Terms: "precision", "indication"

[jeroen79]

Quite often parts will be described as being a "precision" [resistor/opamp/whatever] or for some circuit a "precision" part may be recommended.

As I understand it everything will have a certain degree of (im)precision even if it is very very small.
So at what point does a thing become a precision thing?
How low % or ppm is required to deserve the label "precision"?

And on the other side of the 'spectrum' things may be called "for indication only".
For example the last digit on a DMM.
What does that mean exactly?
What is the difference between a 'proper' measurement and an "indication"?

Is there some consensus on these terms?

[Moon Winx]

I'm not sure about your first question, but it's probably due to marketing more than anything.

The second question, however, has a factual answer. Any measurement ALWAYS has an uncertainty associated with it (in fact, it's helpful to think of a measurement as a statistical distribution rather than a number), but sometimes you don't really care about the uncertainty because you're not using the meter's indicated value for anything. Maybe it's a sanity check, maybe you just care if you see a voltage or not, maybe you want to see if the voltage is within a certain range. If you're not planning on using your meter's exact displayed value for something, you can call it an indicated value.

[Macbeth]

Is the the accuracy vs resolution thing? "Indication" is a mealy mouthed term. "Precision" is not.

[retrolefty]

Is the the accuracy vs resolution thing? "Indication" is a mealy mouthed term. "Precision" is not.

 Without objective specifications, "precision" is just another subjective word.

[David Hess]

For some parts like operational amplifiers, "precision" is associated with specific implementation features like cross coupled quads for thermal balancing and active trimming.

From the buyers point of view these parts have better "precision" specifications like CMRR, PSRR, open loop gain, noise, offset, offset drift, etc. than a "normal" part would have and not because they are graded that way.

[metrologist]

Precision is mostly relative. The numbers matter...

I think of indication as binary, like a temp light in your car. It gets red when the temp is too high, but does not really tell you a specific number. My thought on the DMM last digit is that  usually adds +/- 1 to that reading. I thought that was typical with digital displays like that.

A reading is not a measurement.

[NottheDan]

At my employer's our scales can be run as "indication only" for two reasons. One, they're not calibrated (or not calibrated to the precision required), only tested. Which makes them considerably cheaper to run in internal jobs where ballpark weights are enough.
Two, they are not overseen by Trading Standards. We can use them wherever we want except for giving us weights the customers are billed for.

[ConKbot]

At my employer's our scales can be run as "indication only" for two reasons. One, they're not calibrated (or not calibrated to the precision required), only tested. Which makes them considerably cheaper to run in internal jobs where ballpark weights are enough.
Two, they are not overseen by Trading Standards. We can use them wherever we want except for giving us weights the customers are billed for.

Ah, reminds me of one fine day when inspectors griped at the greybeard in the electronics/cal lab about insufficient 'CNR' (cal not required) stickers on items not related to the cal process on the other side of the lab (electronics half), and he went off the deep end with the stickers.

Wall clock? CNR. Telephone with time on it? CNR. Fire extinguisher gauge, wall warts,  break room microwave, coffee maker, hot and cold taps in the bathroom, computers, printers, socket sets, adjustable wrenches, and a load of other things?  CNR.  He was quite smug when they came back.

[Edwin G. Pettis]

In the case of resistors, 'precision' is generally applied to tolerances of ±0.1% or tighter, such as ±0.05%, ±0.001%.  Capacitors on the other hand tend to have the term 'precision' applied at about the ±2% or tighter level, mainly because it is more difficult to achieve tighter tolerances on capacitors.  Inductors generally start at about the ±1% level for the same reason, difficulty in making tight tolerances.

The term 'precision' really depends on the subject matter as to the definition of when it is applied.  This applies to mechanical devices as well such as gage blocks for example.