Would you invest a grand for a KH Guardian?

[iiluisii]

Maybe you take people for idiots. But most of us see the problem with spending an extra 200 for an upgrade. And no reagent pricing info. Give us a break, they know exactly what that is going to be. If they really don't know the pricing of the reagent then this is nothing but a scam. I'm sick and tired of paying good money to Beta test someone elses product.

I agree with you about the beta testing and coralvue is known for selling products exactly like you described. Beta products

I really don't see how people deal with coral vue. The comments by coralvue in this thread shows how they are. I really think this thread has nothing to do with asking for pricing which I already think they know they will charge over a grand for the unit.

 

[nervousmonkey]

If I offended you @Dumaurier7 , I apologize. Wasn't trying to offend, just sharing my opinion.

Back to topic here: If I could buy one now, I would. Prices will drop though, as always tends to happen with new tech. Only way it doesn't drop is because demand for the product is incredibly high, like all iPhones, iPads, etc., but demand for this product at this price point is not high enough that $1000 pricing stays around, and it looks like almost all the new kH monitoring devices coming onto the market will reflect the same price. Reference use of the Apex device on Terence's tank: https://www.reef2reef.com/threads/gilroy-425g-build-thread-all-apex-all-the-time.198797/page-53. In any case, these devices, not being specific re: the CoralVue product, are expensive, and I have no reason to believe that other manufacturers are going to be any less. Hanna's electronic version of this device, the Mini Titrator, costs $810, with no dosing pumps added. It's the testing that is expensive, not the dosing of alkalinity.

In reference to what people are saying here, and I think this is an awesome thread with a ton of insights into the way people think about their own methods of reefing:
"I do see a lot of people angry about the price but this type of testing is not easy to do and requires pretty hi tech equipment to do it properly." Completely agree with you @reefwiser

"Why not just release the product day one with the best software/hardware possible at a Fair market price?" What determines "fair market price"? In economic terms, it is always demand or what the consumer is willing to pay, uncoupled from cost. Same as the housing market. Fair market is what someone is willing to pay for a house. If the cost of building the house is greater than the price people are willing to pay for it, the house won't sell. I don't own an Abyzz pump for one reason: price.

In terms of accuracy, the KHG is close to all testing equipment, but we just don't know that as the testing kit manufacturers do not provide standard deviations (for the most part, I believe that Hanna does, which is a 0.25 dKH SD). Arguing that these devices are not as accurate as the manual method (a test kit gives a "definitive" number, which we all believe to have no deviation from perfect) is a bad argument to make, and not an argument that is scientifically based in any manner whatsoever. Stating that "Maybe it is ok with +-0.2, but then still a "Salifert" is better" is crazy, Salifert has a standard of measure of 7.2 dKH +/- 0.3. I would say that Salifert alk testing is no more accurate than anything else is. The margin of error on this device or any other similar device is no more or less accurate than salifert, hanna, api, etc.

The argument I do agree with however, was made pretty plain by @redfishbluefish Not making any adjustments to calcium is a problem, but with calcium monitoring modules coming out, the market will self-correct to reflect what reefers actually need. I agree with redfish that now is probably not the time to buy this for most people, as alk is not the only parameter we care about.

Getting angry about any of this makes no sense to me. I am sure that we can all wait a year or two for price, accuracy, manufacturing, additional modules, etc. to come within the range that we would consider worth the cost when compared to the deliverable that we desire.

@NeverlosT has made some great comments, see post #230. Makes a lot of sense. Once again, if someone is willing to pay this price for an alkalinity monitoring tool, then great. This is still free market economics at play. Even Adam Smith would agree I believe. This market will self correct, as do almost all free markets when it comes to competition, supply and demand, and self-interest/third party interests.

There are other great points being made that if this device, or kind of device, were built and sold the way that the consumer wanted it to be built, it would not be an issue to purchase it. I agree, and these devices are still new to the market, not new to development, but new to us as consumers. The companies that make an alkalinity monitoring device that best fits the needs of the consumer will always win, so in answer to these statements, I am in total agreement and hope that manufacturers pay attention to what people are posting here. It would be more worth our time to stop arguing about these new devices and come together to make proposals for change to the manufacturers.
I hope I am not offending anyone, but at this point, there are apparently a lot of people that became offended as soon as the price was announced. I don't own an Abyzz pump, while some here use them and wouldn't use anything else. Not everyone can afford this thing, and that's fine. We can't all afford an S-Class Benz 600, and we all can't afford an Abyzz pump.
My $0.02.

 

[Twocorns]

Nope. I can easily test with my Hanna and dose with my jebao.

 

[JonasRoman]

It is obvious you dislike KHG's 0.33dKH variance.

I just looked up on the Hanna checker webpage: http://hannainst.com/hi755-alkalinity-saltwater.html
It says on their page that the accuracy of the Hanna checker HI-755 is
Accuracy @ 25°C/77°F = ±5 ppm (mg/L) ±5% of reading
Since 1 dKH = 17.848 ppm alkalinity, that means the accuracy of the Hanna checker is +/- 0.28 dKH (+/- 5%)

Hanna checker accuracy variance is 0.28 dKH
KHG is 0.33 dKH
Difference of the two is 0.05 dKH. Can you even tell 0.05 dKH from typical test kits?
............and I can't even find accuracy statement on the Salifert web page

So what is your point?

(It would be great if alkalinity monitor/controller can have better accuracy then the Hanna checker, but I don't know if affordable technology can achieve it.)

The Hanna-checker is not forgiving bad accuracy of another brand. Hanna-checker is also very bad, and is constantly showing too high, because their are some approximations in their method. The Hanna-cheker uses a "third" princip, to simplify the process, skipping the titration procedure. They always add a constant amount of acid, so the destination pH always will be well below 4.00. Then the pH below this point is not CO2-sensitive as it is above (maybe a little), and also they assume that the relationsship between pH and dKH is linear in this pH-interval. Then it is possible according to only pH calculate dKH, BUT with more inaccuracy when dkH is high. I found when comparing to a regular titration (with pH-electrode, lab-pipette etc) that the Hanna was showing constantly 0.2 too high, sometimes more. So, the bad accuracy of KHG is not forgiven. In this debate you have to compare with a simple Salifert or similar, with very good accuracy as long as you do everything correct. I would say at least +- 0.1 or better. That is acceptable. For a machine which shall detect variations and do actions, you have to have at least this accuracy, o even better I think. But, be calm, everything is possible...we have just to wait for the correct moment when the market could offer this. Of course the time will come, sooner or later.

 

[nervousmonkey]

In this debate you have to compare with a simple Salifert or similar, with very good accuracy as long as you do everything correct.

Great point Jonas. To remain within the accuracy limits of these tests, one has to follow instructions perfectly. Adding too much reagent, saltwater, etc. will render the results too far from accurate to start making decisions about dosing given a one point measurement. As with all testing in reefing, the trends are more important that a single point by which we base our decisions. Even Salifert has a standard deviation of 7.2 +/- 0.3 dKH as an alk standard, so Salifert testing is prone to be off as well, although one batch of reagent is almost always off by the same amount, so dosing decisions can be made with the trendline of Salifert alkalinity testing kits, but as you state, *everything* must be done correctly.

 

[JonasRoman]

Great point Jonas. To remain within the accuracy limits of these tests, one has to follow instructions perfectly. Adding too much reagent, saltwater, etc. will render the results too far from accurate to start making decisions about dosing given a one point measurement. As with all testing in reefing, the trends are more important that a single point by which we base our decisions. Even Salifert has a standard deviation of 7.2 +/- 0.3 dKH as an alk standard, so Salifert testing is prone to be off as well, although one batch of reagent is almost always off by the same amount, so dosing decisions can be made with the trendline of Salifert alkalinity testing kits, but as you state, *everything* must be done correctly.

I agree, their are weakness in salifert also(the user mostly). But if you for instance do 10 measures following each other and do it exactly the same, the accuracy is much better than 0.33,i would say 0.1, so as you say, the trend you can see quite well with salifert, but not with the KHG. The point is, that you probably do the same error with salifert, thus have a reasonable accuracy (but maybe a little worse precision but that is not equal important), but with a machine with accuracy of 0.33 (and maybe even 0.1) the errors will vary, so it will be more spreeding of the inaccuracy.

 

[nervousmonkey]

Yep, I agree. Don't know how the technology will become more accurate, but that's what we need. Also agreed on precision. I don't need that, to me accuracy equals stability, which means I need accuracy of trends with an alk device. Had not thought of that. Great point, maybe I will stick to what I do now. It works.

[EDIT: Reverse the accuracy and precision terms, as pointed out by @JimWelsh ]

 

[JimWelsh]

I see many posts here where I suspect that the terms "accuracy" and "precision" are being confused / conflated / misused. "Accuracy" is the closeness of the result to the true value; in other words, whether the answer is "right" or "correct". "Precision" is the closeness of repeated measurements to each other; in other words, how consistent and repeatable the results are. One classic teaching tool for these concepts is the "target" model, where accuracy is shown as the bullet holes being centered around the bullseye, and precision is shown as tight grouping, as in this image:

 

[nervousmonkey]

Awesome! Thank you Jim. I stand corrected and in this case, well corrected. Precision then is more important, at least to me, in an alkalinity monitor. Accuracy being off is not that important. I know you can't speak to anything alk monitor related Jim, but in the case of a Salifert test, what I need for stability is precision of the measurement then, correct? At least to maintain stable levels of what I am testing for, calc, mag, etc. Accuracy in testing, at least on a one point measurement, is not as closely correlated to me being able to keep the same level of say, calcium, to keep this on neutral ground, correct?

 

[JimWelsh]

Awesome! Thank you Jim. I stand corrected and in this case, well corrected. Precision then is more important, at least to me, in an alkalinity monitor. Accuracy being off is not that important. I know you can't speak to anything alk monitor related Jim, but in the case of a Salifert test, what I need for stability is precision of the measurement then, correct? At least to maintain stable levels of what I am testing for, calc, mag, etc. Accuracy in testing, at least on a one point measurement, is not as closely correlated to me being able to keep the same level of say, calcium, to keep this on neutral ground, correct?

Correct.

 

[nervousmonkey]

Correct.

Thank you for the clarification and chemistry lesson! I'll remember now from the graph you showed. "A" on x axis, "P" on Y axis.

 

[JonasRoman]

I see many posts here where I suspect that the terms "accuracy" and "precision" are being confused / conflated / misused. "Accuracy" is the closeness of the result to the true value; in other words, whether the answer is "right" or "correct". "Precision" is the closeness of repeated measurements to each other; in other words, how consistent and repeatable the results are. One classic teaching tool for these concepts is the "target" model, where accuracy is shown as the bullet holes being centered around the bullseye, and precision is shown as tight grouping, as in this image:

Thanks I have to admit that when I wrote accuracy I ment precision and when I wrote precision I ment accuracy. Thanks for clarifying this:_) So, precision is important as a indication of that the machine do the same errors, but accuracy is little less important (but of course not unimportant).

 

[nervousmonkey]

To clarify more then on this topic Jim, and to keep this within bounds so you can respond... Theoretically, when I use my Salifert test to measure calcium, and I know that the SD of the test is 450 +/- 25 ppm, and I use my testing kit to try and keep my calcium levels as close to 475 as I can, what I care about is seeing results from my Salifert test that are consistently precise, correct? Then my issue with keeping my calcium at 475 is really keeping it at what I believe to be 475, although, it could be 450 to 500 in all reality, correct? And since what I am concerned with is keeping my calcium stable, as long as I am keeping it at the 475 value as best I can, knowing that it really has a range of accuracy is still OK, so I am not that concerned with the SD of the testing kit. Assuming a lot here, but assuming that a range of 450 to 500 ppm calcium would keep all of my corals healthy and growing, the standard deviation of +/- 25 ppm is really not that bad. Obviously in this theoretical testing experiment, I want a testing kit that has as low a SD as I can find, to maintain stable levels within an acceptable range. Is there another limit or measurement "error" of my testing kit that I should be thinking about?

 

[JimWelsh]

The Standard Deviation and precision are very closely related, and for the purposes of this conversation could more or less be considered synonymous.

Fundamentally, in analytical testing, there are two types of error: Random error, and Systematic error.

The SD / precision we have been discussing is in the category of Random error, where you analyze the same sample over and over again and get (hopefully) slightly different answers. This correlates with the "tightness of the grouping" in the image I posted above. This type of error is easy to determine statistically, e.g., repeat the test a number of times and determine the SD from the results. This type of error is also easy to address and minimize by increasing the number of analyses being performed, and taking the average of the large number of results, you are able to determine the metaphorical "center of the group" in reference to the target model.

Systematic error is a type of error that tends to consistently affect the accuracy of the result in a particular direction, introducing what is referred to as a "bias". For example, let's say that the syringe I use to take my sample of tank water when doing my Salifert calcium test has markings that are a little bit off, so that I always end up taking 1.9 mL of water instead of the 2.0 mL I'm supposed to take, then my results will consistently read lower than the "true" value, meaning the results were "biased low". This type of error is impossible to determine statistically. This is where calibrating the test against known standards and/or using alternative testing methods that are known to be more accurate need to be done in order to determine the Systematic error.

To answer your question, when I read "Accuracy" numbers cited by manufacturers, I tend to interpret them as "Precision" values, or the "measurement uncertainty".

A good discussion of these topics that explains them much better than I could, for those interested in further reading, can be found here: https://www.wmo.int/pages/prog/gcos/documents/gruanmanuals/UK_NPL/mgpg11.pdf

 

[nervousmonkey]

Fantastic answer Jim! Puts all of this into a much greater perspective. Great article on statistics as well. I have taken many courses on stats, but have never had the need to use lessons from those courses; over the course of time the knowledge just goes away if not used.
I really wish I could ask the question that I want to ask, which is, what is an acceptable "measurement uncertainty" value for testing alkalinity? Given the assumption that all testing was performed equally well by a person. As that equates with something you are most likely bound to keep under wraps, I understand not being able to answer, which just means I wait for the Jim Welsh/Apex alkalinity module to be released.
If you can answer that question in terms of Salifert testing, or Hanna checkers, that would be fantastic!
Thanks Jim, I really appreciate your knowledge!!!

 

[JimWelsh]

Fantastic answer Jim! Puts all of this into a much greater perspective. Great article on statistics as well. I have taken many courses on stats, but have never had the need to use lessons from those courses; over the course of time the knowledge just goes away if not used.
I really wish I could ask the question that I want to ask, which is, what is an acceptable "measurement uncertainty" value for testing alkalinity? Given the assumption that all testing was performed equally well by a person. As that equates with something you are most likely bound to keep under wraps, I understand not being able to answer, which just means I wait for the Jim Welsh/Apex alkalinity module to be released.
If you can answer that question in terms of Salifert testing, or Hanna checkers, that would be fantastic!
Thanks Jim, I really appreciate your knowledge!!!

Well, I don't know about what an "acceptable" measurement uncertainty for an alkalinity test method is, but I can tell you what I think might be considered a quite reasonably precise measurement uncertainty.

Long before I started work on my Alkalinity Monitor, I had developed and used my own DIY alkalinity test method for several months. It is a variation on the Hach burette method, and uses 10 mL of tank water measured with a volumetric pipette for each test. Sometimes I used a pH meter for endpoint detection, and other times I used an indicator dye. The reagent strength was chosen to require appx. 15 mL of reagent per test, so I could fill a 50 mL burette, and then do the tests in triplicate, having enough reagent for three tests in one burette filling. Striving to be as consistent as possible, and using laboratory best practices for working with volumetric glassware, I was able to consistently achieve a Relative Standard Deviation of 0.3%, or about 0.025 dKH for an alkalinity level of around 8.3 dKH.

It should be noted that this RSD correlates very strongly with the relative size of a single drop from the burette vs. the total amount of titrant used for each test: I have calculated the average drop size from the 50 mL burette to be right around 0.058 mL, and a typical test would consume appx. 14 mL of titrant, so the uncertainty due to +/- that last drop would be 0.058/14.0 = about 0.4%.

When I estimate method uncertainty in my work, I use a "coverage factor" of k=2 for a confidence interval of appx. 95%, so, in other words, I use 2*SD, so I consider the measurement uncertainty of my DIY alkalinity burette method to be 0.6% of the measured value, or about 0.05 dKH for alk = 8.3 dKH.

 

[nervousmonkey]

Well, I don't know about what an "acceptable" measurement uncertainty for an alkalinity test method is, but I can tell you what I think might be considered a quite reasonably precise measurement uncertainty.

Long before I started work on my Alkalinity Monitor, I had developed and used my own DIY alkalinity test method for several months. It is a variation on the Hach burette method, and uses 10 mL of tank water measured with a volumetric pipette for each test. Sometimes I used a pH meter for endpoint detection, and other times I used an indicator dye. The reagent strength was chosen to require appx. 15 mL of reagent per test, so I could fill a 50 mL burette, and then do the tests in triplicate, having enough reagent for three tests in one burette filling. Striving to be as consistent as possible, and using laboratory best practices for working with volumetric glassware, I was able to consistently achieve a Relative Standard Deviation of 0.3%, or about 0.025 dKH for an alkalinity level of around 8.3 dKH.

It should be noted that this RSD correlates very strongly with the relative size of a single drop from the burette vs. the total amount of titrant used for each test: I have calculated the average drop size from the 50 mL burette to be right around 0.058 mL, and a typical test would consume appx. 14 mL of titrant, so the uncertainty due to +/- that last drop would be 0.058/14.0 = about 0.4%.

When I estimate method uncertainty in my work, I use a "coverage factor" of k=2 for a confidence interval of appx. 95%, so, in other words, I use 2*SD, so I consider the measurement uncertainty of my DIY alkalinity burette method to be 0.6% of the measured value, or about 0.05 dKH for alk = 8.3 dKH.

Now that is a precise answer, no relation intended re: the "accuracy vs. precision" question! Thank you Jim, that is an amazing method of testing. Had to look up the Hach method, but makes a ton of sense when applied here. Incredible RSD to achieve, essentially what you are saying is that the precision was determined more by an accidental drop of water than by the method or reagents used. That is quite enlightening.

Did you use that method all the way up to the development of the Alk Monitor? Just wondering as it interests me a lot, would love to learn how to do that. I understand the glassware and how to use lab equipment; took 3 years of chemistry in undergrad at Ga. Tech. Additionally, is this the way your brain thinks when thinking about alkalinity measurements? I understand you from the CI perspective, as that is most likely the k value you typically use in testing, since that's what most commercial, regulatory, etc. applications would use. I would think that in this application, setting a CI with k=3 would be unnecessary, as the 95% CI gave you the uncertainty measurement of 0.05 dKH for 8.3 dKH. Seriously impressive Jim. That beats the current application methods for alkalinity by a long way. I'm sure that my testing with Salifert or Red Sea can't provide that level of uncertainty.

This answered all the questions I have about the topic! Thanks Jim, sharing your knowledge with our community is amazing.

 

[eg8r210]

I see many posts here where I suspect that the terms "accuracy" and "precision" are being confused / conflated / misused. "Accuracy" is the closeness of the result to the true value; in other words, whether the answer is "right" or "correct". "Precision" is the closeness of repeated measurements to each other; in other words, how consistent and repeatable the results are. One classic teaching tool for these concepts is the "target" model, where accuracy is shown as the bullet holes being centered around the bullseye, and precision is shown as tight grouping, as in this image:

Pictures make everything easier. Thanks.

 

[iiluisii]

$1269 for the standard version and $1300 for the pro version seriously?? All I can say is money hungry. Good luck.

 

[sk8elenex92]

$1300 !!!!!!!!!! thats crazy