Recalibrate IC40?

[JamesW]

Still fighting the issue of my IC40 cell calculating a low salt reading of 2650 and the k-1776 giving a result of 4K. As stated above I recently replaced the flow switch and it did rectify a no/zero salt reading per my system however the reading has yet to be accurate according to my k-1776 kit.

Even though the K-1766 is a very good test kit, it can still be incorrect.

You should measure the salinity two different reliable ways to be really sure before you decide that the cell is the wrong reading and not the K-1766.

 

[Roberd07]

Followed what’s in the video and disconnected the temp portion of the flow switch and I’m back to a green cell light and not a red low salt indicator. Given that doing this sets a default temp of 77 im guessing I do need to add an external temp probe to take the place of the flow switches thermistor?

 

[JamesW]

Given that doing this sets a default temp of 77 im guessing I do need to add an external temp probe to take the place of the flow switches thermistor?

Probably a good idea.

What is the actual water temperature?

 

[Roberd07]

85 degrees as of this afternoon.

 

[JamesW]

It's probably reading the temperature too high.

I would go with a separate temperature sensor.

I would also check the salinity two new different reliable ways.

 

[Roberd07]

I’ll get the salt tested elsewhere tomorrow.

 

[JoyfulNoise]

Unless you are performing the test incorrectly (operator error) or you compromised the reagents in some way (left them outside), I disagree with the test being inaccurate. The reagents of the K-1766 are incredibly shelf-stable and there are literally no interferences to the test (unless you’re routinely adding cyanide to your pool). So getting the salt measured somewhere else is only going to lead to confusion because I doubt the results will be anymore accurate than what you are doing and there’s a higher likelihood that they will be way off from your testing.

It sounds like the thermistor in the new flow switch is bad. This is not unusual as those combo flow/temp sensors are very poorly made. Pentair should really go with a different solution but they haven’t really taken the problem seriously. Adding an external temp sensor as described above is a much more reliable fix.

 

[JamesW]

Unless you are performing the test incorrectly (operator error) or you compromised the reagents in some way (left them outside), I disagree with the test being inaccurate.

I have seen multiple examples of inaccurate tests and I have found that the test sometimes gives inaccurate readings.

In any case, I never rely on a single test, especially when it is in conflict with a different test.

I would try a calibrated meter and a different K-1766 test.

 

[JamesW]

So evidence seems to be pointing towards something wonky going on with my Taylor K1766 results.

This is where things appeared to go awry for me. The K1766 results compared to the instant salt reading on the older T15 cell indicated the cell was on it’s way out (below the 75% threshold). It wasn’t until I installed a new T15 cell and it read the same as the old cell did it become apparent that something wonky is going on with my K1766 results (new reagents didn’t help) or the SWCG. When I can, I still plan to get a salt meter and calibration fluid to investigate further. For now, however, the new T15 cell is happily generating FC, so I’m in no hurry.

So perhaps my cell from 2015 is fine after all. I thought it was dying based on the K-1766 results I was getting.

 

[JoyfulNoise]

The R-0630 indicator is a potassium chromate solution. It is a shelf-stable solution and not subject to spoiling. The R-0718 titrant is a silver nitrate solution. As long as it remains properly sealed and not left outside in the heat or cold, it will remain stable for a long time. The chemistry of the Mohr method is such that the only positive interference with silver nitrate reacting with chloride would be in the presences of additional nucleophilic anion. Br- and I- will react with Ag(I) as they are strongly nucleophilic. -CN will react as well. None of those anions are usually present in pool water.

Did the reagents go bad … maybe. Perhaps the titrant was compromised. Mixing up a simple solution of table salt (non-iodized) in DI water would make an easy sanity-check.

As for a conductivity meter, that’s going to be inaccurate even with a standard chloride solution because pool water contains calcium and magnesium ions that will increase the total EC of the solution. Without very specific calibration methods to account for those interferences, it’s not going to be any more accurate.

I’d say buying a new K-1766 test kit is probably cheaper and easier than fooling around with an EC probe. But if having a second method is important, EC is probably a little better than test strips.

 

[JamesW]

My main point is that you should never assume that any test is absolutely correct 100% of the time.

Tests can be wrong for a variety of reasons.

If there is any doubt at all, double or triple check.

I have 2 K-1766 test kits and 2 meters.

If the reading makes sense and it matches the box, no need to verify.

Any doubt and I do all 4 tests.

Sometimes, I get a weird result with no obvious reason from a test that was previously reliable.

 

[Saturn94]

The R-0630 indicator is a potassium chromate solution. It is a shelf-stable solution and not subject to spoiling. The R-0718 titrant is a silver nitrate solution. As long as it remains properly sealed and not left outside in the heat or cold, it will remain stable for a long time. The chemistry of the Mohr method is such that the only positive interference with silver nitrate reacting with chloride would be in the presences of additional nucleophilic anion. Br- and I- will react with Ag(I) as they are strongly nucleophilic. -CN will react as well. None of those anions are usually present in pool water.

Did the reagents go bad … maybe. Perhaps the titrant was compromised. Mixing up a simple solution of table salt (non-iodized) in DI water would make an easy sanity-check.

As for a conductivity meter, that’s going to be inaccurate even with a standard chloride solution because pool water contains calcium and magnesium ions that will increase the total EC of the solution. Without very specific calibration methods to account for those interferences, it’s not going to be any more accurate.

I’d say buying a new K-1766 test kit is probably cheaper and easier than fooling around with an EC probe. But if having a second method is important, EC is probably a little better than test strips.

I don’t know if you read the thread where James pulled my posts, but I seem to be having issues with the K1766 kit.

Using and trusting the K1766 lead me to buy a replacement cell which as it turned out I didn’t need (had the cell tested; passed no problem).

At this point I don’t have much confidence in any salt test method (I still plan to try a calibrated meter as suggested in my thread just to have another data point).

If anyone who reads my thread has any info that can explain what is going on, I love to hear it.

Meanwhile, since adding salt as indicated by the SWCG (and two different T15 cells, one 6 years old and the other brand new), the SWCG has been happily maintaining FC with no issues (set at 57%, 10hrs pump run time).

 

[JoyfulNoise]

Sorry you’ve had issues. But I’ve used that kit for years and it’s spot on accurate. I’ve even measured homemade standard solutions and it comes well within testing error limit. The chemistry behind the test is very simple so there’s not much that can go wrong. Perhaps there’s something in your water interfering but it’s hard to say. I’m not a huge fan of Hayward cell salinity measuring methodology as it uses the cell itself to act as conductivity probe. That can be fraught with difficulty as it assumes all cells are the same and infers EC based on cell electrical parameters. Not a good assumption to make. That said, none of the manufacturers do a very good job on salinity measurement so you have to take what you get.

Meters are not foolproof. Even if they are calibrated against a standard, it still isn’t going to give accurate results because pool water has lots of dissolved ions in it. That is going to produce an offset in the measurement that you can’t account for easily. So, in the end, it’s no better than using the SWG because you really have no way of knowing how the EC value compares to the true chloride content of the water. The K-1766 measures chloride content exactly.

Hope you get it sorted out. And, on the bright side, you have a spare cell. Or, maybe you can come up with a creative way to run both cells in parallel and then halve your pump run time.

 

[JamesW]

But I’ve used that kit for years and it’s spot on accurate.

How do you know that it is accurate?

That's my point.

You didn't just blindly trust the reading.

You used a different way to figure out what the salinity was to verify the reading.

You have to have a different way to make the same measurement to know that the first measurement is correct.

So, you are doing error correction as standard procedure.

 

[ParB]

Couldn't you just take some distilled water put into a measuring cup, add salt and then verify the salt test against that?

I'm sure someone can calculate out the volume of salt to water for the rest to be in the let's say 3500 range?

That would let you get confidence in the test if there is doubt of its correctness?

I test hygrometers (relative humidity) using salt and a bit of water in a closed airtight bag to verify their correctness. Those meters are known for being crappy. I seek +-15% accuracy in those meters, I've had them being 50% off target (the meters I use can be calibrated,). It's a good thing to be able to simply verify a tests accuracy.

 

[BowserB]

This may sound too simple, but I'm a simple person. The IC40 manual says salt concentration should be between 3600 and 4500 ppm. Unless I missed it, I haven't seen any report in this thread that your salt tested to at least 3600 with the K-1766. (Yes, it is weird that Pentair gives a range and then specifies "ideal" to be the bottom of that range.) Still, the cheapest thing I can think of to try here, is to dump a bag--or two--of salt into the water and retest after it has dissolved. Maybe take a bottle of water to Leslies and let them test it too. And also, being a retired I.T. consultant, can you shut everything down--even shut off the breakers--and give it a couple minutes and then turn everything back on. I've long ago lost track of how many computer problems I've fixed that way.

 

[JoyfulNoise]

How do you know that it is accurate?

That's my point.

You didn't just blindly trust the reading.

You used a different way to figure out what the salinity was to verify the reading.

You have to have a different way to make the same measurement to know that the first measurement is correct.

So, you are doing error correction as standard procedure.

The proper way to verify reagents in any titration test is to first standardize your reagents against a known laboratory standard. You mix up the standard solution, you test your reagents against that and then you either decide if your reagents are too far off (maybe they sat on the shelf too long) or you calculate a correction factor to them (maybe fresh they were 100ppm/drop but now they are 95ppm/drop). You NEVER standardize one test with unknowns against another test with unknowns, that’s not correct lab practice.

In this case, it would be fairly easy to check the reagents of the K-1766. If you have a precise enough scale, you can make a standardized salt solution out of DI water and non-iodized table salt. If that’s too tricky to do, you can buy a standard salinity solution online and test the kit against that. There are plenty of standard salinity solutions you can buy very cheaply. If the reagents then test against a known standard and are within an acceptable error margin, then the issue isn’t with the K-1766.

 

[JamesW]

Again, my point is that you should verify the accuracy of any test and not take any test result as absolutely correct without question, especially when there is evidence that at least one of the test results is incorrect.

 

[Jimrahbe]

The IC40 manual says salt concentration should be between 3600 and 4500 ppm.

Bill,

The IC40 works just fine all the way down to 2800 ppm.. I usually run around 3200... 3600 may be "ideal" but it is not a requirement for the cell to work.

Thanks,

Jim R.

 

[ParB]

Again, my point is that you should verify the accuracy of any test and not take any test result as absolutely correct without question, especially when there is evidence that at least one of the test results is incorrect.

Totally agree!

If two sensors (or tests) disagree with each other and neither is validated against a known standard then you're in the worst of situations -you really can't trust either sensor (or test). If any decision is based from either sensor/test then you're doing the moral equivalency of flipping a coin.