HOWTO - Make a BrewPi Fermentation Controller For Cheap

AWESOME!!! :rockin:

 

yes i did.

the problem wasn't just the adruino.

for some reason the internet browser on the PI I was using would not work well with brewpi. I used the midori browser and everything popped up like it was suppose to.

 

Hmm. Most people access the web interface from other computers or devices with chrome, firefox, etc. Are you doing all the work on the PI directly? I gotta tell ya, it would be much easier to use ssh and your web browser from your mac.

I'm glad you got it working though .

 

Yeah I've got my UI setup to kiosk a browser in full screen mode. Then it runs a page with 4 iframes each showing a separate instance of the lcd text for each of my fermenters. I can't run anything else I'm at my limit.

 

Say ^what^?

Cheers!

 

Graphically I mean. I can tunnel in and use external browsers and ****. I just can't run anything else on the display

 

Ah, got it.

No problemo: Ask Sandy Claws for a bigger display

Cheers!

 

Running it on a 15' display.

 

No you're not.

Check the calendar. April 1st is months away...

Cheers!

 

Yeah I am it's an old laptop display I got a converter for off eBay

 

"Running it on a 15' display."

You missed your own joke....

Cheers!

 

Hah

 

That ought to be big enough...

 

fwiw, semi-exotic Tek fuses arrived.
Current meter wired into my BrewPints clone - which is identical to its sibling save for being short two flow meters.

First current measurements with a full suite of components all running from a single common 5VDC mini brick (fwiw, they took $10 for two units, delivered, a couple of months ago) peaks at 770 milliamps, total.

That's the peak when a flow meter causes a tap list update, which should be as strenuous as it gets for this little machine.

So here's what's running on that supply:

- RPi Model B
- AlaMode running RaspberryPints sketch
- Sainsmart Uno R3 (slightly modded for direct 5VDC supply) running BrewPi sketch
- Sainsmart 2 channel relay board (cooling channel active)
- 4 port USB2 hub, two ports used
- Edimax Wifi dongle
- nameless wireless mini keyboard dongle
- 8 DS18B20 probes (five for my logger and three for BrewPi)
- 4 SF800 meters
- max of 5 leds on at any time

When the 60mm 5V fan kicks on (second relay channel active) the total draw is 950 milliamps peak.

Pleasantly surprised, I expected closer to 1.3 amps worst-case.

Random stuff:

- wifi dongle ~50 milliamps higher system power versus 100mbit enet, which itself adds 50 milliamps when connected. I suspect most of the 100mbit power consumption is actually for its three on-board leds.

- mini kb/mouse dongle: 10mA max, best as I can tell, sending a stream of repeat characters full speed to a terminal session.

- meters: ~5 mA each. Shouldn't matter whether idle or active, the IR led is always on and that's going to dominate the consumption there.

More when I get to it. But bottom line is these things are remarkably efficient...

Cheers!

[edit] More data. SWAG puts the accuracy in the +/- 20% neighborhood.

- With everything but the RPi disconnected from power, shutdown current is 120mA
- Idle current (booted but not doing much) is 350mA
- Add the AlaMode (with 5 DS18B20s and PIR attached) atop the RPi the idle current rises to 440mA
- Add the R3 Uno (with its 3 DS18B20s) and the 2 channel relay module attached but script stopped and no actuators active, idle current is 560mA
- Add the USB2 hub with wifi and mini-kb dongles and idle current rises to 670mA
- With all the scripts and services running the "idle" current rises to 710mA
- Finally, as mentioned earlier, a flow-meter inspired tap list update peaks the total system 5V current at 770mA...

 

day trippr with your bluetooth idea would something like this work? It's using a DS 18b20, just not sure if it's available as a bus device.

 

Anybody uses I2C LCD 20x4 on your build? Could share the step by step?

 

Just finished mine up, thanks for the great DIY. Here are a few photos of it complete. I added a couple led strips to mine so the inside of the case will light up blue if cooling or red if heating.

 

Very nice!!! I like the enclosure and the LED's

Here's my final rendition, all DIY with working encoder and additional freezestat to keep the AC evaporator from icing up.

(The other STC1000 controls the kegerator which is separate from the fermentation chamber)

 

I just noticed you said I2C... The one I built uses SPI like the one you can buy from Brewpi. Not I2C

Yep, but that will take me some time to write it up...

You'll need a 20X4 HD44780 Display like this

A shift register chip like this

A 10K pot to adjust the contrast on the LCD

three 10K 1/4 watt resistors

a 100 nf capacitor

a 10uf capacitor.

And a protosheild like this

PM me an email address and I'll send you my notes.

 

Wow! 254 pages and 2̶5̶3̶9̶ 2540 posts. I'd say Fuzzes little thread is all grown up. It happens so fast.

After "faking" it for a while (by using my setup as a mere monitoring system then to just run a heating pad in my mash tun converted to a fermentation "chamber") I finally got a chest freezer and put an IPA in last night for a diacetyl rest as it is. Basically just pulling up the temp from ~65 to 70 degrees F over the course of a day and holding it there for a day or so. I may do a cold crash before dry hopping just because. I don't have the remote access set up so I can't say how things are progressing but it looked pretty stable this morning.
Anyway, just thought I'd chime back in and post an official "it's alive" message. Everything is still loose so I need to corral all the components and make some extensions for the temp probes.

<EDIT> More like trying to get of few more points lower on FG than a yeast cleanup. This fermented in my crawlspace and things were running on the cool side so I thought I'd warm things up a bit for any yeast that may still be partying. Plus, a pretty good test of the new freezer and BrewPi.

Todd

 

Interesting. A dual relay module with either a wifi or BT dongle mounted on top. Thanks for the pointer.

It could be very useful in other applications, but it doesn't solve remoting the Arduino via a wireless link to the RPi. Remoting the relay and sensors introduces extra complexities to BrewPi, namely the Arduino now has to pull ds18b20 data from that relay module thingie and send actuator commands to it.

Otoh, there's a really good chance the BT solution I'm going to try won't require any BrewPi changes, other than a device pointer and perhaps a baud rate tweak.

Fingers crossed...

Cheers!

 

As I posted yesterday, I'm messing with the inaugural run of my BrewPi set up. I got home to find the freezer sitting at a chilly 9 degrees F and the beer at a crashingly cold 36! At first I thought the F/C toggle had been switched. Nope. Checking the device configuration showed a duplicate chamber heater on pin 5 which is the same pin as the original chamber cooler. At some point when the BrewPi called for heat it energized the impostor relay and the system was caught in a loop ... running the freezer while thinking it was asking for heat. It had been working fine for about 18 hours prior to this with the freezer and heater tag teaming to hold the target temp. I reset everything, unassigned the rouge device and put my beersicle back in to dry hop at a warmer temp.

So .... gremlins, short circuit? I do have things sort of hanging about until I can get them in an enclosure. Hopefully there won't be a repeat this evening.

Todd

 

Similar thing happened to me when I switched the "Device Slot" of my cooler. Had you done that while you were messing with it? I found I had two cooling devices on the same pin and the web interface claimed one was being controlled while the other was constantly on. This is a bug that Elco is working on. I manually deleted the incorrect devices and everything went back to working as it should.

 

That's probably it. I did switch the pin assignment since my freezer plug design required it to be on the "bottom" outlet (otherwise it blocks the other outlet) which was originally assigned to the heater. I made the pin switch and hit Apply after each one but I guess there was some corporate memory at play that kicked in at some time during the profile. When I discovered the duplicate, I changed all the pertinent selections to None, Unassigned, etc. so I guess that counts as manually deleting it (?). If it cold crashes again today I'll dig a little deeper and try again. Guess I should have done a true shakedown with some water but I thought I had everything figured out.

Todd

 

Remember, you are doing everything from a browser interface. Clicking apply isn't enough you have to refresh sometimes to make sure your settings go through. I click apply then I refresh the device list everytime to keep this from happening.

 

Sage advice. After beating back the glacial advance yesterday evening and resetting the device configuration I did refresh and all looked normal.

Todd

 

So I have a Bluetooth usb dongle running on my office RPi and talking to the Bluetooth Serial module hanging off an Uno. Left the two chattering over night and they're still on good terms (albeit only a couple of feet apart - have to push that out to a few meters tonight).

Next I need to get the BT/serial module into its AT command mode and set the serial side to match the 57.6K baud rate that the BrewPi sketch uses. Hopefully BrewPi will then find the Uno.

If I don't over-indulge while kinda/sorta watching the football game tonight I should get the first part done...

Cheers!

 

Good luck

 

Well, so far no joy. Luck hasn't made an appearance tonight.

I'm trying to use a small AVR script and simple wiring setup that purports to take strings sent to the Uno over the USB side and send them on out to the HC-05 through a software serial port. It also enables a wire function that allegedly forces the HC05 into AT mode at 9600 baud.

You'd think that's all so simple it should just work. Well, I can see symbols going into the serial side RX pin of the HC05 but it never sends anything out its TX pin to ack the AT commands I'm sending.

On top of that, I just noticed that Windows sees the module as an HC-06 not HC-05. Not sure what that means yet.

You'd think you could just program the darned thing through the Bluetooth side but noooo.

More when I get to it. Right now I need a beer...

Cheers!

[edit] The module can actually drive its transmit pin - verified by using BrewPi to try to program the Uno over the Bluetooth link.
Wiggles the heck out of the TX pin - just not at the right speed. So it's not broken, it's just being recalcitrant...

 

I've messed about with these cheap bluetooth modules before. I got it to the point that by modifying the optiboot bootloader and using some cleverness in the sketch, I was able to upload a new sketch using AVRDude (or ArduinoIDE) just as if connected by pure serial.
There are some gotchas. You may need to determine your module firmware and what AT command set it supports. I found that not all baudrates were working reliably (specifically I seem to remember that 9600bps was borked). Software serial has a lot of pitfalls... Higher baudrates may not work, duplex may not work. I'd try to avoid it...

 

Um....What?

Cheers! (seriously, that's some major word salad right there )

 

One mystery solved: the interface header on my BT/serial module is not wired in the "expected" manner, so the setup that was supposed to put it into "AT mode" would never happen. Going straight to the pin on the MCM got a reaction that hadn't happened before - the LED is doing what was described as what should happen - but haven't gotten a response out of the thing.

Going to need to back up and verify this guy's serial-serial implementation is sound. Should be able to do a loop-back test to prove it actually captures received characters - which right now is the missing piece of the puzzle.

Double brew day tomorrow so this may have to sit for a day...

Cheers!

 

I am mounting 2 arduinos uno with one raspberry and my doubt is if have a way to turn one arduino off but no unplugging the respective cable. This situation occur when only one fridge is fermenting, and the other one is waiting the next batch.

Thks,

Fabiano

 

If using a pc as the RPI, does the computer have to be attatched to the arduino at all times? I have an acer aspire1 that would be perfect as a linux based device for this purpose. However, I want to use it as more of a monitoring device whereby I may carry it with me and check the data on my brew. If it must be connected is there anything (economical) I can use as an alternative to transfer the data to this pc via wifi? Keep in mind I know nothing of this stuff. Just want to nerd out on my brews.

 

THe RPI logs, so the PC logs, the Arduino runs the temp-controlling show.

So, the ferm chamber would run the temp control just fine without the PC but you'd have no logging to see what was going on. Also making changes to the K constants to load to the Arduino if you wanted to change them.

 

You need the computer to log the data, the arduino can control the fridge on it's own but it wont save the data for you to look at those pretty graphs.

day_trippr is working on getting a wireless connection to his arduino using bluetooth, if he can get that to work that may be a solution for you.

There are other wireless options, such as using a pair of rf transeviers to relay messages, but they quickly get costly. At that point you might as well get the rpi.

 

My BrewPi is currently set to 40 degrees Fridge Constant, and it says it is cooling, but it is not running. I suspected the fridge may have died, but a reboot caused it to cool for a bit, and then give right back up. Everything was fine, and nothing was touched to cause this change. Any thoughts?

 

You can simply use the web management page for the idle BrewPi instance to set the Temperature Mode to "Off". The setting is right below the graph.

As mentioned, if you want to capture the data to fill out the chart, you need to have something grabbing the data packets from the Arduino in real time (the AVR has no room to queue up that data or the code to do so).

As also mentioned, I've been working on getting an essentially transparent wireless link between the Arduino and RPi running under BrewPi using cheap Bluetooth devices.

So far it has been an adventure that started with trying to identify and purchase the right module. Not only are many ads showing something other than what they're actually selling, the market is utterly polluted with knock-offs that express varying degrees of functionality and equally varying wiring configurations. Add to that receiving a clearly defective module and my efforts slowed to a stop as of the wee hours this morning, and a handful of these things are going back to Amazon today.

But now that I've been educated about these things I finally found someone selling what appears to be the right design (yay!) and a pair of boards should arrive in a day or two. Fingers crossed.

Anyway....Bluetooth doesn't have much range, so it may not be a viable solution depending on your spread. In my intended use it's to tie my brewery fridges to the RPi running my keezer, and they're all within 10-12 feet of each other.

What BrewPi really needs is support for a wifi link as the AVR-RPi transport. But I expect that'd take some major tinkering with both the AVR and RPi code, and Elco is deep into his brewery controller development these days...

Cheers!

 

First thing is to plug the fridge straight into an outlet and see what happens. If it solidly comes back to life we can work on figuring out why your BrewPi system is losing control...

Cheers!

 

Fridge sometimes needs to have its internal thermostat set to coldest, mine was set to somewhere normal and while STC+ was trying to make fridge run, fridge thought its own thermostat was making it plenty cool enough. Answer was to set fridge to absolute coldest setting, then when STC+ sent power, fridge would run, regardless of its thermostat