AccuBlue Home test tolerances

[lymanepp]

I received this from Leslie's technical support. These are terrible and explain the wildly erratic results from AccuBlue Home and their in-store tests.

 

[Bperry]

I received this from Leslie's technical support. These are terrible and explain the wildly erratic results from AccuBlue Home and their in-store tests.

View attachment 538151

Welcome! They are way less accurate than even that spec says. Those tolerances are best case and only when it’s been calibrated, cleaned and someone trained is using it. A primary reason why TFP exists is to convince users to test their own water because the pool store tests aren’t reliable.

 

[JamesW]

Their actual results are probably about as accurate as this random number generator.

RANDOM.ORG - True Random Number Service

RANDOM.ORG offers true random numbers to anyone on the Internet. The randomness comes from atmospheric noise, which for many purposes is better than the pseudo-random number algorithms typically used in computer programs.

www.random.org

 

[Donldson]

Even worse than that massive margin of error is they still program it to show incredibly precise numbers. If you are testing a sample of 10 ppm FC it will register something between 7.5 and 12.5 but still show something like 8.1 to make one think it is way more precise than it is. We see people fall for that imaginary precision almost every day during the swim season, to the point of argument sometimes.

 

[Newdude]

FC it will register something between 7.5 and 12.5 but still show something like 8.1

FC is to the hundreths place, like .06 FC ever saved anyone.

'Don't worry, if it was a 3 you'd be in trouble, but it's 3.06'

 

[Rocket J Squirrel]

For comparison, what is the tolerance of the Taylor reagents, assuming perfect drops & technique?

 

[Newdude]

+/- 1 drop, or +/-10 for the CYA.

Using salt as an example, at 3500 Taylor may have a 200ppm variance and Leslie's may have a 875 variance. (Per the above).

I guess that proves the iron test is useless too with a .3 variance at .3.

 

[JoyfulNoise]

+/- 1 drop, or +/-10 for the CYA.

+/-15ppm for CYA ... ackshullalee

 

[JJ_Tex]

Just because the dumb kid in math class gets a fancy new calculator doesn't mean you should trust his answers.

* Stolen from Newdude, but seemed appropriate here.

 

[JoyfulNoise]

Just because the dumb kid in math class gets a fancy new calculator doesn't mean you should trust his answers.

* Stolen from Newdude, but seemed appropriate here.

I knew all the good looking girls sat around me in math class for a reason ... and I thought it was because I was rocking my 80's hair George HW Bush glasses ...

 

[Rocket J Squirrel]

+/-15ppm for CYA ... ackshullalee

That's a little scary. My CYA has consistently read 30 for a while, but it might ackshullalee be as low as 15? Yikes.

 

[Newdude]

My CYA has consistently read 30 for a while, but it might ackshullalee be as low as 15?

-15 rounds up to -10, so it's a 20.

 

[JoyfulNoise]

Let's just say that the CYA test results should be taken with a grain of salt ... but not too many grains or else you'll mess up the K-1766 test results ...

 

[JamesW]

For any test, you need to know the accuracy, precision and resolution of the test.

For example, a salinity test might show 3,200 ppm +/- 250 ppm with 99.9% certainty assuming the test is done exactly right and the test equipment and materials are in good condition.

So, this gives you a range of about 2,950 ppm to about 3,450 ppm, which is a range of about 500 ppm.

Note: There is still a 0.1% chance that the number is not within the identified range.

If you have more than one test, you can see if the test ranges all overlap.

If all ranges overlap, then you can estimate that the actual number is likely to be in the range where all tests overlap.

In the illustration below, there are 4 tests with 4 different ranges and they all overlap the area between the two black lines

If all ranges do not overlap, then you can estimate that one or more of the tests is likely to be inaccurate.

This is why it is usually better to have certain tests done using several different methods.

For example, a salinity test where you have zero prior knowledge of what the actual salinity should be should be done with a K-1766 salt test and a calibrated meter at a minimum.

If you can get other estimates of the salinity, then that is usually a good idea as well.