Author Topic: Calibration test of RS 22-812 -- how accurate is an inexpensive meter?  (Read 5796 times)

0 Members and 1 Guest are viewing this topic.

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9693
  • Country: us
There's a lot of discussion around the forum on accuracy, calibration, inexpensive meters, and so on, so I thought it would be fun to look at one of my favorite inexpensive meters to see how it measures up. The meter in question is the venerable Radio Shack model 22-812. I've posted before about why I like this meter (solid build quality, large clear display, computer interface for data logging). Here, I'm curious about how it might behave in terms of measurement accuracy. I have previously adjusted the RS meter against an independent voltage reference (not the 3456A). That was some months ago and I have not touched it since then.

The test I performed is somewhat narrow in scope, looking at DC volts over a moderate range typically encountered in low voltage circuits. I was mainly interested in the statistical picture that would emerge, so I did not look at other ranges like AC volts, current, or resistance.

To perform the test I put the 22-812 alongside an HP 3456A and compared readings with steady voltages generated by my TTi EL302P power supply. For each pair of readings I looked at the percent difference between them. I don't know the exact calibration status of the 3456A, but based on a few comparisons with other references it seems to be about right.

When considering the results, it is instructive to consider the best case accuracy that a 4000 count meter like the 22-812 can achieve. At full scale, the true value could be 3999.5 counts. The best the meter can indicate is either 3999 or 4000, either way it has a display error of 0.5 parts in 4000, or 0.0125%. This represents the best the meter could possibly achieve. At the lowest end of the scale the true value could be 400.5 and the meter could indicate 400 or 401. In this case the display error is ten times greater at 0.125%.

So if the meter keeps its readings within 0.125% of the expected value we can say it is doing as well as it can.

Plotted graphically, the results look like this:



The mean signed error is -0.007% with a standard deviation of 0.035%. The mean absolute error is 0.027% with a standard deviation of 0.025%.

The worst error is about 0.1% with the readings mostly falling within 0.05%.

It is interesting to see a kind of cyclic S pattern to the errors, suggesting a degree of non-linearity in the instrument.

In case anyone is curious the full table of data is shown below.

Voltage readings compared between RS 22-812 (adjusted) and HP 3456A
HPRS
0.10240.1023
0.19960.1995
0.30070.3004
0.39780.3976
0.49870.499
0.59750.598
0.69790.698
0.79670.797
0.89750.898
0.99570.996
1.99912.000
2.99843.000
3.99994.00
4.99735.00
6.00186.00
7.00587.00
8.01018.01
9.01059.01
10.006210.00
11.007911.01
12.00812.01
13.00413.00
14.00314.00
15.00915.01
16.01316.01
17.01317.01
18.01518.02
19.01119.01
20.01920.02
21.02021.02
22.02022.02
23.02623.03
24.02924.03
25.03025.03
26.02726.03
27.02927.03
28.02828.03
29.02329.02
30.02930.03
« Last Edit: July 21, 2013, 11:18:19 pm by IanB »
I'm not an EE--what am I doing here?
 

Offline Robomeds

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Thanks for posting the data.  I've generally found (but never verified with a bench meter) that even my HF meters have proven to be accurate enough for much of my meter needs.  Of course as a tool monger what I want and what I need aren't the same thing. 

I would add that a better meter will promise to meet it's specs over a larger range of temperatures and the like.  Also, part of the factory spec may assume either poor adjustment at the factory and/or mild variations in the trim pots over time.  It would have been interesting to see what the factory vs adjusted performance was.  Either way, the nice thing about your data is you have a good idea what you are dealing with.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9693
  • Country: us
Thanks for posting the data.  I've generally found (but never verified with a bench meter) that even my HF meters have proven to be accurate enough for much of my meter needs.  Of course as a tool monger what I want and what I need aren't the same thing. 

I would add that a better meter will promise to meet it's specs over a larger range of temperatures and the like.  Also, part of the factory spec may assume either poor adjustment at the factory and/or mild variations in the trim pots over time.  It would have been interesting to see what the factory vs adjusted performance was.  Either way, the nice thing about your data is you have a good idea what you are dealing with.

Both my samples of this meter were significantly off from the factory, by about 0.3% in each case. I found that disappointing, however I am pleased that after adjustment they seem to do pretty well.

When using a meter, it does not matter so much what the actual accuracy is, it is more important that the reading is predicable and that you have a good idea what to expect from it.
I'm not an EE--what am I doing here?
 

Offline Spunky

  • Regular Contributor
  • *
  • Posts: 63
  • Country: wales
Agreed, consistency if the key thing not accuracy or resolution. My most accurate meter is an old Vichy99 which is about 6 years old and cost under £30 off ebay. I can't fathom how it's been so accurate for so long and for so little money but it's stayed well within it's specs when others have drifted.

I think a lot of it also has to do with environment, if it's protected from cold, damp or heat it has a good chance. Every meter I ever lent to my dad would end up in the shed (damp) and be way out of spec by the time I got it back. It never bothered him as he only knew how to test continuity, but I was never amused. I know now only to lend him total shite.
 

Offline Robomeds

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Well sometimes absolute accuracy does mater.  I think it maters most when you need to see small changes and you can't assume you will use the same meter each time you test.  At my job a decade ago we had a number of the bench HP meters for that reason.  Sure the Fluke 83 I had at my desk was good for most stuff but when those last mVs counted it was nice to know they didn't change just because the meter changed. 
 

Offline Spunky

  • Regular Contributor
  • *
  • Posts: 63
  • Country: wales
Yes if you're working with millivolts I'd agree, I never needed to go to that level. But you're talking about some expensive meters here, I was thinking more along the lines of the nearest 0.1 of a volt, beyond that you can't be looking at any of the cheap brands.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9693
  • Country: us
Yes if you're working with millivolts I'd agree, I never needed to go to that level. But you're talking about some expensive meters here, I was thinking more along the lines of the nearest 0.1 of a volt, beyond that you can't be looking at any of the cheap brands.

If you notice, the $70 meter here is reading accurately to the nearest 0.01 V. The only time it errs is on the 400 mV range, when it is sometimes out by as much as 0.3 mV.
I'm not an EE--what am I doing here?
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4302
  • Country: ca
  • Canadian po
I am not trying to bad mouth that multimeter but I would like to interject something here. The only reason you know that meter is close to correct is because you first adjusted it to a high precision reference and are now checking it against another much higher priced and accurate meter. How far out was it when you needed to adjust it the first time?

A person buying any multimeter can only hope or rely on the reputation of the manufacturer to have adjusted the multimeter to meet its specifications before it left the factory. One cannot claim that based on a sample of one that it will be representative of the whole production run, whether it be a $5 meter or a $400 meter. That is why in reality any review that shows the meter meeting its specs with some reference is not really valid, mine included.

I have had a multimeter arrive almost ten times out of specification. It was adjusted and over the past year it has settled down to stay within its spec. If I was not better educated on the subject I would have quite happily believed that I was measuring and adjusting things to within 0.05% accuracy when it fact it was out 0.4%. It was not a "cheap  meter as far as price goes, so you can't rely on one meter and believe you have the truth.

So the 22-812 might be a good meter, but can a random person buy one and trust it to be even within the 0.3% you say your meters had when purchased? Probably, but you can't be sure until you can check it against another meter. It is almost a sure thing that a buyer won't get one that is within 0.125% like yours because you adjusted it to this accuracy. What your experience shows is that you can trust your personal meters to not drift too much over a few months, which is a good thing!
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9693
  • Country: us
Right, but I think you can trade experience and testing against expenditure. I can weigh an inexpensive meter with good knowledge of how it performs against an expensive meter where I have to trust the factory calibration (but which I will nevertheless try to verify before trusting it).

It was curious (and puzzling) that both samples of the 22-812 in my possession were out of adjustment by exactly the same amount from the factory, as if the factory calibration reference was itself out of adjustment. However, I adjusted them myself against the inexpensive DMM check 5 V reference, and about a year after that they both verify similarly against the HP meter that I only recently obtained.

What I like about such inexpensive meters is that I can adjust them, unlike the more expensive meters that I can't, or don't dare, fiddle with. It's a bit like the difference between older cars where you could tune them up yourself, and new cars that have black box electronics and only the dealer has the compatible diagnostic computer.
« Last Edit: July 23, 2013, 04:07:42 am by IanB »
I'm not an EE--what am I doing here?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 31414
  • Country: au
    • EEVblog
A person buying any multimeter can only hope or rely on the reputation of the manufacturer to have adjusted the multimeter to meet its specifications before it left the factory. One cannot claim that based on a sample of one that it will be representative of the whole production run, whether it be a $5 meter or a $400 meter. That is why in reality any review that shows the meter meeting its specs with some reference is not really valid, mine included.

That is the gripe I have when people hassle me for not doing serious "accuracy" measurements on meters I review. It is essentially a pointless test if it's within spec, but outside of spec is different though, that tells you something. Accuracy test one one-off meters simply encourage the clueless to use that as proof to validate that a cheaper meter is as "accurate" as an expensive meter costing 10 times more.
In practice you can certainly better rely on the reputation (and methods) of a major proven manufacturer.
Who would you trust, a $40 no name cheapie, or a $400 Fluke or $400 Agilent with a cal certificate with values? The answer is obvious.

And yes, if this Radio Shack meter was adjusted first, then that essentially invalidates the entire point of the test, not that's it's a really meaningful test to begin with anyway!
« Last Edit: July 23, 2013, 05:03:19 am by EEVblog »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 31414
  • Country: au
    • EEVblog
Right, but I think you can trade experience and testing against expenditure.

That comes down to the entire concept of calibration. It's all about "confidence" and measurement history.
You could use a $100 cheapie as a gold lab reference standard if you really wanted to, if you had the data to back it up to give you enough confidence to assert that.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9693
  • Country: us
And yes, if this Radio Shack meter was adjusted first, then that essentially invalidates the entire point of the test, not that's it's a really meaningful test to begin with anyway!

But every meter is adjusted first, whether you did it or someone else did it (at the factory).

What can then be learned, after adjustment, is information about things like time and temperature stability, and linearity. These aspects are interesting and educational in a meter of any cost.

It turns out that for much less than $400 I can make measurements to my satisfaction that suit my purposes. If it doesn't suit your purposes that's up to you, but I think everyone can benefit by having a better understanding and appreciation of their tools before using them.
I'm not an EE--what am I doing here?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf