Refractometer question

[Neto]

So i have a refractometr and i calibrated it with 35ppt solution that came with it a while ago. After that, i test fresh ro water and it reads like 5ppt below 0. I know some people calibrate with fresh ro and others with a 35ppt solution. I am now concerned that my calibration solution is bad?

 

For those of you who calibrate with 35ppt solution, whe. You check RO water, does it read at 0?

 

Thanks

[gmerek2]

I have 2 calibration fluids. They are both ever so slightly of from each other. You kept it sealed right? No evaporation. I'll check with fresh out of the *****-it RO water later and will report back.

[gmerek2]

Yea mine reads well under 0. Yours too? I flushed the refractometer surfaces with copious amounts of water and wiped clean before all tests. Kind of concerning. Maybe a temperature difference?

[Neto]

I have 2 calibration fluids. They are both ever so slightly of from each other. You kept it sealed right? No evaporation. I'll check with fresh out of the *****-it RO water later and will report back.

Yep I keep the calibration fluid sealed.

 

Yea mine reads well under 0. Yours too? I flushed the refractometer surfaces with copious amounts of water and wiped clean before all tests. Kind of concerning. Maybe a temperature difference?

Thanks for checking, yeah mine is well under 0... 

[jaddc]

This comes up from time to time. I'm a scientist, and I use RO water to zero my refractometer.

 

It is most likely that your 35 ppt is not actually 35 ppt. If some water evaporated, then the concentration would be higher. 

 

I'm a scientist, so forgive me for being anal-retentive. 

 

The refractometer's optics are calibrated at the factory and hobby-grade refractometers cannot be multi-point calibrated by the user.

 

At the factory, they use standard solutions to allow them to place the scale sticker in the right place on a lens. This principle is just like an analog thermometer. If you watch how they make thermometers, they etch the scale on last.

 

http://www.sciencechannel.com/tv-shows/how-its-made/videos/how-its-made-thermometers.htm

 

The end users cannot actually calibrate a thermometer. If you boil water, and stick in a thermometer and it reads 50 degrees (Celsius or Fahrenheit) then there is nothing you can do but throw it away.

 

What we do is a "single-point calibration"  or zero the device. It doesn't matter if you zero with RO water or with calibration solution. It is easier to zero with RO water because it is more constant. RO water will be always zero. 

 

There is a good argument as to why to use 35 ppt calibration fluid for more accuracy in professional settings. But our hobby grade devices do not have the resolution to take advantage of high accuracy. For example, we need to know if the water is 35 ppt. We don't need to know if it is 34.8 or 35.2, nor can our devices give us that information.

 

So, zero with RO water and worry about one less thing in this hobby. 

Edited August 3, 2014 by jaddc

[lnevo]

That's not entirely true...being that they are hobbyist devices the farther away your sample is from the target calibration, the more error deviation you introduce. If you use 35ppt fluid all that matter is the line that you compare any other mix. If you take a faulty thermometer and it says 50 when you put it in boiling water you add a mark and consider that 100. All you need to know if something is now at 100 you can just compare with your line same thing with a refract omelet.... Now put that same thermometer in ice and marking it at 0 does not make the 100 mark suddenly accurate and you should not go use that calibration line to measure 100 degree water any longer.

Edited August 3, 2014 by lnevo

[Rburgee]

Temperature has a large part in this. When I got my refractometer manufacturer recommended to test at 78 degrees Fahrenheit but the salt manufacturer recommended testing at 79 degrees. I split the difference and recalibrate the instrument once a week to maintain consistency. Make sure there are no air gaps between the glass and plastic as well.

[jaddc]

That's not entirely true...being that they are hobbyist devices the farther away your sample is from the target calibration, the more error deviation you introduce. If you use 35ppt fluid all that matter is the line that you compare any other mix. If you take a faulty thermometer and it says 50 when you put it in boiling water you add a mark and consider that 100. All you need to know if something is now at 100 you can just compare with your line same thing with a refract omelet.... Now put that same thermometer in ice and marking it at 0 does not make the 100 mark suddenly accurate and you should not go use that calibration line to measure 100 degree water any longer.

I disagree.

 

The precision of the refractometer does not change given same temp and consistent measuring practices. In other words, the error deviation does not change depending on how one moves the scale. The error is the same and mostly due to user habits.

 

The hobby devices do not have high magnification to the scale and reduces the accuracy. Zeroing at 35ppt does not increase accuracy significantly. High end refractometers use high end microscopy optics to zoom into the scale. They also have better prisms and optics as well. That leads to better accuracy. Well trained analysts lead to great precision.

 

Let's be clear. When you single point calibrate the device, all you are doing is moving the prism to align the printed scale with the refraction line. Since you are not changing the prism's shape (and thus refraction physics) nor the rulered scale, you do not affect the precision of the device.

 

The situation you described with the thermometer is indicative of a work around with a faulty thermometer and not proper practice. Not standard procedure. If you etch 100c in boiling water and put it in ice water and mark 0c, then when you put it in boiling water it should read 100c. If it does, not then it is non functional. Usually this happens when a bubble is introduced into the liquid.

Edited August 3, 2014 by jaddc

[jaddc]

Temperature has a large part in this. When I got my refractometer manufacturer recommended to test at 78 degrees Fahrenheit but the salt manufacturer recommended testing at 79 degrees. I split the difference and recalibrate the instrument once a week to maintain consistency. Make sure there are no air gaps between the glass and plastic as well.

Yes. The easiest way to be precise with respect to temp is this:

 

Add a drop or two of water on the prism. Remove air and close cover slip. Wait 30 seconds. This will equalize the temperature between the prism and the water.

 

If you want to see the effect of temp in the reading, then put a drop of cold tank water on the prism and look through the refractometer as the water warms up. It is not dramatic, but the line does move.

Edited August 3, 2014 by jaddc

[lnevo]

Your assuming the scale and prism are accurate. If you mark the line at 0 and 100 on a faulty marked thermometer you are changing the scale and it is no longer faulty since its been calibrated. The scale on the refractometer doesnt matter when you only need one point. If you are only mark 0 you have no clue that the 35 mark is anywhere close to accurate. You are assuming the scale and prism are accurate. When you use 35ppt fluid you know that you are matching the reference. The precision and magnification are irrelevant except in the case you described to measure to a specific precision 34.8-35.2. But zero'ing with RO may work great on scientific grade equipment but on a hobbyist device you only care about the measure you are trying to reach, in which case 35ppt fluid is the best practice. Advocating otherwise is dangerous plain and simple. It may work on your refractometer and maybe on others, but on ones manufactured cheaply etc?

[lnevo]

And sorry to harp but a better analogy would be calibrating a level. By the principle of zero with ro/di that would like calibrating a level with a 45 degree incline versus a known level surface... If you calibrate at 45 degrees you are trusting the lines marked by the factory and the worker that lined it up.

[jaddc]

Your assuming the scale and prism are accurate. If you mark the line at 0 and 100 on a faulty marked thermometer you are changing the scale and it is no longer faulty since its been calibrated. The scale on the refractometer doesnt matter when you only need one point. If you are only mark 0 you have no clue that the 35 mark is anywhere close to accurate. You are assuming the scale and prism are accurate. When you use 35ppt fluid you know that you are matching the reference. The precision and magnification are irrelevant except in the case you described to measure to a specific precision 34.8-35.2. But zero'ing with RO may work great on scientific grade equipment but on a hobbyist device you only care about the measure you are trying to reach, in which case 35ppt fluid is the best practice. Advocating otherwise is dangerous plain and simple. It may work on your refractometer and maybe on others, but on ones manufactured cheaply etc?

I am assuming the prism and scale are accurate. If you buy a refractometer from a company that does not have a reputation for producing an accurate device then you are wasting your money and your time. If you buy a 10 cent ruler, but 12 inches does equal 12 inches, then what is the point?

 

Most refractometers that you will find at reputable hobby stores are accurate. You can zero at zero with rodi and save time and worry. Check the reviews.

 

If you go on eBay and buy a ten dollar device from Slick Willies' discount things from a stolen truck, then all bets are off.

 

All measuring devices: ruler, level, thermometer, etc. are calibrated by factory workers. The better trained, the higher the salary and thus higher the price. You get what you pay for.

Edited August 3, 2014 by jaddc

[lnevo]

You aren't wasting anything if you use an accurate reference. My 10 cent ruler that is 9" and says its 12" measures a 6" line just fine after i calibrate it with an accurate 6" measurement. But i wouldn't use its scale and calibrate only one inch.

 

Point is not to assume and calibrate as close as possible to what your measuring. Same principle as calibrating a ph probe with 7 and 10 fluid and not with 4 and 7. Yes 4 and 7 will calibrate it, but it will be more accurate to what we measure if you use 7 and 10. It's just best practice.

[jaddc]

No. You are wrong.

 

Well, look if you are ok with using an inaccurate ruler, then I got nothing add. I like to buy rulers that are correct without the need to spend time calibrating. I zero the ruler at the beginning of what I measure and read out how long the distance is like almost everyone else.

 

I don't go to NIST and calibrate my ruler against a standard foot long rod, re etch my scale, and then make my measurement.

 

Think of everything that use daily that we use zero as a reference. Scales. Thermometers. Levels. Rulers. Speedometers. Odometers. There is nothing fundamentally different about a refractometer that warrants treating it otherwise. In fact, if you treat it carefully and don't knock it or drop it, you shouldn't need to zero often. We zero because the lens can move over time. That's all.

 

The ph scale is not a linear scale. It is logarithmic. If you think a refractometer is to calibrated like a pH probe, then we have identified how your logic is faulty.

Edited August 3, 2014 by jaddc

[jaddc]

Also, we cannot change the scale on the refractometer. If the factory screwed up a ruler, then can recalibrate and make a new scale with a sharpie. You can re-etch the thermometer. But you cannot change the scale of a refractometer. Thus you cannot 2 point calibrate the device.

 

You can only zero, or a one point calibration.

[lnevo]

Exactly my point. One point calibration. Calibrating at zero is useless on a measuring device

[jaddc]

Do a google search or YouTube search on refractometers and how to use them.

http://maple.dnr.cornell.edu/kids/refractometer.htm

http://www.homebrewstuff.com/refractometer-how-to

 

Among others

[lnevo]

I can use google too... Make your own decision..

 

http://blog.captive-aquatics.com/captive_aquatics/2011/01/refractometer-calibration-the-myth-and-risks-of-using-rodi.html

 

Mine is pretty clear to me...

[lnevo]

And here's some expert advice laying out the real issues as well when using hobbyist grade refractometers. You may not know Boomer as he sticks to Manhattan Reefs, but he's otherwise a pretty well known authority.

 

http://www.manhattanreefs.com/forum/chemistry/159347-calibration-fluid-refractometer.html#post1444522

[AlanM]

Independent of equipment, calibrate close to what you want to measure. Universal principle.

[jaddc]

Independent of equipment, calibrate close to what you want to measure. Universal principle.

Yes. But we are not two point calibrating anything. We are one point calibrating a linear scale. Do you calibrate your bathroom scale with a standard weight? Or just make sure it says zero when you are not standing on it

 

Whatever. When I put rodi water on my refractometer it reads zero. When I put 35ppt on my refractometer it reads 35 ppt.

[AlanM]

If I had a calibrated 200 lb weight, yes, I would calibrate the bathroom scale with a single point to near my weight. No brainer. I dont happen to have one, so I use zero. Every fish store that sells refractometers, however, also sells 35ppt calibration solutions, so it is silly, except to avoid paying $5 and accepting a less accurate measurement, to use deionized water to calibrate your refractometer. You are picking a known point to calibrate to, but you are picking the one farthest away from what you want to measure.

[AlanM]

btw, I have no idea what mine reads with di water on it.

[jaddc]

I just want to be clear. There is nothing wrong using a standard solution. It is just more work, time and expense.

 

My point is that it is unnecessary. In many optical readings, a blank works just fine as means of rough calibration.

 

For our needs, a blank is sufficient.

[jaddc]

btw, I have no idea what mine reads with di water on it.

Try it!