Gettin’ gooey

It is that delightful time of year where the pace slows a little and I get some creative time to dream and build in the brewery! And while summer has well and truly hit with a heat wave in South East Australia (a week of 40C+ temperatures for Melbourne) I’ve been in front of my computer rather than in the brew shed. Those who have followed this blog for a while will know that I’m keen on simple, efficient, intuitive user interfaces. I’d rather a few extra discreet buttons that make sense of function rather than combining functions in complex ways. I’m not a fan of needing to press two buttons to stop a pump, for example. While in some ways a rotary encoder is a simple and powerful user input device, with the BrewTroller it gets clunky when used to input a brew name. To be fair it is simple and cheap, but not quite what I’m after for a user interface. Most commercial user input devices (ie. buttons and LCDs) are built to a price which makes sense when you produce thousands or millions of them and every cent saved makes a difference to the bottom line. I think it is hard to justify the same level of thinking to a home built brewing machine. To put things in perspective, I brew to make the best beer that I can and the cost of a few grams of grains or extra hops is not relevant to that end. I estimate that I spend around $1000 on ingredients each year for brewing so that helps frame costs of a new machine build.

A brief controller history

I’ve gone through a few iterations of control device over the last 14 years. The first model HERMAN used an old laptop PC with a parallel (printer) interface board. It was pretty much state of the art in its day but all the computing intelligence was in a Visual Basic 3 program. The interface was just a dumb peripheral so if Windows crashed, so did everything. Windows 95 had a wonderful ‘blue screen of death’ that prompted me to look for something a little more reliable and less cumbersome. Still I was proud of what I achieved with this technology. It was revolutionary for its day …

From there the HERMAN control system was rebuilt around a picaxe microcontroller. Micro controllers are the logical choice for simple process control like brewing machines. In a sense, what we do is not that far removed from controlling a bread maker, a dishwasher or a washing machine. Micros, LCDs and a few buttons make sense. It is lightweight, there is no ‘blue screen of death’, and micros reboot in seconds rather than minutes. I went with a PICAXE 40X as it was the most powerful in that range at the time. It was easy to program but that came at the expense of flexibility, especially in multi-tasking and timing in general. It was good, and it certainly did its job, but I always felt I was pushing against its limitations.

It was about that time that I discovered the growing Arduino movement. It had many advantages over the PICAXE – many real life examples, a big community of users, and far more grunt than the PICAXE range. To be fair PICAXE was focussed on education, and it was excellent for its market. I’d still recommend using their processors to build intelligent and cheap peripherals like a hop dropper, for example.

It wasn’t long before I found the BrewTroller project. Here was a fledgling Arduino based open source group of enthusiasts who had something up and running and saved me hours of coding. It seemed like the perfect solution for its time. I own two BrewTroller boards and a number of peripherals and it is BrewTroller that currently controls my original BrauBushka machine. Despite my annoyances with the system, none have been bad enough to get in the way of the usability of this device. It remains a great brewing controller. For anyone who is not a hard core experimenter, I’d still be happy to recommend it. It works, and works well. It is not designed for single vessel brewing and yet a few simple code modifications out of the box and it does a better job than anything I’ve yet encountered. If I were not the ‘restless, wanting to build the best possible thing’ type, I would stick with BrewTroller … … except for perhaps Matho’s BrauMeister clone controller. This is specifically one for single vessel brewing and therefore has a simple elegance that the BrewTroller lacks. Matho has done very well to mimic the BrauMeister control system. My only reservation is that I would build a control system differently to Spiedel. To summarise the last two options – BrewTroller and Matho’s controller: If you do not want to build or code from scratch, either of these are good solutions. The BrewTroller is more complex but will handle different configurations; the Matho controller is specific to single vessel and is excellent for that. But … the BrewTroller is much too complex in my opinion, and it’s user interface is nasty. The Matho controller is better but as a reverse engineered BrauMeister control system suffers from the ‘save money at the expense of the interface’ issue. So, my idea for the next system …

Gooey Pie

I’ve always wanted a user interface to be easy to use. Brewing parameters (mash times and temperatures, boil time, times for hops etc.) can be cumbersome to input or set up. I’d ideally like a machine to have quick and easy ways to get a brew started. One way to do this may be to build a recipe in a program and have a controller read parameters automatically. While a reasonable idea, there are many different programs and then the issue of how to get that program as input. My direction involves an easy, intuitive user interface that is a breeze to use. Simple and powerful is the aim, rather than cheap and complicated. This search has led me to a touch display from 4D systems. Pretty much every technophile these days uses a touch screen on a smart phone or similar device. The screen responds to whatever type of input we need. If we need buttons, buttons are on screen. If we need pretty colours, they are there to help the eye figure out what is going on. If we need numerical input, a numerical keyboard pops up. Same also for QWERTY input. The beauty of 4D screens is that they are intelligent as a standalone screen, and easy to program via free software. User input can be put on screen quickly and easily, and parameters are then made available by a host processor to either read or control. In other words, the hard/grunt work of a graphical user interface (GUI or gooey) is done on the screen, leaving the controller to do simple stuff like timing and maintaining temperatures. It means that complex brewing parameters can be set quickly and easily, and the machine can be up and running in no time.

In my case I’ve been experimenting with a 3.2″ PTU screen. At around $100 AU, it is not a deal breaker, especially considering that is roughly my ingredient cost for a month. Below is a screen shot of my first experiment in this field. It doesn’t yet do anything, but coding is simple from here and it promises a lot! The other part is welcoming a new controller. The Raspberry Pi is gaining popularity and it seems about time to experiment with one. It does some things better than Arduino, and some things not quite so well. One feature that has caught my eye is that Model B is natively equipped to connect to the world of things. I’m far from there, but it means that in theory the nice GUI can be replicated on a web page and control can be either local or distant via smart phone or any web connected device. As usual, I have so much to learn, but I intend having fun doing so 🙂 Happy new year, may you have success in whatever brewing ventures you have in 2014.