I need to verify the accuracy of my new Refractometer

[mladencovic]

I just got a new refractometer from marinedepot and when I compared it with my hydrometer, the discrepancy was way too large: 1.030 refr. vs 1.023 hyd.

Now, I know that there might be Ca deposits, dirt and wear on my 1 yr old hydrometer, but just to make sure, I would like to check the refractometer/water with someone in my neighbourhood - Plainsboro (near Princeton), before I fully trust it and start correcting my salinity.

p.s. Refractometer seems to be calibrated properly since RO water tests 1.0

 

[panmanmatt]

hat's actually not an uncommon difference between a hydro and refractometer. You would actually be amazed how far off those swing arms can be.

I wish I was local to help you check.

 

[Carlo]

I'm not exactly local but if you want to take a trip to Bordentown I can help you out. I have 53ms solution on hand that I use to calibrate my digital salinity meter with.

If you were to bring the reference temp card it came with we could warm the solution to that of your normal tank and get a precise calibration for you so you don't have to cross reference the temp.

BTW, if it's a swing arm type hydrometer then it's probably on the low side. If you have the glass float type I bet it's pretty darn close to the real sg. The glass floating type about 1 foot long with built in temperature correction are darn accurate.

Carlo

 

[GregW]

Couldn't it be calibrated with RO/DI water which would be 1.000?

 

[IrradiatedReef]

Do not be shocked to see between .004 - .010 variation error on swing arms! I see it all the time at work when people come in because all their crabs just died and I find their salinity is 1.034!! Soak your swing arm in distilled white vinegar overnight and and deposits should be removed. I have seen new out of the box, including mine, that are off by .004...

 

[Carlo]

Couldn't it be calibrated with RO/DI water which would be 1.000?

No, that's about the worst thing you can use. You want to use a calibration fluid that is at sea sg.

Carlo

 

[GregW]

Couldn't it be calibrated with RO/DI water which would be 1.000?

No, that's about the worst thing you can use. You want to use a calibration fluid that is at sea sg.

Carlo

If you search "calibrate refractometer" on Google, almost every site says to calibrate with distilled or RO water.

 

[Carlo]

I've read that numerious times and just shake my head every time I read it. It's no wonder why so many hypo treatments fail.

I just don't get why people try and calibrate at base ZERO instead of using a known solution in the center of the usuable range.

Carlo

 

[mladencovic]

Well, strictly speaking for calibration of ABSOLUTELY ANY instrument in the world you need 2 points if it measures any linear quantity.

(this cool nerd emoticon from Yahoo should be added to NJRC)

Measuring RO water 1.0 establishes base point and it MUST be done so that you do not have parallel shifts in the measurements, and another point MUST be calibrated as well, to establishes the correctness of the scale. That is why, for example, pH metters must be calibrated with both 7 and some other solution (usually 10). Speaking of pH metters, their calibration is actually recomended with 3 points (4-6,7 and 10) because pH is nonlinear (logaritmic) function.

 

[Carlo]

I understand the logic however it doesn't work with sg devices very well for marine aquarium use. You can quote text books all day long but I guarantee if you calibrate a cheap (under couple hundred bucks) refractometer or hydrometer using RO/DI water you are going to be off mixing your salt water.

Most digital pH meter manufactures will recommend NOT to try and calibrate with all three points but only to use the 2 solutions closest to the range you work in. Of course the farther you get out of range from the set point solutions the more "off" your reading is likely to be from the true result. For marine environments you would always want to use the 7pH and 10pH salinity solutions and not the 3pH solution.

Same goes for setting up a cheap hydrometer or refractometer. First you have to either warm/cool the solution to a specific temperature or find the temperature of the solution and cross reference this with the manual. Then depending on the model you may need to do an extra step if it doesn't have ATC (automatic temperature correction). If you have a higher end model you can set it up using two reference points. However most only really work with one reference point. Even some models are better being set using only a 53.0 mS reference solution.

There are a couple of issues with using RO/DI water. First you are assuming that the water is a perfect reference point and it probably isn't as the RO/DI unit isn't perfect. Second and biggest concern is that it's way to far away from the specific gravity of our tanks to be a truly useful reference point. If you set just one reference point using a 53.0 mS reference solution then you just calibrated your device at tank salinity so the device is going to be a LOT more accurate.

PS I do agree with your first statement mladencovic but in reality we don't care about a linear measurement. We really only care about a couple points from the 53 reference point as that is what we are aiming for.

PPS If you plan on using the device for a hypo treatment then the best thing you can do is mix RO/DI water together with a 53mS reference solution in proper parts to create a new reference solution in the range of the hypo target specific gravity. Then once calibrate you will be much more accurate.

BTW, the whole RO/DI refractometer thing has been kicked around a lot on RC and pretty much the end result was that almost everyone was surprised how far off they really were in the calibration once they used a true 53.0 mS reference solution.

Carlo

 

[JohnS_323]

And to bring this back to the original post, this is an order of magnitude question only. Calibrating a refractometer for normal aquarium use with RO is fine. Yes, maybe you wouldn't want to rely on it 100% for hypo'ing a tank, but that was't the question.

Sorry, sometimes I suffer from big brain overload reading these posts. I'd wager 99.9% of the people on this site used RO to calibrate their refractometer. I'd further wager that none of them experienced any problems as a result.

 

[Carlo]

John, I do agree if you use the same device (regardless of calibration) to always mix your salt and are having good results then sometimes "good" is "good enough" even if not optimal.

In mladencovic's original post he was getting results of 1.030 with the refratometer vs 1.023 with the hydrometer. That's a drastic difference assuming they were both calibrated with RO (he didn't say the hydro was). The real sg is probably somewhere in the middle (lets hope). Even if both devices show a different result the difference would probably be no where near 0.007 if they were both calibrated with 53.0 mS solution instead of RO. The difference would probably be more like 0.001 which for practical purposes is not a concern nor important.

Taking hyposalinty treatments out of the conversation since that had nothing to do with the post itself the main reason I like to calibrate at 53.0 is to save money! (covered below)

Typically in my experience swing arm hydrometers will read low and refractometers will read high when calibrated with RO. If you mix up salt water based on that you end up using more additives with the hydrometer and use a lot more salt then needed with the refractometers when doing water changes (throwing away money). Either way the cost of these items will far outweigh the cost of the calibration solution if you do routine water changes. This IMHO is the main reason to calibrate accurately based on a solution at the desired end point. SAVING MONEY longterm. I wasn't thinking of "tank problems" as the reason to calibrate.

If anyone is interested I could bring a big bottle of calibration solution to the next meeting and everyone can bring their devices with them and see how accurate they are against a base solution of 53.0mS/1.025sg. I'd bet we'd see a wide discrepancy in the results.

Carlo