# Common-mode voltage outside the specified range

I'm reading the datasheet of TL331. Here a summary of the relevant specification for my question:

ABSOLUTE MAXIMUM RATINGS
VCC Supply voltage 36 V
VID Differential input voltage ±36 V
VI Input voltage range (either input) –0.3 V to 36 V

ELECTRICAL CHARACTERISTICS
VICR Common-mode input voltage 0 to VCC – 1.5 V


with this note:

[...] The upper end of the commonmode voltage range is VCC+ – 1.5 V at 25ºC, but either or both inputs can go to 30 V without damage.

I'm going to use this component as a comparator (i.e. without any feedback) in this situation:

• single power supply: 12V
• IN-: 9V
• IN+: 18V

Now, the absolute maximum ratings are satisfied and also the second clause of the note. But the common-mode voltage is 13.5V, and this is outside the specification (VCC-1.5V = 10.5V).

So, I cannot use this amplifier at all? Or, because I'm going to use it as a comparator it would work anyway? I mean, the outputs would saturate without actually be able to make V+ = V-.

• Sorry, I'm not sure to understand. I'm going to use it as a comparator, indeed. The output will drive the diode of an opto-isolator with 12V (and a resistor) to its anode. – Mark Oct 8 '18 at 15:25
• Still not understand. I know its an oc output. What would prevent me to connect a series of a resistor and a diode, and connect the anode to +12V? I said I have no feedback, hence I'm using it as a comparator, what it's designed for. Am I wrong? – Mark Oct 8 '18 at 15:31
• If you know about all this, then its fine, its just that when people talk about dedicated comparators as amplifiers, its often an indication they mixed these things up and expect that thing to be able to output significant current. – PlasmaHH Oct 8 '18 at 15:34
• My guess would be that it's designed to survive common-mode voltages up to 30V, but the manufacturer doesn't guarantee that it will work correctly under those conditions. I would recommend looking for a different comparator. – Hearth Oct 8 '18 at 15:35
• @Mark -- if you're making a few, you don't need to save components. If you're making a ton of these, and need them to work, most engineers wouldn't go with that part selection under your use scenario. – Scott Seidman Oct 8 '18 at 15:45

There wouldn't be an a priori way to determine what the failure mode is.

The good news is that since you're not exceeding absolute maxima, it's fairly safe to say that the failure mode will be repeatable, but that doesn't help you know in advance what it is.

My guess is that under your use scenario, the manufacturer would be wary about timing issues. I suspect lock up will be an issue. Will it be within your needs?? Hard to tell, as I don't know what they are. There may also be thermal issues, and bias current issues. If this is important, serious testing would be in order before you run with your proposed design. Such testing is expensive with regards to time and money.

I suppose my biggest point is that you've done your job well, and identified that this part isn't for you, unless you somehow tweak your design (like with the zener I mention in comments).

• I'm looking through the documentation of tons of comparators and op-amps but I cannot find one that would satisfy this scenario. But please take a look to this question: electronics.stackexchange.com/a/329761/122430. It seems (at least that specific IC) would, even if the ds not is the same. – Mark Oct 8 '18 at 16:01
• @Mark that part looks like it will work, so long as both inputs stay below +36. Barring a data sheet entry similar to that, I wouldn't assume the TI part would work. The way to get a real answer that you can take to the bank is to contact the TI engineers. Start with a sales engineer, and say "the data sheet seems a bit ambiguous to me" – Scott Seidman Oct 8 '18 at 16:13

Most comparators (including the TL331) require that only one of the inputs be within the common-mode range for correct operation. The data sheet schematic reveals the reason for this. Because the input differential amplifier uses PNP transistors, the transistor input with the higher voltage will be off; it doesn't matter if it exceeds the common-mode voltage range. As long as the input with the lower voltage is within the common-mode range, the output will be correct.

It will function, provided at least one of the inputs is within the common mode range, as @user28910 indicates. This is spelled out explictly in 9.2.2.1 Input Voltage Range. However I would expect the timing might be degraded, perhaps significantly. The TL331 is not particularly fast and is not even guaranteed to be any particular speed so you can say it will meet the specs (since they are silent on guaranteed response time).

If you want guaranteed performance you can look at a part such as the LT1716 comparator.

"Input Common Mode Range Extends 44V Above V–, Independent of V+"

However it's a boutique part and won't come as cheap as the jellybean TL331, maybe 5x as much.

Edit:

Since you're looking for very low frequency response (1kHz, so many microseconds response time will be okay) , I don't think you'll have any trouble using the part.

• Thanks. I would add the maximum input frequency (is a square wave) is 1 kHz. – Mark Oct 9 '18 at 5:49