Accuracy of KH director

[chelmon]

Hi,

I own a KH director for around 6-7 weeks. The start was quite difficult till the GHL apps accepted me in the GHL connect, rejections for a couple of days days till suddently it worked.

After several days getting weird values the problem turned out to be that I got a lemon probe since the ADC calibration values were out of expected range. I got a new one from GHL Germany. Then for about 4 weeks I have been getting values with some error respect to the Hanna or Salifert values, around 0,3 KH units. Not really happy since Salifert and Hanna match quite close together, always the KH director looks the wrong one. In the last days the bias started to grow and reached around 1 KH unit.

Yesterday I recalibrated the pH sonde and it was quite the same ADC values as the las time. For the dosing pumps recalibration the one that is filling the probe changed from 32 in the previous calibration one month ago to 36 ml/minute. It looks a better value since the nominal of the pump is 40 ml/min. After correction I started to get values that are above 1 KH offset, see the table below.

Now I wonder if in addition to the lemon pH probe I also got a lemon dosing pump. I do not understand why I cannot open the unit and see how much it is being refilled, or why the prove is not visible, it would help in the debugging.

Below the comparison table of various days whe I took measurements with Hanna and/or Salifert and the display in the app showing very noisy values. I am very sorry to say that this product is quite disappointing for me.

It is a 2 months old lps tank with Tropic Marin all for reef one component dosing.

Hanna	KHD
7,7	8.00
7,3	7,1
7,1	6,9
7,3	7.00
7,7	7,5
7,7	7,4
7,9	7,6
8.10	7,8
8.00	8,2
8.00	8,5
8.23	7,9
8,23	7,8
7,56	7,4
7,33	7,1
7,72	7.00
7,89	7,3
7.80	7.00
7,9	7,1
8,17	7,1
8,17	7,3
8,17	7,7
8,4	7,4
8,3	7,4
8,23	7,6
8,57	7,4
8,68	7,4

 

[wmb0003]

How did you calibrate the dosing heads? Graduated cylinder or scale? Also the sample tube volume? Did you use their calculator or actually empty the tube and measure it? I would suggest calibrating everything with an accurate scale.

Also it appears something happened at point 15, you had an increase in alk according to hanna but not KHD and the difference continued for the rest of the graph.

Also for the first half of the graph it appears you are running consistently at .02-0.4 less on the KHD than hanna. I would say this is normal and would be happy with those results. but then for some reason the results flip flopped and your KHD was higher than hanna.

I would not worry about trying to get the hanna and KHD to match, if they are within 1dkh of each other and are on average about the same difference each time, that is good in my opinion. Does it really matter if you are 7.7 or 8.1 dkh? Not really, but it does matter if you drop 1.5 dkh in a day. And the KHD can easily alert you if that happens.

 

[chelmon]

Thanks for the anwer. I am calibrating them with a 20 ml syringe by collecting in 2 steps the liquid poured in a probe. I also measured the liquid from the tubes and used their calculator. That value has not changed and I think it is OK even with the filter.

the values are not consecutive, sometimes they have gaps of a few days, and I calibrated the pH probe a couple of times, but with very similar results. As you say now the bias has started to increase. However when I measure with hanna and Salifert they are always really very close within 0,2 KH.

On day 15 I increased the All For Reef dosing from 10ml to 20ml per day. The KH has started to increase (according to Tropic Marin, AFR systems tend to KH 9)

The KH has started to increase in the last days but the KHD is not detecting it, well really 3 days ago I got a KH of 8 with the KHD and on that day I recalibrated the dosing pumps resulting in a drop from 8 to 7,4KH and I registered the 7,4. I wonder if that dosing pump is inaccurate and is the one causing the noisy results. GHL could solve it just by allowing to see how much is filled in the probe. That would help really to the users and would remove one uncertainty.

Anyway I am not really very satisfied with this device. Dosing the alkalinity based on the noise measurements does not look very feasible.

 

[wmb0003]

Testing the dosing pump should be easy enough, continue to run calibrations and see what amount you are getting each time. They shouldn't vary much, if they do then you have a problem. Alternatively you can tell your doser to dose X amount and then compare. You can also get the KHD reference fluid to compare.

The way the KHD measures alkalinity is very accurate. It is the way Randy Holmes Farley suggests measuring alkalinity, per this article https://www.reef2reef.com/ams/a-diy-alkalinity-test.12/ The KHD just automates this process using dosing pumps and a ph probe.

I would still not let the KHD dose based on the readings, I would manually adjust the dosing amounts as needed.

 

[chelmon]

It is quite disappointing that such an expensive device is not at least as accurate as the manual tests, needs so frequent calibration and cannot be trusted for dosing the alkalinity liquids as advertised. May be there is not a big difference between alkalinity 7 or 8, but if you try to maintain alkalinity of the NSW and get a 1 KH error it looks too much.

Google is pointing to this video every time I was looking for information before buying it. I thought it was exaggerating

I am not the one who resells things, I will probably open it to see how much it is refilling every time and will refuse to the warranty, that is one of the three things that depend on the calibration and I suspect a problem in the dosing pump as the video was also pointing. Are there around pictures or videos of the device dissassembled? The pH probe calibration needs to be done from time to time (I have already suffered a lemon probe in the brand new product, GHL send me a new one but not even more calibration fluid to compensate all the one I wasted trying to figure out the problem). BTW, why is it a different pH probe? Is it due to the diameter to make it fit or is it optimised for low pH? The calibration of the reagent dosing pump is more tricky, it could make drop dosing to improve it, and tell how much it has dosed to help the user to understand problems, this is pointed out in the video.

I am sorry to say this, hopefully I will manage to change my opinion on the device, but so far I am having many problems with it.

 

[LOVEROCK]

What are the number for adc1 and adc 2 values ?

 

[wmb0003]

It is quite disappointing that such an expensive device is not at least as accurate as the manual tests, needs so frequent calibration and cannot be trusted for dosing the alkalinity liquids as advertised. May be there is not a big difference between alkalinity 7 or 8, but if you try to maintain alkalinity of the NSW and get a 1 KH error it looks too much.

Google is pointing to this video every time I was looking for information before buying it. I thought it was exaggerating

I am not the one who resells things, I will probably open it to see how much it is refilling every time and will refuse to the warranty, that is one of the three things that depend on the calibration and I suspect a problem in the dosing pump as the video was also pointing. Are there around pictures or videos of the device dissassembled? The pH probe calibration needs to be done from time to time (I have already suffered a lemon probe in the brand new product, GHL send me a new one but not even more calibration fluid to compensate all the one I wasted trying to figure out the problem). BTW, why is it a different pH probe? Is it due to the diameter to make it fit or is it optimised for low pH? The calibration of the reagent dosing pump is more tricky, it could make drop dosing to improve it, and tell how much it has dosed to help the user to understand problems, this is pointed out in the video.

I am sorry to say this, hopefully I will manage to change my opinion on the device, but so far I am having many problems with it.

To clarify about auto dosing, it’s not that I don’t think it’s accurate enough, there are other issues, mainly user errors that could lead to a tank crash. Ex. Fill line partially out of water, air bubbles in dosing line, reagent evaporation, etc. according to my KHD over the past week there has only been a .4 swing in dkh. I would say that’s very consistent. If it starts trending lower. I will up my dose by .5-1ml per dose.

I have also attached a picture of my khd from when I first bought it and had issues. The area marked in yellow, you need to ensure the tube is all the way in the holder, it needs to be pushed in to the left. I bought it used and it was a pain because the ph probe was left to dry out and it was bad. Then I got it replaced and the mixing internal pump wasn’t working and it ended up needing another board. GHL customer service was awesome at getting my issues resolved.

I am not 100% sure about the ph probe but I believe it is designed to be used in a horizontal position rather than vertical, which is used for most in tank ph readings.

I will see if I can’t find another pic of the inside of the KHD where everything is right.

also I added a picture of my adc values for comparative purposes, I’m not sure how close they should be from user to user though.

 

[chelmon]

Thanks for the picture. When my pH electrode was replaced the common ADC values are assumed around 48500 and 28000. I had the following calibration readings: 56370 and 35338, 57132 and 35036, 55088 and 34780 and the probe was replaced.

Even if the device is open, It does not look easy to check how much fkuid has been dosed inside the probe with that shape, and I assume it is not possible to rotate 90deg the device to a vertical orientation, the inputs and outputs need to be on one lateral side, not from the top