View on GitHub


Jay Slater

Patents and paths forward

So, I’ve been informed that the people behind the Tilt hydrometer have a US patent on the concept, so that’s out the window. C’est la vie. If it weren’t a good idea, it wouldn’t have been worth pursuing or, indeed, patenting.

Ever onward, though! We are creative folk, and surely we can figure out some way to take the same measurement.

Wireless floating hydrometer

The patent covers tilting hydrometers in particular, not wireless floating hydrometers in general. If you build a tilting hydrometer without the tilt, what you get is an ordinary hydrometer. The trouble then becomes detecting the liquid level. There’s a perfect sensor for this, manufactured exclusively by Milone Technologies: a long resistor-tape whose resistance varies based on the depth to which it’s immersed in liquid. It even comes with a food-safe enclosure and a premade connector.


  • We don’t have to redo very much work; we can just remove the GY521 and add a connector for the resistor.
  • It probably simplifies the circuit design somewhat, although I’ll have to set up two separate circuits going into the ADC.


  • The water level sensor is $50, and none of the other options are nearly as good, or even solid-state. Cleaning a float sensor would be a huge pain.

Wired sinking hydrometer

Hydrometers are devices which read specific gravity, but measure buoyancy. That’s what your average hydrometer does: measures the difference between a reference buoyancy (namely, the hydrometer’s neutral buoyancy) and an observed buoyancy, when the observed buoyancy is greater than the hydrometer’s neutral buoyancy.

Is the direction of that inequality important? In fact, it is not, as my brewing partner observed. Equivalent to the floating hydrometer is a sinking hydrometer, where the observed buoyancy is less than the hydrometer’s neutral buoyancy.

Say you have a sinker which masses 50 grams and occupies 20 cubic centimeters. (I have no idea if these are realistic numbers. I’m only using metric because it’s convenient for water mass/volume.) It displaces 20ccs/milliliters/grams of 1.000-gravity water, for a net weight of 30 grams. If the liquid’s gravity is 1.100, the displaced liquid masses 10% more and the buoyant force is 10% greater; the net weight of the object is now 26 grams.

It’s much harder for humans to pull on a string and say, “This weighs 26 grams,” than to look at a scale on a float and say, “That’s between 1.095 and 1.105.” Crucially, though, it’s easier for computers.


  • Waterproofing becomes almost a non-issue. The only things going in the carboy are an aluminum load cell with insulated wires and an entirely inert sinker, and only the sinker actually goes into the wort. (A separate temperature sensor float might, too, but could be designed to be permanently sealed.)
  • The microcontroller/PCB no longer needs to fit into the cylinder, which opens up room again for using the Wemos dev board. (I’m still going to make the PCB, though. It’s cool, and without having to hit the 20mmx70mm target, I can make it user-friendly, with molex connectors for the various sensors.)
  • May actually be more accurate than the tilting models, given that it could easily be sized to hang between the trub and krausen.
  • No longer depends on battery power or indeed an ESP8266-compatible microcontroller.


  • A bigger departure from the initial concept.
  • More difficult to sanitize.
  • Requires a specially-modified stopper.


Ultimately, the guiding light for this project is cost, followed closely by ease of assembly. I think the wired sink hydrometer is definitely the way to go.

Written on June 19, 2018