

edit
Olin objects. He's right: it suggests a built-in hysteresis, rather than a required one. So second attempt:

enter image description here

\$\endgroup\$
  • 2
    \$\begingroup\$ Do real-life comparators include hysteresis? The LM339 apparently does not, the data sheet even shows a circuit that adds hysteresis. \$\endgroup\$ – Wouter van Ooijen Sep 3 '12 at 15:39
  • 1
    \$\begingroup\$ @Wouter - Some do, but I don't use comparators frequently enough to have a good idea how many. \$\endgroup\$ – stevenvh Sep 3 '12 at 15:45
  • 1
    \$\begingroup\$ That datasheet even shows the symbol with hysteresis :) \$\endgroup\$ – Wouter van Ooijen Sep 3 '12 at 15:50
  • \$\begingroup\$ @Wouter - Heck, yes, I hadn't noticed that :-), (but only once, in most schematics they use the "opamp" symbol) \$\endgroup\$ – stevenvh Sep 3 '12 at 15:56
  • 3
    \$\begingroup\$ I think this is a reasonable symbol only if the comparator really does have hysteresis built in. Otherwise it would be misleading. \$\endgroup\$ – Olin Lathrop Sep 3 '12 at 20:46
8
\$\begingroup\$

I'm not an expert on using op-amp or comparators but from my understanding, many op-amps can be used as comparators, the main difference is that the op-amp uses feedback to define the amplification value and the comparator doesn't care much about a linear output since usually we only need a high or low value from it but essentially they perform quite similarly.

\$\endgroup\$
  • 4
    \$\begingroup\$ I'd rephrase as "many op-amps can be used as comparators"...Some won't work well as comparators becuase they have slow recovery from saturation conditions. Others might even draw damaging currents when their input voltages are allowed to diverge. \$\endgroup\$ – The Photon Sep 3 '12 at 15:31
  • \$\begingroup\$ AFAIK the same (ideal) component serves both functions, hence the same symbol is used for both. An (ideal) op-amp does not need to be used with feedback. \$\endgroup\$ – Wouter van Ooijen Sep 3 '12 at 15:37
  • 1
    \$\begingroup\$ @Wouter - the opamp may be usable, but performs far worse. Some comparators are Damn Fast™, like < 500 ps propagation delay. An opamp has frequency compensation and therefore will be much slower. \$\endgroup\$ – stevenvh Sep 3 '12 at 15:50
  • 1
    \$\begingroup\$ @stevenvh I'm not claiming that using a comparator is the same as using a op-amp. The question was about the schematic symbol, and ideally both work very similarly hence the same symbol. They are used in different situations so they have different requirements. \$\endgroup\$ – Bruno Ferreira Sep 3 '12 at 15:57
  • 3
    \$\begingroup\$ @stevenh compare the situation of a hf-transistor versus a 3055-type slow-but-high-current transistor. They have the same symbol, but no one in his right mind would use one when the other was required. \$\endgroup\$ – Wouter van Ooijen Sep 3 '12 at 16:12

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.