I'm having a little problem with hydrometers not being linear when it comes
to error. If I understand correctly, the hydrometer is going to float based
on it's weight and the amount of water, or solution, that it displaces.
Since the change in displacement is based on the thin upper extension of the
tube which is cylindrical, the change in the volume of water it displaces in
different gravity solutions is a function of height. The cylinder's volume
is equal to pi * radius^2 * height. If the radius of the upper tube stays
constant, the displacement, height, is directly proprotional to the gravity
of the solution. To be non-linear, either the radius must change over the
length of the tube or the paper scale must have been printed in a non-linear
scale. I don't see either of these as being very likely.
I should have that issue of Brewing Techniques*** around. I'll see if I
can dig it up and post the Author's basis for non-linearity.
> That is correct, the scales a not linear, I just measured my Dujardin
> Salleron hydrometer. at around 1.000 its a .40" for 10 degrees, at
> around 1.100 it's around .32" for 10 degrees, that is a 20%
> I have Kesslers, cheapos and the French hydrometers mentioned above
> and have the ability to calibate them. In a month or so I will have
> some time available and will calibrate them and post the results.
> Nice catch, I am a metrologist and always considered them reasonably
> linear, that what winemaking texts usually say.
> PS: Jack, I have half of those hydrometer values entered into a
> spreadsheet and will post that too. ( I got the hydrometer scales
> from NIST and am hand entering them into a spreadsheet for
> Best regards
> > I recall reading an article in "Brewing Techniques" vol 7 #3, a great
> > magazine which died a couple years ago, that suggested the error curve
> > not simply linear, and that a hydrometer reading 1.000 in water at 60F
> > in fact be off as the gravity increases. If your unit reads, for
> > 1.002 in your calibration water, you cannot just subtract .002 from a
> > gravity reading as the error increases as gravity increases. The cheap
> > hydrometers typically have an error of 10% +/- This means that in a
> > solution of 1.050 a $5 hydrometer could read between 1.045 and 1.055.
> > for $5 you get a unit that is OK for low readings, but not so great for
> > readings. A more accurate scientific hydrometer which has a small range
> > spread over a long tube can be had for about $30US
> > Steve