measurememtns

[leonski]

We have become a measurment dependent society.      Cars must have so much HP and be so fast to 60mph and so on.     Even the cabin must be a certain DB quiet.

In stereo we depend on VOLTS and AMPs and many other mesurments.    

In other threads I have touched on some issues with measurement but I'd like to expand that.

In a stereo shop, you may have several voltmeters.    A DVM or in the old days a VTVM or VOM.    All measure voltage, resistance and have current functions, too.   

But NOBODY talks about accuracy, repeatability or the a REAL the real boogie man, that of CALIBRATION.    Where I worked, meters were annually exchanged for

fresh calibraded meters.    Some other equipment was calibrated by specialty houses which calibrated to NIST standards.    Our tempreture themocouples for

furnaces came with deviation printouts in 1/10 degree increments for the entire range of the thermocouple.   THese were replaced on a regular basis and it was

a task of mine to enter the new data into the computer where the part was used.   THAN one of the maintenance guys would come along with what was called

'The 8mv supply'' which was supposed to read exactly some fairly high temp.    We'd press a couple buttons and the machine would snap right in.  And than would

be checked periodically.

The lesson is that a LOT of what comes down to us as specs are probably taken with only so-so calibrated equipment.    Even internal standards are better than

nothing.    I would read SOME machines 3x daily.....once per shift....and record the results on a control chart.  If the machine went nuts, we'd know pretty quickly.

I was called as the first-responder IF there were any questions.    

Since I was so heavily involved with the quality end of things, I guess I'm still a little obsessive about such things.  

[vcautokid]

Sorry the title threw for a moment. "Measurmemtns"

Interesting post. Test equipment can drift and needs to be certified calibrated. I don't know quite as much today, as I did maintain our test equipment back in the 80s. I maintained a good relationship with General Radio in the day. Known as GenRad, we had machine that got certified and calibrated if needed periodically to maintain quality.

[leonski]

There are So MANY factors in measuring something. Here's one many would not have thought of.

You are going to measure a medium to high power amplifier. You'll measure ONE channel. Everything is plugged in and you are measuring away.

Do you Control power voltage TO the amp using a Variac? It might matter especially when the amp starts dragging the line voltage down some. In extreme cases

the voltage may go under 110 or so. Are you getting 'valid' readings of what the amp can do OR is this 'real worl'd' reflecting use in a real system?

We can worry about the Calibration Stickers on each piece of measurement gear in some later post. This is NOT a trivial expense if you have a big setup or multiple

technicians with equuipment needing calibration.

[Boomzilla]

For some things (industrial hygiene air monitoring equipment, radiation badges, covid 19 tests, etc.) calibrated accuracy is VERY important. For manufacturing companies whose products MUST be consistent (pharmaceuticals come to mind immediately), calibrated accuracy is vitally important.

But for a home enthusiast, how important is it, really to have ultra-accurate measurements? If I'm measuring the bias of a vacuum tube, accuracy to plus or minus a tenth of a volt is normally sufficient. Even the cheapest "credit card" VOMs better that by an order of magnitude. And once factory-calibrated, there just isn't much drift.

That wasn't always so - Even a decade or two ago, even the "test instruments" had to be calibrated before each use. I've spent many an hour calibrating LEL (lower-explosive-limit) meters before use. But anymore, most critical meters are self-calibrating and automatically verify their own accuracy before & after each use.

But I wander away from home use... For the hobbyist, I contend that modern instruments are factory-calibrated, and drift so little that they can be used for decades and still meet their tolerance ratings. Even inexpensive power conditioners and surge suppressors anymore offer digital displays of voltage on the supply line. Many also have ampere meters that indicate the total current draw of items plugged into them. Are they "laboratory accurate?" Probably not. But for home use, they seem pretty darned good.

Boomzilla

[leonski]

The main issue may NOT be ultimate accuracy to a standard but rather REPEATABILITY and CAPABILITY.

Manufacturers making various claims may desire or even require both high accuracy AND high repeatability. Someone in another thread mentioned a Gage R+R study.
In that technique, 3 persons Indenpendently measure samples and record the results. In doing these studies myself, sometimes one operator is way off the others in measured results.
Sometimes ALL 3 operators measure in such a way that it looks like 3 groups. Once the data is collected, the calculations are made and Measurement Tool capability can be calculated.
Not how 'accurate' it is.

When I did this sort of thing for a living as part of my tasks, we made IN-HOUSE standards which were internally vetted and used. Some machines were checked 2x daily and the measured
result of OUR standard entered in a control chart. That's when things could potentially get weird.

My remarks, above, were aimed primarily at manufacturers, NOT home users. I personally have a CHEAP amp clamp and a Kill-A-Watt meter. Neither are particularly calibrated. And are of unknown repeatability.....

I disagree with most of your last line. Unless you are willing to accept a drifitng value over time. I think a small 'niche' market may exist for a product used to check certain meter functions. For example? I'd trust
a selection of 1% or better resistors to check THAT function of my meter.