FrankRizzo - 30-10-2008 at 17:21
Do any of you guys think it might be possible to hack the internals of a good sonar stud finder into measuring the density/s.g. of a water solution?
The idea would be to use the information to continuously monitor the fermentation of a malt solution (beer). 
watson.fawkes - 30-10-2008 at 18:57
I can't speak to what's on the inside of a stud finder, but ultrasonic range finder units are readily available. The one whose spec sheet I've read
personally is the Ping))) from Parallax. That unit is designed for use in air. Getting the transducers bonded to your container or pipe may take some doing. The
industrial units I've seen use a transducer pair mounted on opposite sides of a pipe; they're calibrated on that pipe only. I'd recommend doing
likewise and figuring out your mechanical requirements early in the process. Then figure out how you're going to calibrate it.
FrankRizzo - 30-10-2008 at 20:04
Another guy mentioned that measuring density this way would be troublesome due to the small bubbles created during the fermentation. I think this is
a dead idea. Thanks though.
watson.fawkes - 31-10-2008 at 06:50
The effect of bubbles is to spread out the signal. It speaks against using reflection geometries, but straight-line transmission ones should work. The
received signal will be weaker in the presence of bubbles, but it will still be there. The trick is to look at just the first peak from a ping. There
will be a longer tail that contains a bunch of junk.
Also in your favor is that this is a continuous monitor, so at any given time (after startup) you've got a previous estimate of the density to work
with as a starting window. And you don't have to get clean signal from each ping; you can do multiples, toss out outliers, and average the rest.
This technique has been used in environments far noisier than beer.
Swede - 1-11-2008 at 07:21
Frank, I think that is a cool idea, and WF nailed it in that you execute a run or two using both the ultrasonic, and a more traditional hydrometer.
Parallel the data and look for correlations. If there is a repeatible measurement using the ultrasonic that corresponds to the change in density as
fermentation progresses, then simply plot it, and if the signal to noise is useable and repeatible, you've got it knocked. Doesn't matter what the
signal value is, so long as you've verified it does in fact correspond to specific gravity.
There might be problems if you change the beer recipe, from one that creates more/fatter bubbles to one that may create fewer/finer, but perhaps the
signal will be good enough to use.