BrewMonitor project

[megabyte]

Hi Guys,

Just a quick intro to a project I've been working on-and-off (mostly off ) for far too long.

It's called BrewMonitor and the idea is pretty simple - to hook up a recording device to your fermenter and record temperature and airlock activity. That way you can easily answer questions like "what was my lag time?", "what was my peak temperature?" and "how well is my fridge controlling the temperature?".

Obviously it's not a replacement for taking specific gravity readings so it won't tell you when to bottle or keg, but it could help to determine when you should start taking those final gravity readings.

The project incorporates an electronic circuit (the BrewMonitor) and some software that communicates over USB allowing you to explore the data it records.



The airlock sensor uses an infrared light gate and a 3D printed mould to hold it in place on a standard gooseneck type of airlock. I have also been playing around with serial CCD cameras that could be an option in the future for all you glad-wrappers! :lol:


The simplest mode of operation is to run the BrewMonitor stand alone (without a computer attached) by powering it from a spare USB charger or battery bank, this way it just records the data internally and you can download it from the device later over USB. Alternatively, you can leave it connected to a computer (a spare laptop perhaps?) throughout fermentation and the software can send notifications to your phone when certain conditions are met.

I'm currently coding integration with Pushover API to send notifications to your phone like:

• "You've reached 17C, it's time to pitch your yeast!"
• "Your temperature has crept up to 25C. You might wanna ice-bath that baby"
• "Your bubble rate has slowed to 50% of its peak, now might be a good time for a diacetyl rest"
• "Your airlock is slowing down. Time to take some S.G. readings and free up a keg"

Here's an example notification:

I'm also working on live data uploads to thingspeak.com for those that like to watch their yeast reproducing over the internet. h34r:

The source code, 3D models and PCB designs are all going up on GitHub over here along with some more information about the project. Let me know if you'd like to help contribute and I can help out with some PCBs to get you started!

 

[goatus]

Very cool.

Have you attempted converting the airlock flow rate sensor into an estimated gravity based on the amount of CO2 estimated to have left the chamber? If you can work out some kind of algorithm to do that even remotely accurately given the starting gravity and the amount of flow that would be an amazing tool. You would have to feed it your starting gravity, wort volume, fermenter volume, and somehow estimate what an average "bubble" volume is, and then you could get a somewhat accurate volume of CO2 produced, which could be converted to a current gravity estimate? I'm not sure if different (or even the same yeast) would consistently produce the same amount of CO2 in the same wort, but well worth a crack at testing the data.

I'm currently using the BrewPi for ferment control / logging. The creator keeps teasing of a way to log gravity readings along with temp (I believe with some kind of weight on a line that dangles into the wort). There was also a kickstarter bluetooth hydrometer that sits in your wort and tilts as the gravity changes and takes readings via a tilt-sensor and feeds it back via bluetooth. Loving that there is also another approach (and your way nothing has to touch the beer).

 

[megabyte]

Have you attempted converting the airlock flow rate sensor into an estimated gravity based on the amount of CO2 estimated to have left the chamber? If you can work out some kind of algorithm to do that even remotely accurately given the starting gravity and the amount of flow that would be an amazing tool. You would have to feed it your starting gravity, wort volume, fermenter volume, and somehow estimate what an average "bubble" volume is, and then you could get a somewhat accurate volume of CO2 produced, which could be converted to a current gravity estimate? I'm not sure if different (or even the same yeast) would consistently produce the same amount of CO2 in the same wort, but well worth a crack at testing the data.

I'm currently using the BrewPi for ferment control / logging. The creator keeps teasing of a way to log gravity readings along with temp (I believe with some kind of weight on a line that dangles into the wort). There was also a kickstarter bluetooth hydrometer that sits in your wort and tilts as the gravity changes and takes readings via a tilt-sensor and feeds it back via bluetooth. Loving that there is also another approach (and your way nothing has to touch the beer).

You're spot on, and regretfully the answer is no I'm not planning to make it measure S.G at the moment. Currently there's the BrewNanny and BeerBug projects that try to accomplish this, but I'm still yet to be convinced that either are reliable or convenient enough to substitute manual S.G. measurements. There's a fantastic thread over on HBT where a number of people are throwing alternative ideas around but we're yet to come up with a robust solution.

Regarding that kickstarter, are you referring to the Brewometer that fell short of funding? I thought that was an elegant solution except that as soon as it touches the side of your fermenter it will read the wrong gravity. I was working on a similar setup with a linear ccd array inside a clear floating tube, but avoiding the krausen and holding it away from the walls are the pitfalls. I still have hopes for a self-resonant piezoelectric sensor and I've run some correlation experiments between wort reflectivity throughout fermentation detailed over here that are very promising alternatives for the future.

Alas, the BrewMonitor is much less ambitious, it just watches the airlock and temperature and it does this pretty well. We can still gain a lot of insight from these parameters and augment them with a couple of manual S.G tests IMO.

 

[professional_drunk]

Have you thought about modifying brewpi to support your airlock bubbler? It already does graphing functions and displays them via a webserver on raspberry pi. Instead of a custom pcb, how about a spark photon?

 

[klangers]

Very cool.

Have you attempted converting the airlock flow rate sensor into an estimated gravity based on the amount of CO2 estimated to have left the chamber? If you can work out some kind of algorithm to do that even remotely accurately given the starting gravity and the amount of flow that would be an amazing tool. You would have to feed it your starting gravity, wort volume, fermenter volume, and somehow estimate what an average "bubble" volume is, and then you could get a somewhat accurate volume of CO2 produced, which could be converted to a current gravity estimate? I'm not sure if different (or even the same yeast) would consistently produce the same amount of CO2 in the same wort, but well worth a crack at testing the data.

I'm currently using the BrewPi for ferment control / logging. The creator keeps teasing of a way to log gravity readings along with temp (I believe with some kind of weight on a line that dangles into the wort). There was also a kickstarter bluetooth hydrometer that sits in your wort and tilts as the gravity changes and takes readings via a tilt-sensor and feeds it back via bluetooth. Loving that there is also another approach (and your way nothing has to touch the beer).

A million times easier to just put the fermenter on load cells. This will give you an accurate reading of how much CO2 has escaped as the mass will gradually reduce.

 

[TheBlackAdder]

A million times easier to just put the fermenter on load cells. This will give you an accurate reading of how much CO2 has escaped as the mass will gradually reduce.

Skeptical on measuring the weight of air with a load cell...I would think the mass difference would be less than the uncertainty of measurement

 

[klangers]

Fair comment, let's do some calcs.

Let's say we have a 20L wort at 1.060. This will weigh 1.060 x 998 x 0.02 (SG x density of water in kg/m3 x 20L in m3) = 21.158 kg. After dropping to 1.02, this would weigh 20.3592 kg. This is well within the accuracy of load cells. Granted, they would need to be decent quality but they can typically resolve down to 0.001 kg (1 gram).

 

[TheBlackAdder]

Point taken
I was too lazy to do the maths

This would be a good method to trail, adjusting for solids that fall out of suspension, which may be negligible

 

[Drick]

Love the idea.
Even if it only monitors the airlock and temperature, it'll should still sell at the right price.

 

[megabyte]

Have you thought about modifying brewpi to support your airlock bubbler? It already does graphing functions and displays them via a webserver on raspberry pi. Instead of a custom pcb, how about a spark photon?

Yep I think the BrewPi inputs would support the airlock sensor but I fear the BrewPi code might need some tricky changes since you have to sample the ADC at a high frequency and do some calculations to monitor the bubbles.

The particle photon is a great candidate for adding WiFi to the BrewMonitor but I'm also considering adding an ESP8266 module to the existing design and porting the code to Arduino. Honestly the biggest deciding factor is what the community prefers the most out of the two options.

A million times easier to just put the fermenter on load cells. This will give you an accurate reading of how much CO2 has escaped as the mass will gradually reduce.

You'd also have to factor in yeast production but I've heard it being done over on the HBT forums with good results. What does a gram-accurate load cell that can handle 25kg cost?

 

[sponge]

Now to modify for the glad-wrappers

 

[Mardoo]

We'll need a kitten recognition system to start...

 

[megabyte]

Aww snap.. and here I was thinking we might get through the first page without being struck my dem lolcatz :lol:

Mardoo! (I'm saying that in in a Colonel Klink "Hoogan" voice)

 

[professional_drunk]

I think the end game really needs to be a temperature controller. The thing we are missing is a smart temp controller that is also legal in Australia. STC1000 and Brewpi's with SSR's aren't legal. Fully assembled 240v power tail switches don't exist either.

I've recently put together a temp controller for a cheese fridge using arduino and a rf transmitter that can control 3 arlec rf power sockets (fridge, fan, humidifier). That was my solution to getting over the legalities.

So my suggestion is some sort of

• open hardware (arduino, particle)
• legal 240v switching such as the rf example above
• rasbpi webserver or cloud for graphs and stuff
• optional humidity probe for cheese makers.

Add your hardware design skills and that would be a product that would be interesting.

 

[megabyte]

Nah, the end game replaces your fridge and heater belt altogether -

Let's not get confused - temp controllers control temperature, BrewMonitors monitor your brew. Like compare the market, compare the meerkat right?

(respectfully though, awesome idea with the Arlec switches and great suggestion with the cheesemakers too!)

 

[idzy]

All very exciting stuff, well done Michael.

 

[klangers]

You'd also have to factor in yeast production but I've heard it being done over on the HBT forums with good results. What does a gram-accurate load cell that can handle 25kg cost?

Shouldn't have to do that. It's a simple mass balance. Unless you're conducting nuclear reactions inside the fermenter (changing mass to energy or visa versa E=MC^2), net change in mass = mass of escaped matter. In this case, CO2 (and any samples taken or moisture ingress which has then condensed).

The yeast doesn't appear from nothing, it simply grows from molecules which were already present.

I will admit I do not know the cost, and this may be prohibitive.

 

[Mardoo]

I was checking out load cells on eBay a few weeks ago. You could probably come in at a reasonable cost.

 

[MichaelM]

A million times easier to just put the fermenter on load cells. This will give you an accurate reading of how much CO2 has escaped as the mass will gradually reduce.

 

[MichaelM]

I agree with you Klangers. It's one of the methods some production breweries monitor the gravity.