Cedar Rapids Brewing Society

I was sick and bored yesterday, so I decided to calibrate my refractometer and hydrometer. Since standard Brix solutions are the most straightforward to accurately produce, I mixed up 250 gm each of 0, 5, 10, 15, 20, 25, & 30 Brix solutions.

If you're not familiar with the Brix unit of measurement, it's based on a percent weight of sugar in solution and is effectively identical to degrees Plato and Balling. To make 250 gm of a 10 Brix solution, for example, you mix 25 gm of Sugar with 225 gm of water (25/250 = 10% sugar solution, by weight). Since my blade scale measures down to 0.1 gm with good consistency, this seemed like the easiest approach to accurately generate a known solution.

I measured each of the solutions with my refractometer and hydrometer. I corrected the hydrometer values for temp and then converted them into equivalent degrees Brix, using an average of 7 calculators (Beer Tools Pro, iBrewMaster, http://onebeer.net/refractometer.shtml, http://www.northernbrewer.com/refractometer-calculator/, http://seanterrill.com/2012/01/06/refra ... alculator/, http://www.brewersfriend.com/brix-converter/, http://brew.stderr.net/refractometer.html). After capturing the data, I repeated the measurements, did some simple math, and built the following plots of deviation from ideal. One is expressed in degrees Brix, while the other is in percent.

I was surprised that my refractometer was off high by a very consistent 2% up to 20 Brix, but it leveled off to 1% above that. My hydrometer is all over the map, but pretty consistently reads low. Given that I found it in the basement of a house I bought in the UP 10 years ago, it should only surprise me that it's even this close. I rarely use it anymore, since I have a finishing hydrometer and the refractometer-given values have always been within the margin of uncertainty in my brew-house efficiency.

On a side note, I also learned that if you have a refractometer that reads in Brix, you need to apply a fudge factor to get from the calibrated scale (done with a pure sugar solution) to the actual degrees Brix/Plato of beer wort. This is due to the small percentage of non-sugar components that comprise beer wort but still influence the refractive index of the solution. Most sites I've found indicate a fudge factor of 1.04 is effective. In my case, then, I should take a wort reading in Brix, divide by 1.02 (since it reads 2% high), and then divide by 1.04 (wort fudge factor).

I'll post a poll in the Tech Topics forums, but if anyone is interested, this would be a pretty simple Tech Topic to assemble. I would make up some standard sugar solutions and anyone that is interested could build calibration tables for their hydrometers and/or refractometers.

Also, for anyone that is interested, the raw data and various additional notes can be found here:

https://docs.google.com/spreadsheet/pub ... utput=html

Does this imply that you could calibrate your refractometer using a known standard for a more accurate reading?

I've been wondering how far off the readings are at sparge vs lager pitching temperatures, I may have track these on my next brew days.

Maybe. Normally you calibrate a refractometer with water (tweak the screw until it reads 0), but that really only adjusts the zero crossing. Mine obviously has an error with slope. I would assume I could calibrate it to a commonly-occurring gravity, say 10 Brix, but that would just shift the error distribution around.

I would expect digital refractometers have a two-point cal process, much like pH or DO meters. Since the scale in my optical refractometer is printed (etched?) the slope is fixed so only a 1-point cal is possible.

Why are you concerned about refractometer accuracy at different temps? Mine has ATC, but I'm not sure why it matters since the wort sample is at the refractometer's temp within a few seconds. The one thing I have noticed is that the refractometer will read a different value if it gets too cold (around 40F). Maybe it's compensating for ambient temps?

I remember reading ATC was more to stabilize the reading as the device temperature changes slightly as opposed to sample temperatures. Although you are only using ~50 µL of wort so the thought is that the lens can fairly easily quench the sample.

The couple lab grade refractometers have used contained heated stages to address the issue of sample temperature, more than anything this is the reason why I question the temperature-error of refractometers.

I believe that is correct about ATC. This has limits, however, so you should try to keep the refractometer at around room temp. Certainly don't calibrate too warm or cold. I keep mine in the kitchen while brewing during the winter.