Fully Automated Brewery Senior Design

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
We weren't looking to beat up on jgalati. We wanted to see if there is any additional information that would lead us to where the network broke down. We spoke with him and worked everything out. We are a small company and do not like to see our assets stolen and cut apart. Although, to most, the message of stolen kegs seems to be a tired one there are still a lot of people that don't understand the ownership. That being said, I apologize for "spamming" the forum... I was not aware of the private message.
 
Although I haven't updated this forum recently... here's a list of things we've accomplished:

(this is still a work in progress)

Note this is for an electric brewery, but I'll add expandability for natural gas or propane users soon. (As well as hybrid support for those who use gas to heat quickly, and electric to maintain temperature)

1) The brewery will require at minimum, 11KW to operate. However, only two plugs will be required to physically plug into the panel.
2) The system allows you to be boiling one batch while mashing the other in perfect timing with each other.
3) We've successfully parsed the BEER_XML file format and the system understands single infusion, n step infusion, and decoction mashing profiles.
4) The mobile interface successfully has connected to the beaglebone, and we can actuate pumps, valves, etc.
5) I've multiplexed the 7 analog inputs of the beaglebone to allow for up to 28 inputs.
6) We used the LM34 temperature sensor to measure temp instead of a thermister.
7) I've created a more efficient P&ID for the herms system. Meaning, the footprint of the brewery is now smaller. Also, the design allows the user to reclaim heat from the kettle.
8) I've eliminated the control box.
9) I've eliminated a ton of useless wiring
10) I've reduced the time to brew the first batch of beer by 23 minutes (on average), and subsequent batches by almost 2 hours (assuming that you start a new batch after completing the old one).
11) We've replaced this with a $55 beaglebone black.
12) I eliminated all of those PID controllers that people have been using. This eliminated complexity as well as useless expenses.
13) We're currently working on finishing a CIP loop.
14) We're also working on integrating fermentation chambers with a wireless interface that will connect to your beaglebone.

I believe there is more than enough here to constitute what my senior design culminated into...

I'll be posting a more full work up as I finish knocking out the final kinks of the project. Currently it's in transit back home, but should be here in the next two weeks.

Questions? Any clarifications?

-JG
 
Although I haven't updated this forum recently... here's a list of things we've accomplished:

(this is still a work in progress)

Note this is for an electric brewery, but I'll add expandability for natural gas or propane users soon. (As well as hybrid support for those who use gas to heat quickly, and electric to maintain temperature)

1) The brewery will require at minimum, 11KW to operate. However, only two plugs will be required to physically plug into the panel.
2) The system allows you to be boiling one batch while mashing the other in perfect timing with each other.
3) We've successfully parsed the BEER_XML file format and the system understands single infusion, n step infusion, and decoction mashing profiles.
4) The mobile interface successfully has connected to the beaglebone, and we can actuate pumps, valves, etc.
5) I've multiplexed the 7 analog inputs of the beaglebone to allow for up to 28 inputs.
6) We used the LM34 temperature sensor to measure temp instead of a thermister.
7) I've created a more efficient P&ID for the herms system. Meaning, the footprint of the brewery is now smaller. Also, the design allows the user to reclaim heat from the kettle.
8) I've eliminated the control box.
9) I've eliminated a ton of useless wiring
10) I've reduced the time to brew the first batch of beer by 23 minutes (on average), and subsequent batches by almost 2 hours (assuming that you start a new batch after completing the old one).
11) We've replaced this with a $55 beaglebone black.
12) I eliminated all of those PID controllers that people have been using. This eliminated complexity as well as useless expenses.
13) We're currently working on finishing a CIP loop.
14) We're also working on integrating fermentation chambers with a wireless interface that will connect to your beaglebone.

I believe there is more than enough here to constitute what my senior design culminated into...

I'll be posting a more full work up as I finish knocking out the final kinks of the project. Currently it's in transit back home, but should be here in the next two weeks.

Questions? Any clarifications?

-JG

Sorry if my comment before wasn't clear - I'm sure that you have done (more than) enough to justify a senior design project :D I was more meaning only throwing together a few kegs that basically just heat/cool water and nothing else is probably not going to cut it.

But I really have to point out that you are overstating 11 a bit - total build of the "panel" is going to be a bit more than just the $55 for the BBB, you still have all the other control hardware, boards, wiring, cabinet, etc. to account for. But long run you will have more flexibility/control than a Kal clone using the Brewtroller as the brains for much less.

Congrats on Graduating though, huge achievement! :mug: I am a little confused though, has this project been marked already? :confused:
 
It has. I got an A in it. But I'm still working on it to get this puppy finished. Next project on the list is to run 240 volt to the shed as well as fixing any algorithm errors that will come up while brewing a simple water batch
 
It's sanitary as in you can drink it and not get sick. The probelm is the water gets tainted with a rubber taste. Go outside and drink some and you'll not want to use that to brew with :D


That the reason you let it run for a couple of minutes. You can also buy the RV Hose at Home Despot or Lewdes or any RV outlet.:tank:
 
I use a GE whole house water filter with a Carbon activated filter. It has 3/4in in/out so you can attach a hose. I use a garden hose QD on the in and a type F cam lock on the out. This way I can run a silicone brewery hose directly from my filter to my HLT. I've drank water from the hose and hated it, after the filter... loved it. Makes a big difference.

IMG_0394.jpg
 
Although I haven't updated this forum recently... here's a list of things we've accomplished:

(this is still a work in progress)

Note this is for an electric brewery, but I'll add expandability for natural gas or propane users soon. (As well as hybrid support for those who use gas to heat quickly, and electric to maintain temperature)

1) The brewery will require at minimum, 11KW to operate. However, only two plugs will be required to physically plug into the panel.
2) The system allows you to be boiling one batch while mashing the other in perfect timing with each other.
3) We've successfully parsed the BEER_XML file format and the system understands single infusion, n step infusion, and decoction mashing profiles.
4) The mobile interface successfully has connected to the beaglebone, and we can actuate pumps, valves, etc.
5) I've multiplexed the 7 analog inputs of the beaglebone to allow for up to 28 inputs.
6) We used the LM34 temperature sensor to measure temp instead of a thermister.
7) I've created a more efficient P&ID for the herms system. Meaning, the footprint of the brewery is now smaller. Also, the design allows the user to reclaim heat from the kettle.
8) I've eliminated the control box.
9) I've eliminated a ton of useless wiring
10) I've reduced the time to brew the first batch of beer by 23 minutes (on average), and subsequent batches by almost 2 hours (assuming that you start a new batch after completing the old one).
11) We've replaced this with a $55 beaglebone black.
12) I eliminated all of those PID controllers that people have been using. This eliminated complexity as well as useless expenses.
13) We're currently working on finishing a CIP loop.
14) We're also working on integrating fermentation chambers with a wireless interface that will connect to your beaglebone.

I believe there is more than enough here to constitute what my senior design culminated into...

I'll be posting a more full work up as I finish knocking out the final kinks of the project. Currently it's in transit back home, but should be here in the next two weeks.

Questions? Any clarifications?

-JG
Can you explain in a bit more detail (and maybe some pics) how you're utilizing the BBB?
 
The BBB can be thought of as a Programmable Logic Controller, but one you don't have to program in ladder logic. The beaglebone can take up to 28 analog inputs, and 50 digital inputs (why you'd need that many idk...) and can then control any electronic device such as a pump, ball valve, solenoid valve, heating element, gas solenoid valve, etc.

The beaglebone also serves as a server for the webapp. Because it has an ARM core processor, this puppy can output 1080p, as well as host serve as a server for the mobile interface. In doing so, we've eliminated any need to install any small LCD screens or workout a complex user interface using a single button like various toggle switches you see on control boxes.

The beaglebone controls an array of FETs that act as switches for the relays. Given that the beaglebone can output 6-8mA per channel, it's unwise to directly power a relay, many of which require 20-40mA.

The beaglebone also parses beer XML files, strips the information pertinent to the brewing process, and then plugs those values into the .config file.

Does that make more sense?
 
here's a demo of the interface with dummy drivers built in:

i'd suggest 720p so you can actually read it lol

[ame]https://www.youtube.com/watch?v=VkiRQkAa12M&feature=youtu.be[/ame]
 
The beaglebone controls an array of FETs that act as switches for the relays. Given that the beaglebone can output 6-8mA per channel, it's unwise to directly power a relay, many of which require 20-40mA.
This is the part I was scratching my head over. I thought "there's no way you can use that thing to control SSR's directly" and this makes a little more sense. So did you make a cape with these FETs or is this something off the shelf?

Is this something you're going to open source, keep to yourself, or sell as a product?

The demo video was neat, but the MLT was at 220 ... what?!?!

I am also curious as to why you've stripped so much out of the control panel. I understand that multiple PID units would be excessive and maybe you don't want any screens or LCDs, but not including any physical buttons seems like a mistake. At bare minimum, I'd think you'd want an e-stop. Seems a bit unsafe without something basic like that. What was the design criteria that lead to this decision?
 
I soldered up a relay board and used the FETs as switches to "relay the relay"

The estop is the GFCI which will be installed right after the panel, of which there will also be a disconnect bar.

If I have some money to throw around I'll put a shiny red button in, but as of now, if the beaglebone is running, then I'll be able to control it through HTML posts.

1401679312368.jpg
 
I read almost all of this and found only one issue...

3) You shouldn't use solenoid valves, they actually end up clogging because of the sticky wort. you should use these electric ball valves instead- https://www.oscsys.com/store/valves

There are actually sources for brass versions to save some coin, instead of the stainless ones.

I didn't like the limitation of those ocsys 3-way valves in that you can only go from input A to output B, or A-to-C. The manual 3-way valves can go any-to-any, and have all three open for easy cleaning. Mated to a rear window wiper motor from a bone-yard Chrystler (Cherokee or similar), via gears printed on 3D printer. It is significant additional design/deployment effort, but I liked the result
 
Gents I swear to god eventually I'll update this to its full glory, but I just started working and no kidding that once you get a job you don't have time for anything. Always thought you guys were crazy. HAH.
 
Gents I swear to god eventually I'll update this to its full glory, but I just started working and no kidding that once you get a job you don't have time for anything. Always thought you guys were crazy. HAH.

Congrats on the job - who are you working for?
 
I didn't like the limitation of those ocsys 3-way valves in that you can only go from input A to output B, or A-to-C. The manual 3-way valves can go any-to-any, and have all three open for easy cleaning. Mated to a rear window wiper motor from a bone-yard Chrystler (Cherokee or similar), via gears printed on 3D printer. It is significant additional design/deployment effort, but I liked the result

I built a quick little Arduino sketch that uses one of the OCSsys 3-ways on the cold water input of the chiller, and opens/closes it incrementally, in 100ms jogs and measures the output temperature until it reaches target. When it initially opens, it quickly opens for a longer period to try to intercept the target temp quickly, then adjusts from there. The initial flow of the wort can be quite off from the target temp, but by the time a gallon or so is in the fermenter, it's pretty close.

Those 3-ways work well for that since they have a separate Open and Close wire. The 2 ways are controlled on board the valve and only fully open or fully close.
 
Those 3-ways work well for that since they have a separate Open and Close wire. The 2 ways are controlled on board the valve and only fully open or fully close.

I can stop my two way valves too, depends on the type you purchase.
My two wire 3 ways I stop by removing the voltage at the center position to close off both port for a special application.

Cheers,
ClaudiusB
 
Back
Top