Robots and Kegerators

So after the brew experiment in June I have continued to brew several batches of beer. The latest and greatest is a Scotch Ale. Hops and Berries here in Fort Collins has made it far too easy to continue with this hobby. It has officially begun to run amok. In order to keep up with production I have purchased a kegerator with all the tubing and pressure vessels required for kegging. All told it was not cheap but I don't have to wash as many bottles. What used to take an hour now is done in 10 or 15 minutes.

As for the bubble counter I'm laying out a pcb so I don't have to worry about dropping it. I had one instance of little fingers exploring the colorful wires and thought it time to convert the design. The schematic so far looks like this:


Inspired by another Robothon went and dug out parts to rebuild a rover platform. The first version was built back a few years ago but it was parted out for other projects. I went back and parted those out so I could rebuild the rover in hopes that I can work on some basic object detection / avoidance. Maybe put a camera on it and explore my crawlspace. Here is a quick video of it on my kitchen table.

Data!

After several attempts a getting some good contiguous data I have succeeded in logging a fermentation...er sort of. The graph below shows the bubble rates (the number of bubbles through the bubble lock divided by minutes), the ambient temperature (few inches above the board) and the temperature of the outside of the barrel. Ok so the temperatures aren't all that exciting, the one exception is the temp of the brew bucket coming down as the wort cooled at the very beginning. The cool part are the rates of gas production that give a clear idea of when things are really cooking.

You might notice a spike the evening of the fourth. I realized after some thought that I had bottled the latest IPA that evening and had to move my fermenter to get to the carboy. I can only assume that the spike in gas production was caused by agitating the brew as I slid the barrel along the floor. Otherwise a rather nice curve with a bit of a left shift. Time to get back to building the pages to help archive the brews.

The pages...

So we have some sketchy looking forms / pages with some dodgey functionality. In the screenshot we can see where we can fetch a brew record from a drop down. This populates the brew info, journal, and ingredients. We can then add ingredients or journal entries from other forms that post back to the db. The cost / unit is calculated on the fly everything else is pulled out directly from the db. We can also add new recipes and start new brews with the buttons at the top. If anyone wants to chime in with suggestions on how to purty this one up let me know.

Database

Here is the data model for the mySql database thus far (click on the image to enlarge it to something legible). It needs some work (the relationships are undefined) but its a start. The work on the php pages has begun but have stalled with the discovery of Aspect Oriented Programming. I'm currently torn on weather to proceed with the quick and dirty pages that are in development or to scrap everything and take a more methodical approach. We'll see what time allows...

Das Brewmeister!

It started with a photointeruppter, and ended with fermented barley and hops. My brother in-law needed a way to analyze biomass fermentations and the volume and or rate of gas produced by a waste management system. My brother who is an electrical engineer setup a sensor to count bubbles. I took some of the hardware and added some code and a database so I could pipe a count to a database and graph the results.

Enter Das Brewmeister!

The microcontoller on the breadboard takes sensor data from two one wire digital temperature sensors- one on the barrel (taped with only the finest blue electrical I have in the garage) and another a few inches from the board (for ambient temp) . On the water lock sticking up from the barrel there is a photo interrupter that counts bubbles for us.


The laptop is running a VB program to parse a serial string from the microcontroller and push it out to a mySQL database. The database then serves as a back end to a PHP page that either displays the data or graphs it on a line graph applet. Initially I thought we could sample every minute for several days for a nice robust log.

Once things got setup and the fermentation began to kick I noticed a few problems:

1. The microcontroller code was insufficient to count the 4 million or so bubbles coming off the water lock and an unsigned 16bit int was not going to do. The counter rolled over - the microcontroller got hung and manual resets put gaps in the data.

2. The increase in fermenter temperature I expected never materialized. The temp data was not all that interesting in that it varied only 1 degree C between the barrel and the board. I concluded that the difference could be accounted for by my drafty kitchen and not some cold fusion/fermentation reaction.

3. I can make drinkable beer. So much so that I've cranked out 3 runs since without taking time to test the corrections to the counter code (see below) and finish the php pages so they are fit for public consumption.

Maybe after things slow this summer and before Robothon!