Hi there, I'm trying to find/build a specific gravity sensor in order to measure fermentation in my homebrewing setup. Does anyone know if such a beast exists?
If not, what do you reckon is the best way to interface an arduino to a hydrometer? I've thought of using a potentiometer with (ideally) zero physical resistance, or a light sensor array to sense the location of the highest point on the hydrometer. If I go along this route, I need millimeter accuracy, so I'm not sure how easy the light sensor option would be.
Google returns a bunch of links on search line : "electronic hydrometer".
How to interface it to arduino would depends on specification of the device, analog , digital, modbus, 4-20 mA etc.
a potentiometer with (ideally) zero physical resistance,
I doubt such a thing exists.
I tried Googling "Digital hydrometer" and got many results, however most seemed to be very expensive industrial grade or the poster got confused with a "hygrometer" which I think measures humidity.
I gather you're happy to let your existing hydrometer float permanently in the brew, you just want to read it's level.
I do home brew and would like something similar, but no obvious method springs to mind.
Maybe if you confine the hydrometer in a tube that's open to the brew, place a light-weight disc of material on top, then read the distance from the lid with ultra sonic or one of those nifty (Sharp?) range finders.
@Graynomad. Yeah, I was thinking of a tube open to the brew, I hadn't thought of using a rangefinder though, I can't think how I missed it now that I think about it. I'll have a think about that one.
I had a brief look about, and alll the options that I saw for buying sensors were wither too expensive or too annoying to interface to an arduino (I'm already using one for temperature control, might as well just add to it), or both. I'll keep looking though. One that I saw uses a hall effect sensor, which may be another option.
O'K, if we are not into design high-tech sensor here, one of which could be using change in laser beam polarization depends on sugar level ( they should be cheap nowadays, any diabetic has this device in his pocket), than I'd consider contact-less transformation mechanical movement (cheap existing hydrometer) to electrical value.
Options:
optical: led/lens - lens/photo-diode and nontransparent object closing light path;
capacitive/inductive: like it done it theremin, advantage - direct frequency reading possible;
IR/ultrasound: range-finders, disadvantage - price, not easy to mount ;
and yes, hall sensor.
Submerged weight hanging by a thread from a $10 load cell (as used in kitchen scales).
The measured weight varies as the density of the surrounding liquid changes.
mjbmikeb:
Submerged weight hanging by a thread from a $10 load cell (as used in kitchen scales).
The measured weight varies as the density of the surrounding liquid changes.
The problem would be that buoyancy would always yield a measure of zero. But if you could keep the object rigidly in place, then you'd have something to measure, as the liquid tries to force it up (high density), or its own weight tries to pull it down (low density).
You could also put the weight on an arm, and use a hall effect or opto to measure angle of declination as it sinks or rises in the liquid - then translate that into density.
Assuming the weight is made of something like solid glass (density 2.5) which is denser than beer, then the thread will always be under tension.
If you reduce the beer density, there is less force pushing the weight up, therefore the tension on the thread increases.
Take away the beer and you measure the full weight minus the small effects of the surrounding air buoyancy.
mjbmikeb:
Assuming the weight is made of something like solid glass (density 2.5) which is denser than beer, then the thread will always be under tension. If you reduce the beer density, there is less force pushing the weight up, therefore the tension on the thread increases. Take away the beer and you measure the full weight minus the small effects of the surrounding air buoyancy.
I was thinking that if the object was floating in the water, then it was at equilibrium, therefore any scale would register zero. But of course if you can somehow consistently keep it out of equilibrium, you could measure the difference as density changes. So for instance if you kept a balloon underwater, the scale (measuring from below) would register it's desire to 'pop up', which would change as density changes (poor example - the balloon would likely expand/contract with pressure/density - a rigid object, like a toilet tank float, would be better, but I really didn't want to talk of toilet floats and beer in the same topic). Likewise, your example of the glass - since it would normally sink, it was never in equilibrium, and so would always have a (changing) force. Sorry I didn't catch that - I was thinking of floating objects.
For anyone who cares, the math on buoyancy (full or partial) is spelled out at Buoyancy - Wikipedia
Thanks guys, this is a really good idea. I had been thinking along the lines of using an existing sensor and figuring out an interface, but this is much better. I have a load cell from some scales in front of me as I write, and will do some prodding this week. I'll post my findings. Thanks again!
As specific gravity can be a tricky and possible costly thing to measure directly there have been many creative solutions used in the present and past, some posted here already.
I'll just throw out another method. If you have a way to measure the absolute liquid level height of a liquid in a vessel (many ways) and a way to measure the pressure head of the liquid (pressure sensor tapped off lowest level of vessel) then you have the two variables needed be used to calculate the specific gravity of the liquid material.
I really like the hanging weight solution. My O'Haus Cent-o-Gram balance has a little hook and a special platform to put the beaker of liquid on specifically for measuring specific gravity.
Let's say you could find a stainless steel ball bearing that had a volume of exactly 1 cubic centimeter. You would weigh it in air in grams then weigh it in your brew, also in grams. The difference between the two weights would be the specific gravity, right? Of course you would have to scale it to the actual volume of your weight but since we know the specific gravity of water, it would be easy to calculate its volume.
Edit: Looking a little deeper into alcohol production, it looks like the delta of specific gravity measurement needs to be accurate to one part per thousand. That might be outside the capabilities of the $10 Chinese load cell. Maybe not.
The ina125p instrument amplifier came in the mail just the other day, and I had a little play last night, but I have nothing worth reporting yet. I like the pressure/volume solution, but airlocks will negate the pressure value after a certain point.
but airlocks will negate the pressure value after a certain point.
Air lock? If the vessel pressure isn't at atmospheric, just use a DP pressure sensor, Sensor and HP leg mounted equal or below vessel bottom, LP leg to top of vessel, that will cancel out vessel pressure changes, probably should liquid pack LP lead and back out the resulting offset in the calculation.
Aah, I get you. I was thinking that you meant measure the increasing pressure caused by the release of CO2 into the tank, but the airlock is designed to prevent this buildup. That's a good idea though, and I think I have a DP sensor lying around somewhere too. I'll keep working on the load cell option for now, because I want to avoid putting holes in the body of the vat, but if it doesn't work out, then I'll give the DP a shot. Thanks Lefty!
EmilyJane:
Edit: Looking a little deeper into alcohol production, it looks like the delta of specific gravity measurement needs to be accurate to one part per thousand.
Wow – one part in thousand is quite a challenge.
Even pure water, which is generally considered non compressible, will change its specific gravity approximately one part in thousand for every 10C. I would also assume that other “contaminants” (such as alcohol and whatever goes into a decent brew) may be further influenced by temperature.
Yeah, that's probably the case, but I'm happy with a ball-park idea. The original plan was just to notify me when the beer has stopped fermenting for three days, and given that the brew will stay within 2 degrees of a particular temperature, I reckon it should be good enough to do that, even if it can't give me an accurate alcohol content reading. Also, the scales my load cell came from were accurate to 0.01g, so I feel I should be able to get close enough for what I want to do.
That should give you pretty good accuracy, especially if you calibrate with distilled water at a known temperature just before you test your beer. Then I would just kick back with a Bundy and Coke and wait for the right results.
