Measurement on sample tubes in TFTestkit don't match Taylor sample tubes. Which is correct?

[PoolBrews]

I have recently noticed that the 10ml and 25 ml lines on the TFTestkit and Taylor sample tubes do not amount to the same amount of water. I had always used the TFTestkit sample tubes, filling to the appropriate line, then dropping in my speed stir pellet, then testing.

I purchased a Taylor Sample sizer, and it indicates it only works with the Taylor 9198 sample tube. If I add water to the tube, then use the sample sizer, I get exactly 10ml or 25 ml in this tube. If I then pour the contents into the TFTestkit tubes, it is always less than the 10ml or 25ml markings. I have attached photos, but which is correct?

 

[Newdude]

Mine were off too. My speedstir tube does not match my TF-100 tubes. My pool was very forgiving and being about 5% off didn’t matter.

 

[Newdude]

I’m sorry I forgot to answer your question. Lalalalalala. Anywho the sample sizer is designed for the specs/size of the Taylor tube so use that to ensure accuracy.

 

[Jimrahbe]

PB,

I don't seem to have that problem.. When I use the Taylor tube with the sample sizer I get 10 and 25 ml readings.. If I pour that water from the Taylor tube into the TFP tubes they read 10 and 25 ml.. But... if I use the TFP tubes and the sample sizer, the readings do not show the correct levels...

Odd..

Jim R.

 

[PoolBrews]

I know I have to use the sample sizer with the Taylor tube - the real question here is which tube is actually 10ml? Obviously one is 10ml and the other is ~11ml or one is ~9ml and the other is 10ml. (Or neither are correct)

Having been an engineer in a manufacturing facility for a good part of my life, I find it hard to understand how either of these tubes could be off - yet one is! Based on measurements, we are looking at a 10%-15% error here. That's quite a bit in my book.

 

[PoolBrews]

PB,

I don't seem to have that problem.. When I use the Taylor tube with the sample sizer I get 10 and 25 ml readings.. If I pour that water from the Taylor tube into the TFP tubes they read 10 and 25 ml.. But... if I use the TFP tubes and the sample sizer, the readings do not show the correct levels...

Odd..

Jim R.

I did the test twice to ensure I didn't do something wrong... I just found it interesting! I want to keep doing my chlorine tests only in my chlorine tube. I guess I'll use the sample sizer in the Taylor tube, then pour the water into my chlorine tube.

 

[PoolBrews]

Maybe I got a defective batch of tubes in my original kit

 

[mknauss]

@OTPirate

FYI

 

[PoolBrews]

@OTPirate

FYI

Not sure what this means?

 

[mknauss]

Not sure what this means?

Notifying the owner of TFTestkits about this thread.

 

[Jimrahbe]

PB,

I have two Taylor (sample sizer) test tubes.. One that I use for FC tests and the other for all my other tests.. Because I always use the sample sizer, I never use the original TFP test tubes any more.

I did rummage through my cabinet and found my original TFP test tubes and that is what I used when I ran my test in response to your post. My original TFP test tubes are about 6 years old, so maybe something has changed in the meantime.. ???

I am not sure what the problem is, but my recommended "fix" would be to buy another Taylor sample sizer test tube..

Thanks,

Jim R.

 

[HermanTX]

I purchased a Taylor Sample sizer, and it indicates it only works with the Taylor 9198 sample tube. If I add water to the tube, then use the sample sizer, I get exactly 10ml or 25 ml in this tube. If I then pour the contents into the TFTestkit tubes, it is always less than the 10ml or 25ml markings. I have attached photos, but which is correct?

In my opinion, the Taylor 9198 sample tube using the sample sizer would be the most accurate. Taylor has been making these for a long time so feel confident they should be the most accurate.
I do not have a sample sizer but I did put 10ml in my Taylor tube and poured it in the TFT tubes (2 different tubes) and I got 10ml fill in each. So mine appear to be equally accurate.
You may have gotten a bad batch from TFT.

 

[OTPirate]

If I am reading correctly, the sample sizer does not work with TF-100 cylinders. it only works with the Taylor cylinder.

 

[proavia]

If I am reading correctly, the sample sizer does not work with TF-100 cylinders. it only works with the Taylor cylinder.

You are correct - you can't use the sample sizer accurrately in the TFTestkit cylinder. But I believe the OP is saying that if the Taylor cylinder is filled to 10 ml and then that water is poured into the TFTestkits cylinder, the TFTestkits cylinder only shows 9ml.

Personally, I use a 10ml syringe (the kind one might use to give a young child liquid oral medication) and don't worry what the reading is on the testing cylinder. Of course, I cross checked the accuracy of the syringe against both cylinders and all 3 match. It's just quicker for me to use the syringe.

 

[Stoopalini]

I've noticed this as well when comparing the TF-100 tubes to the Taylor tubes. I figured the amount of variance was small enough to not cause an issue, but the variance is definitely there.

 

[dschlic1]

I purchase my test kit some years ago. About a year after I purchased it, I checked the markings accuracy using a graduated cylinder. Mine were definitely off by about the amount shown. I just took a permanent marker and marked the correct levels. It seems that occasional some of the TFTtestkit vials come out wrong. there was a thread about this several years back when the test kit first cam out.

 

[mgtfp]

I checked the markings accuracy using a graduated cylinder

Or a graduated syringe, if you don't have a graduated cylinder, should be cheaper for a one-time verification.

 

[BenB]

Could try using a digital scale with 0.1g accuracy. Reminder for the non metric folk, 1ml of water weighs 1g.

 

[Dirk]

As someone that works in the print industry, I can guess what's going on. The markings on the Taylor viles are raised plastic. They are part of the mold that creates these vials. If one assumes the markings are accurate, Taylor would get very reliable, repeatable accuracy from every vial. There would be more than one mold, of course, but they likely are all manufactured from the same master mold (the mold that makes the molds).

Where as the markings on the TF vials are printed on (or screen printed or some such similar ink transfer process). That would require the printer to calibrate the printing machine regularly to align the markings correctly on the vials. And even if he did so consistently, and often, and perfectly (which is iffy) the little vials themselves probably get jostled a bit in the process and may not perfectly align while they're getting their markings applied.

Something like that. That was a greatly simplified explanation. The point was, the TF vials undergo a multistep process that is more likely to introduce a margin of error, which is why some of you are seeing this, and varying degrees of this, and others are not. I like the darker printed markings, but the printing process is bound to be less accurate than the injection molding process.

I bet if @OTPirate pulled a bunch of her vials and set them on a flat surface and pushed them together, she'd see that the black lines don't always align...

Or, yes, it could have been just a bad, isolated batch, but the error would have been caused by what I described.

If she cared to, OT could filter out these hooligans. She'd need to find a vial that was accurate, and make that her "master standard." Then it would be a simple matter to gauge the vials just before she assembles the kits to make sure each vial's markings align with her master vial.

 

[Dirk]

If you look at the TF vile in that first pic, you can see how much farther away the 5ml line is from the bottom than it is from the 10ml line. That suggests the whole set of lines is shifted upwards, misaligned. The Taylor vial is similar in that respect, but the Taylor vial has a concave bottom that would explain why that is. The first 5ml segment looks to be larger in volume than the next 5ml segment, but it isn't, because the concave shape of the bottom is taking up some of that volume.

The bottom of the TF vile looks to be flat (I don't have one to verify that.)