OPA2171 Input interaction I don’t understand?

Thomas,

The OPA2171 has clamps between the input terminals to protect the input transistors against large differential input voltages. Many types of op amps have similar input clamps and are not suited to use as a comparator.

The type of transistors used in the OPA2251do not require input clamps and thus function successfully as a comparator in your circuit.

Regards, Bruce.

Bruce

Thanks for the reply.

I did not see this in the data sheet what is this list as?

Tom

Thomas,

I regret that many of our data sheets (and those of our competitors) don't always make it clear that an op amp has differential input clamps. Our presumption is that an op amp will be used as an op amp, not a comparator. When used as an op amp, it's generally not noticed that these clamps are present (though there are exceptions). I admit it, I have used an op amp as a comparator (usually a case of one channel in a quad package). We should be more diligent in explaining this.

In general, op amps with NPN transistors have input clamps. Examples would be OPA227, OPA277, etc. An exception would be the old '741 which has NPN input transistors but additional PNPs in series that provided inherent protection of the NPNs

Op amps with lateral PNP inputs generally do not have input clamps. OPA2251, OPA244, OPA234, LM324, etc. These can often be identified as the input bias current is listed as a negative number.

JFET and CMOS amplifiers that operate on higher voltage (greater than 20V, or so) may or may not have clamps. It depends on the characteristics of the process and the particular transistors used.

Most low voltage CMOS op amps do not have clamps. There is a special exception for auto-zero or chopper types that may have behaviors that look like clamps.

The bottom line... if you use an op amp as a comparator, know your op amp and validate its behavior. If in doubt, ask us.

Hope this helps and regards, Bruce.

Bruce,

 Thanks for the fast reply!

 This helps to the point that I now understand what is going on but as you know now it brings up the issues of what do I do.

 This is funny though I have been doing EE work for 20 years and have used op-amp 100s of time as comparators and never had this issue.

 Will need figure out what to do with it now!

 Thanks for all the help!

 Tom

Tom,

So far as your previous designs... you may have been lucky in your selection of the op amp. Or, your differential voltages may not have been enough to turn on the clamps. Or, perhaps the clamps worked as they should but it did not matter that the voltages were altered.

You are using both channels of a dual as a comparator. Why not select a dual comparator rather than an op amp?

Bruce

Bruce,

That is what I'm doing now I'm looking at the TLV3702-Q1.

I need the push-pull output because I cannot put pull up resistors on the board at this time.

What is your though?

Tom

Bruce,

As it turns out the TLV3702-Q1 is a non-stock item by Digi-Key but they have some in stock at this time (i dont get that).

The non-stock always gets me I need something that I can get a year from now.

So in looking farther I found the LMC6762.

This looks like it would work just fine.

What is your though on these two?

Tom

Tom,

I don't have much personal knowledge of either of these parts but it appears they meet your basic requirements. You have not provided any other details that would appear to disqualify either product. Certainly a comparator will give you a much better speed/power ratio but I'm not sure that matters in your application.

The "Q1" version of the TLV3702 is an automotive version and may be more difficult to find in distribution. (More expensive, too.) I'm not sure if you really need this version.

Bruce

Thomas;

This question comes up now and then, so a long time ago I created a PowerPoint presentation for our Burr-Brown field applications engineers about using op amps as comparators. You may be interested in the test data, so I'll attach it.

Op amps have a few advantages over true comparators-- in general, their input offset voltage is lower and the output voltage can have a large bipolar swing, depending on power supply voltages. They are far slower, however, so they are NOT suitable for high-speed applications. On the other hand, their output slew rate is lower so they generate less RFI than a comparator.

Neil

Would you see any issues with using the OPA2251 as a comparator?

This is the op-amp I use for my prototype but it got changed do to cost.

 Tom

Thomas;

Bruce can better answer this question; I have no test data on it-- sorry.

Thank you

Bruce and Neil

Tom

Tom,

The OPA2251has lateral PNP inputs (notice that the input bias current is negative). Thus, it does not have a differential clamp.

You will find that many op amps made on a bipolar process that can swing close to the rail on the output have another property that may affect you:  Their quiescent current goes up substantially when the output hits the rail. This is because the output stage must overdrive the base of the output transistor to swing close to the rail. On op amps, only a modest effort is made to control this current increase. On comparators it is a much more important issue. I don't know the magnitude of this current increase on the OPA2251 and I don't know whether it affects you. I suggest that you check it out.

Regards, Bruce.