The reason for using blotting paper is to get around the extreme difficulty of getting the solution in the unit. As this shrinks in size, as the amount of fluid in it decreases to improve the time response, the difficulty in getting the fluid in it, goes up exponentially. If the fluid were a gel it would be much easier (I think), but I have no idea how to work with that.

I agree that this is moving past the simple DIY concept, where most people will opt out of trying to deal with the solutions, the tiny pieces of plastic, the "invisible" membranes, etc. So, getting Seachem or someone similar to produce something along these lines would be by far the best way to go.

To change the subject, kinda: you recommended making 4 dKH water by starting with 5 liters of DI or distilled water, etc. That is a lot of water to hold in a clean container, only to dump most of it. Why not start with 1 liter of distilled/DI water with 4.99 g of sodium bicarbonate? That gives 200 dKH water. Add 50 ml of that to 50 ml of pure distilled water to get 100 dKH. Add 50 ml of that to 50 ml of pure distilled water to get 50 dKH. Keep that in a 100 ml container. Then mix 10 ml of that with 90 ml of distilled water for 5 dKH or mix 8 ml of that with 92 ml of distilled water to get 4 dKH. That is more measuring and mixing steps, but it requires only a 1 liter graduated flask and a graduated cylinder to do the measuring. Do all of those measuring steps degrade the accuracy too much? I suspect it does.

Vaughn,

That 1 liter version will work fine, I used a 4 liter as an example.
When you increase the volume and use a concentrated solution, the steps from there do not increase the errors.

Example:

If you are off by 0.05 grams in the 200KH solution, this means it's only 0.2 ppm off or so(I'm just guessing) at 200KH.

When you go from high concentration down to dilute, this error becomes much less. You also dilute the error in a sense.

Take a solution comparison: 201 KH vs 200 KH.
Error range is 1 KH or %0.5
Take a 20 KH vs 21 KH solution, now the error is 10X what it was at the higher concentration. 5%
Take 2 vs 3 KH, now the error huge

So starting at 200KH and diluting down will yeild pretty low errors, as each order of magnitude will dilute the errors you made at the large volume and the high concentration.

This is how we can make serial dilutions and do colony counts in microbiology. Counting billions of bacteria/colonies is not something anyone would/could do.
We just scale back up to get the bacterial forming units. 10-^12 dilutions etc.

Same type of thing here.
Start with 1 liter volumetric flask, a decent gram scale to 10mg's, make 200KH etc and then dilute to whatever level you want.

Being off 1 degree KH at 200 KH is no big deal, being off 1 degree KH at 3-4 range is.

Now factor in pH/KH and the range of CO2 there.
One KH unit off at 4-5Kh @ pH 6.6 will = 7ppm CO2
Now take a solution of 200-201KH assuming your volumes for the dilutions are accurate: KH4 dilution will have 0.02% error, or about 0.36ppm KH.
I'd say about 0.1 to 0.15ppm of CO2 error over that same range.

So about 7ppm vs 0.1ppm CO2 error.
This assumes a lot, like the baking soda is all baking soda, that the volume of water is accurate in 1 liter, and the dilution volumes are accurate etc.
But even here, you have much less requirement to have accurate 1 liter volumes.

If that is off say about 2 %, not that big of a deal. 2% of .15 ppm of CO2 is not that far off. Nor 5% etc.

It's when folks make 100mls or less, are sloppy etc, use inaccurate scales, try and make dilute solutions etc........ then they get large errors.

Regards,
Tom Barr

Update time!
Paper doesn't work to make a drop checker. It doesn't absorb enough of the fluid to change color enough to see the color change. I used some art paper, acid free, supposedly highly absorbent, and pretty thick. It bombed!

So, I thought some more and decided "SPONGE!!". I rushed out an got a pack of the little white sponge wedges women use to apply or remove make up. I cut a thin slice of one, fitted it into a membrane drop checker between a piece of Tyvek and a piece of Cole Parmer membrane. Then I soaked it liberally with 5 dKH water, and added a couple of drops of pH reagent. Then I sealed it to the end of a small acrylic tube, about 1/2" in diameter. Placed it in the tank, and it worked! It changed to green in about 5-10 minutes, is still tracking the color of the ADA style drop checker, and it was pretty easy to make. With some more work this could be a good way to cheaply make a fast reacting DIY drop checker.

Just out of curiosity I soaked a piece of the sponge in some distilled water for an hour, then checked the pH and KH of the water. Zero KH and 7 pH.

Next result: Tyvek Priority mail envelop material works fine as a membrane. It is thinner than building wrap Tyvek, so it is easier to use, and the cost is right where we DIYers like it to be. Of course it isn't transparent, so either it has to be a single sided membrane device or has to have a transparent membrane on one side.

Another result: I kept a 1/4" thick cylinder of fluid membrane device, with a single Tyvek membrane in the tank for 3 days. It never stopped working. This was just a 1/4" long piece of 1/2" diameter acrylic tube, cemented to a flat piece of acrylic to make a cup, with the fluid in the cup and the membrane held on with an O-ring acting as a rubber band. Easy, small, accurate, reliable, but no faster than the ADA style drop checker.

It seems to me, there's no good reason that the liquid inside the checker must be all liquid.

The sponge notion or porous paper etc is a good idea, it increases the surface area for colorimetric exchange and speeds up the phase into the liquid from the gas. You tend to get even coloration with less time involved.

Remember to take out the KH solution and run test of those as well to make sure the cheapy membranes are holding the water back and maintaining the ref KH.

I'm leery about those cheapy membranes, although I know that they can be used, I just want to be more certain.

Regards,
Tom Barr

I suppose a good membrane check would be to leave out the indicator solution, use a substantial portion of reference KH water, like 5 ml or so, and just leave it in the tank for several days. Then, test the KH solution to see if it is still at the set KH. The way I have been evaluating them is to watch for changes in the indicator dye concentration. If water is being exchanged with the tank water the indicator should be leaking out too. (maybe) But, I see a bigger problem - the method of using an O-ring as a rubber band to hold the membrane in place and seal it isn't exactly a perfect sealing method. Far better would be somehow heat sealing the fluid inside the membrane sandwitch.

However, a DIY device isn't supposed to be foolproof, just cheap, easy, and reasonably accurate. If someone were to make these for sale they would want to set a considerably higher standard for reliability. The really BIG problem is that we will all make a run on the local post offices and use up the national supply of Priority Mail envelopes.

I made another, almost foolproof membrane drop checker. It has a 1/16 inch thick layer of KH reference fluid, for rapid response, an O-ring to seal the membrane, and is easy loading.




The blue fluid photo was when I first put both drop checkers in the tank. The green fluid photo is 1 1/2 hours later. So far I haven't checked on how fast the membrane one reacts. I had no trouble at all putting the fluid in it, with a couple of drops of pH reagent.

My next effort is one that I think will be easier to see the color on from outside the tank, and which can be used with either a tyvek or clear membrane. But, I haven't tried making it yet. (This is fun!!)

I checked the response time for this latest membrane drop checker. It took about 30 minutes to make about 2/3 of the change in color from one ppm of CO2 to another, both for increasing ppm and decreasing ppm. Disappointiing.

It looks almost like something like Heisenburg's Uncertainty Principle - you can either get fast reaction or ease of reading the color, but not both at once. Maybe I will call it "Hoppy's Uncertainty Principle" and find the constant that connects response time to readability. (T/R=K, time constant divided by readability equals a constant.) If I get to it tomorrow I will try this with a sponge in the cavity to reduce the fluid volume, but I know that will make it even harder to see the color.

Adding a small disc of white sponge, cut from a makeup applier/remover sponge, cut the response time in about half. That is pretty good, but not as good as I hoped.

In addition I assembled a unit using Tyvek Priority Mail envelope material as a membrane approximately as shown in this sketch:



This one had the same size cavity and the same size sponge in it with the same amount of KH water and indicator reagent. It responded the same as the other one, so I think Priority Mail envelop Tyvek is a good membrane material, and again, the results are pretty good, if not as good as I hoped.

This one is very easy to make - it took me about an hour start to finish.

Vaughn,
You might consider stopping by and getting a sample of the concentrate.
I can leave out where you can grab it and pick it up.

Regards,
Tom Barr

So is this one any faster in response times?
10-30 minutes is a good goal I think.

I do not have to deal with color resolution using a pH meter
So I'm not bound by my eye perception.

Regards,
Tom Barr