[a10t2]

I know some people have had good results, but I've never found the refractometer correlation to FG that everyone seems to use to be very accurate. I modified it slightly to give better results, based on my last ten sets of readings. If you have data from both a hydrometer and refractometer I'd be interested in hearing how the new numbers compare with the old.

Refractometer Estimates of Final Gravity

[tygo]

I just started using my refractometer with my last batch so I have almost nothing in the way of consistent data.  I found that with the three pre-fermentation readings that I took (1st runnings, 2nd runnings, total pre-boil volume) using both the hydrometer and the refractometer the refractometer was consistently about 2 points higher than the hydrometer.

The other reading that I took was as the fermentation was winding down in the primary.  I sampled the wort and took a measurement with both the hydrometer and refractometer.  I was surprised that after using the refractometer tool in beersmith that both readings were exactly the same at 1.015.  So based on that extremely limited sample...well I guess it's too soon to draw any conclusions at all.

Good write-up.  I'll be continuing to gather data on future batches starting with the two I'm doing tomorrow.

[richardt]

Sean, good stuff--look forward to the emergence of a quick "rule of thumb" with refractometer usage as a result of your investigation.  I certainly like the covenience of the refractometer.

Tygo, by "2 points higher" do you mean gravity points or Brix points?

[Kaiser]

I also noticed a significant discrepancy between the corrected FG from the refractometer compared to a hydrometer reading. That's why I have given up on using a refractometer for beer. Maybe it is time to not only calculate a correction factor but come up with a new fit for the Brix reading to FG correlation altogether.

Kai

[a10t2]

Maybe it is time to not only calculate a correction factor but come up with a new fit for the Brix reading to FG correlation altogether.

That's my long-term plan, but with two independent variables, I'll need a lot more data points. I may end up doing some small test mashes just to get outliers at the extremes of gravity and attenuation.

[tygo]

Tygo, by "2 points higher" do you mean gravity points or Brix points?

Gravity points.  My refractometer has both a brix and SG scale.

[Kaiser]

Sean,

You should be able to get the data point you need by not even having to brew beer. Model solutions made from water, malt extract and alcohol should work although some of them should be compared to equivalent beer samples.

Kai

[a10t2]

Gravity points.  My refractometer has both a brix and SG scale.

Has any sort of correction (dividing by 1.04) been done on the scale itself? If it reads in °Bx and SG but is still calibrated for a sucrose solution you'd have to compensate when measuring wort.

It could be due to evaporation, if the wort was hot enough.

Model solutions made from water, malt extract and alcohol should work although some of them should be compared to equivalent beer samples.

That's a great idea. Thanks Kai!

[a10t2]

So now that I had a little more free time to play around with it, it turns out that I was able to get a really good fit for a new, independent equation. The SD for the new correlation (only 12 data points, though) is 0.0005 SG.

http://seanterrill.com/2010/07/20/toward-a-better-refractometer-correlation/

[narvin]

I've noticed that the approximation for abv ends up about 3 gravity points low for me.  However, I also did a series of tests against different concentrations of sugar water and the gravity according to my refractometer is around 0.4 brix lower than my hydrometer.  This implies that there's some systematic error in my unit.

Have you investigated how accurate your refrac is against a hydrometer on sugar water?

[a10t2]

I've noticed that the approximation for abv ends up about 3 gravity points low for me.

In the spreadsheet? Low relative to what? It uses an approximation based on real extract, so it will be lower than (OG-FG)*131, but also more accurate.

Have you investigated how accurate your refrac is against a hydrometer on sugar water?

Yes, I calibrate using both water and a 20% sucrose solution. My hydrometer is two points low for both ("paper slip", presumably) but the refractometer is dead on at 0.0 and 20.0. In your case, are you sure it's the refractometer that's off, and not the hydrometer?

[narvin]

I've noticed that the approximation for abv ends up about 3 gravity points low for me.

In the spreadsheet? Low relative to what? It uses an approximation based on real extract, so it will be lower than (OG-FG)*131, but also more accurate.

Have you investigated how accurate your refrac is against a hydrometer on sugar water?

Yes, I calibrate using both water and a 20% sucrose solution. My hydrometer is two points low for both ("paper slip", presumably) but the refractometer is dead on at 0.0 and 20.0. In your case, are you sure it's the refractometer that's off, and not the hydrometer?

My refractometer is low compared to the promash calculation... I just attributed it to being low in general.  My hydrometer is accurate with distilled water, so I trust it, but i'm not positive which was correct. 

[Kaiser]

Sean. Thanks for sticking with this. I recently used the hydrometer and the refractometer on beer again and yes, the discrepancy was rather large.

Kai

[bluesman]

I haven't had any consistent results using my refractometer on beer and/or wine "after fermentation". My readings as compared to my hydrometer readings are significantly skewed and inconsistent.  Good Luck!  I hope you can find a solution.

[a10t2]

Thanks guys. My main concern at this point is over-fitting to the current data set. That's why I'm hoping I can get a few other brewers to keep track of both measurements.