Meter accuracy

[ConsultantERP]

I believe there is something very important missing from this articale. My background is in primary standard pressure calibration industry. This is not to say there is misinformtion in the article, but there is missing information. In particular, how can I find out how far OFF my meter is reading. Your example of using the standard thermometer to test the accuracy of the digital is what I am talking about. Question, when compared to a temperture standard (your thermometer) that is know to be accurate are you not actually calibrating that digital device to be more accurate. The answer is yes. The instrument that is said to be OFF can now be considered to be more accurate because of the comparison to a known standard (you thermometer). This is how the device got its accuracy rating in the first place (typically). If you own say a digital PH instrument, and you do not trust it, then it is a simple matter to obtain solutions that have a know PH of 0, 7, and 14, test your instrument with the solutions and note the difference in the readings of the instrument at each point. Graph this information and you now have an instrument that has know deviation from the standard and a little math brings the reading closer to accurate. Example, my Ph meter reads 0 with a standard solution that is know to have a 0 PH , 7.1 when I test it in an 7.0 solution standard, and in a 14.0 standard I am reading 13.8. Graph it on paper, draw a 45 degree line, the origin is 0 the other end is 14.0. Plot your reading above or below that line depending on wheather your insturment reads high or low. Draw a curve along the point and you now know how much to adjust your reading (mathematically) at each point along the scale between 0 and 14.0 PH. This does not mean that you PH meter reads more accurate, but now you can calculate a more accurate PH reading.
Now lets say you tested you PH meter a month later, and in the 7.0 standard solution it reads 7.2 and in the 14.0 standard solution its still reading 13.8. Go back to your original graph, average the first reading last month of 7.1 with the reading this month of 7.2 and since the 13.8 did not change of course you make no adjustments to the end point, redraw your graph. You now have an instrument that may very well be more accurate than the manufacturers stated accuracy. Test your instrument each month for a year and your instrument becomes more accurate with each passing month.
I must add a note here, if you test your instrument in a 7.0 standard solution, and get a reading of 7.1, 10 minutes you try it again and the reading is 6.9 and 10 minutes later you are reading 7.2, then your witnessing the performance capability of the instrument. If the fluctuation is beyond what you can tolerate, for example, my PH reads 8.2 right now, but in 10 minutes is reads 7.9 then throw the instrument away, it was never worth the money in the first place. Buy just because it reads OFF does not mean its a bad instrument. Consistant readings is what you are looking for, and a history of those OFF readings makes your insturment more accurate.
Also remember that manufacturers will use statistics to rate their insturment, they test 1 in every 10 or 100 sometimes and state the accuracy based on that information. You could by 2 of the exact same insturments and find that 1 of them perfoms consistantly while the other fluctuates. If it is consistant than having of history of its accuracy or inaccuracy makes it more accurate.
With all that said, understand that the insturments that typically tell use what a LB. (pound of weight) really is, are instruments that have a long history of use and have many mathematical adjustments to get the end result. I have been out of the pressure standards industry for about 10 years and at that point the most accurate primary pressure instrument had an unknown (you call it accuracy) of about 3 ppm, that number has surely become smaller simply because of the history of the instrument in use. A scienetist named Claude Mix (a pioneer of the dead weight gauges) told me once that NIST has a dead weight gage that has been in use since the 1940's. Great article, by the way. Thanks

[ConsultantERP]

I must add that adjusting your digital instrument each time takes away your ability to obtain the history that makes the above method usefull and tells you nothing about wheather the manufacturers claims of accuracy are valid.

[rick s]

ConsultantERP

Thanks for your comments.

The scenario you described is one of the reasons why I wrote the article. If someone is looking into buying a pH meter it must be because they know where their pH should be, and they want to find out where their pH actually is. If they would do their homework and purchase one that has:
a resolution of .01,
an accuracy of +/- .01,
quality calibration solutions,
automatic temperature compensation (preferred),
and the device does take a good calibration,
then they won’t ever have to go through the steps you suggest and they won’t ever have to wonder how accurate their meter is. Any fluctuations and deviations of a meter with the specs listed above would be relatively insignificant because an advertised accuracy of +/- .01 is certainly “tight� enough for the parameters we are striving to achieve.

A pH meter with an accuracy of +/- .1, that won’t calibrate to 7.0, would already have me doubting its abilities. This meter simply wouldn’t assure me the level of confidence that I’m looking for. But, that’s just me and my opinion.

My goal with the article was to help people understand what to look for before they purchase a device so they will know what they’re getting and be satisfied with it from the beginning. In this case, the cost of a +/- .01 pH meter compared to a +/- .1 meter is not that significant when you consider what it will do for you.

[gochinj]

So who are the players in the test meter industry? What brands should I be looking at when making my comparision?

[Scuba Oz]

I liked that article a lot. It does open the eyes to make the research more indepth.

So what brands do you recommend, and what brands were you refering too in the article?

[rick s]

Scuba Oz,

Thanks for the kind words. I really don't want to name names but, with the specs I quoted and a little research, you could find them.

I really can't make any recommendations as I have not used enough different instruments to fairly compare brands. Aside from the accuracy specs issue, I was amazed at the differences there can be in these meters. In the case of a pH meter: I wanted one that not only claimed accuracy, but the display had to be constantly on, bright enough to see from across the room, and no reports of electrical interference from other electrical devices. Out of all the pH meters available, only one fit the bill for what I wanted. Some do not have the option to display the value constantly. All the LCD screens were not bright enough for how I wanted to use it. Interference is reported as a problem with one brand. A couple days of thinking about what I wanted, research, and phone calls, and I found one that had everything I wanted.

As you pointed out: I'm suggesting people do a little homework instead of rushing into buying something just because someone else says, "I like mine."

[cnelson]

Yay Rick! Great article! Congrats!