Monthly Archives: February 2011

Hop droppers

via Brewbot Mk2

I saw that tiny chinese servos can be had for $2 on ebay so bought a few with plans to use them for the hop droppers.

The plan was to have some little cups on hinges driven by the servos.

Unfortunately I couldn't find any small stainless steel cups locally, and am too short on time to wait for a delivery, so decided to go with some PVC pipe for now.





New pot and bag

via Brewbot Mk2

The pot I have been using is a cheap Chinese aluminium one, which is pretty thin. It's actually good for prototyping. Very easy to drill and work with.

But the downside is that it dents and bends and corrodes. And it's not easy to weld/solder to.

So time to upgrade to the real thing, stainless steel. More stronger and I can braze and silver solder my fittings to it.

I wanted pretty much everything in contact with the wort to be stainless steel. So I got some 316 stainless steel mesh to make the bag with.


The 1" socket for the heating element:



The socket for the temp probe:

This was my first try at silver soldering. It wasn't that hard, but the results aren't perfect. Much better than leaky weld-less fittings however. 

I also spent the money on good brand solder and flux than comes with a MSDS so I knew there were no nasties in the solder or flux.

The 3-piece valve body soldered directly to the kettle: 

The inside with all the fittings in place.



Next on to the bag. First I had a 12" square sheet of thin gauge 316 stainless. Decided to make a 1.5" wide hoop to go around the top edge. So cut and bend.


Next the stainless mesh was cut into a circular bottom piece 10.5" in diameter and some walls 8" high.


 I stitched the bag together with 316 stainless steel wire. Surprisingly easy.




Then I began work on the stirrer.
 Here it is all together.


Performance (MilkyMist)

via OSHUG

It stands to reason that hardware which is open to being studied, modified and improved would be well suited to performance environments, and just as F/OSS has proved popular in support of creative practices so is OSHW similarly gaining favour. With designs ranging from simple electronic instruments that make for an ideal first electronics project to vastly more complex processing and synthesis devices.

At the eighth OSHUG meeting we'll be hearing about the Milkymist™ project which "develops a comprehensive open source solution for the live synthesis of interactive visual effects for VJs (video performance artists)".

MilkyMist - An FPGA-based open-hardware video synthesis platform

The MilkyMist project develops a stand-alone device in a small form factor that is capable of rendering MilkDrop-esque visuals effects in real time, with a high level of interaction with many sensors and using live audio and video streams as a base. The flexibility of the FPGA used as a central component enables advanced users to modify the design, and also permits compact integration of many interfaces (Ethernet, OSC, MIDI, DMX512, video inputs, GPIO, VGA output, USB, Irda ...), making Milkymist™ a platform of choice for the mobile VJ. But Milkymist™ is more than a visual synthesizer - it is also one of the leading open source system-on-chip designs. It is today the fastest open source system-on-chip capable of running Linux, and it comes with an extensive set of features and graphics accelerators. The IP cores that make up the system-on-chip are entirely written in open source synthesizable Verilog HDL and come with test benches and documentation, which makes Milkymist™ a great library of re-usable logic cores to serve as a base for other open source hardware.

Yann Sionneau is a twenty two year old Frenchman and soon to be graduated from the telecommunication and networking engineering school Télécom SudParis. His current interests in the main are low level software development, FPGA design, embedded systems and networks. He read his first C language book when he was 12 and fell in love with the language. He met Sebastien Bourdeauducq (aka lekernel), leader of the Milkymist project, in 2008 while doing whilst an intern at a startup co-founded by Sebastien. He ported RTEMS to Milkymist as part of the Google Summer of Code 2010 program and has been following the project for some time.

Note: Please aim to arrive for 18:00 - 18:15 as the event will start at 18:30 prompt.

Performance (MilkyMist)

via OSHUG

It stands to reason that hardware which is open to being studied, modified and improved would be well suited to performance environments, and just as F/OSS has proved popular in support of creative practices so is OSHW similarly gaining favour. With designs ranging from simple electronic instruments that make for an ideal first electronics project to vastly more complex processing and synthesis devices.

At the eighth OSHUG meeting we'll be hearing about the Milkymist™ project which "develops a comprehensive open source solution for the live synthesis of interactive visual effects for VJs (video performance artists)".

MilkyMist - An FPGA-based open-hardware video synthesis platform

The MilkyMist project develops a stand-alone device in a small form factor that is capable of rendering MilkDrop-esque visuals effects in real time, with a high level of interaction with many sensors and using live audio and video streams as a base. The flexibility of the FPGA used as a central component enables advanced users to modify the design, and also permits compact integration of many interfaces (Ethernet, OSC, MIDI, DMX512, video inputs, GPIO, VGA output, USB, Irda ...), making Milkymist™ a platform of choice for the mobile VJ. But Milkymist™ is more than a visual synthesizer - it is also one of the leading open source system-on-chip designs. It is today the fastest open source system-on-chip capable of running Linux, and it comes with an extensive set of features and graphics accelerators. The IP cores that make up the system-on-chip are entirely written in open source synthesizable Verilog HDL and come with test benches and documentation, which makes Milkymist™ a great library of re-usable logic cores to serve as a base for other open source hardware.

Yann Sionneau is a twenty two year old Frenchman and soon to be graduated from the telecommunication and networking engineering school Télécom SudParis. His current interests in the main are low level software development, FPGA design, embedded systems and networks. He read his first C language book when he was 12 and fell in love with the language. He met Sebastien Bourdeauducq (aka lekernel), leader of the Milkymist project, in 2008 while doing whilst an intern at a startup co-founded by Sebastien. He ported RTEMS to Milkymist as part of the Google Summer of Code 2010 program and has been following the project for some time.

Note: Please aim to arrive for 18:00 - 18:15 as the event will start at 18:30 prompt.

Maiden test brew

via Brewbot Mk2

Pretty early on in the construction I wired up the heating element, Solid State Relay and got the MCU reading temperatures from the DS1820.

So it would already have been able to do a mash and boil with manual intervention at various points.

Kelly has been wanting to test a Hoppy Red Ale recipe in a small batch for a while and I have been telling her to avoid doing it on the stove (temperature is hard to regulate there by hand). The brewbot can do the temperature part.

Well we finally did tonight.

It was good to get a sense of how the system dynamics are and if the basics are working.

I set up FreeRTOS on the RDK with some simple tasks to monitor temperature and the push buttons.

Up/down would control the temperature set point, and after some experimentation, left/right buttons would control the heating duty cycle. On the LCD was display both those parameters as well as the current temperature in the thermowell.

We did a test boil in the pot the night before and the boil was very vigorous, splashing water out of the pot, so I decided to add the duty cycle adjustment feature before doing a mash.

Kelly had a bag that fit the pot perfectly. Keep in mind that the true bag will be stainless steel mesh and quite a bit smaller.

The test recipe called for a OG of 1.066, so a relatively big beer. We were aiming for a final volume into the fermenter of about 2 gallons. This should nicely validate the volumes and sizing of the pot.

The first step was to setup in the laundry near the 240V outlet.


The laptop is there for final code tweaks and eventually won't be needed.

Then dough-in and some experiments.

Without constant stirring it seems a hot pocket would form and rapidly overshoot but the overall mash temp would slowly drop.

Constant stirring seemed to solve this.

Luckily constant stirring is part of the final design, just not yet implemented.


We also discovered that for the mash I needed to turn the duty cycle right down on the heating. Down near 10% to prevent too much overshoot.

I may also implement the PID control algorithm to help with this.

The boil was run at 50% duty cycle. So about 1750W. This means I probably didn't need to use 240V power, and could probably run this off an ordinary 120V outlet.

The SSR and it's heatsink didn't get warm at all, so I think we are well within spec there (I have heard that these need to be massively overrated and can explode quite violently in some conditions).

The boil causes this rocking of the wort to form.

Post boil, whole hops floating:

 No-chill into a 10 litre sankey keg I got off ebay:
 We did have some problems with hops clogging the tap. Will have to come up with a more permanent solution that the strainer Kelly held over the outlet.
 Some minor caramelisation on the element.

But overall seems to be a successful first test batch.

Now on to build the bag and stirrer. Build the hop dropper. Write a ton more code. And write it all up and submit my contest entry.

Mounting the MCU, electronics and probes

via Brewbot Mk2

I've been a bit slack on the blogging.

Recent progress includes:

1) A housing for the RX62N including acrylic splash guard and push buttons.


The idea is to have the LCD at an angle where it is visible and to avoid the board getting wet should there be any splashes.

The cover is hinged so that I can get easy access to the board for programming.

2) Electronics re-work.

As seen in the other posts I had the electronics board mounting directly on top of the RDK board via some DIP pins. I decided to redo that with cables instead to give me more mounting options.


I didn't have enough IDC connectors to put another set on the electronics board side of the cable, so had to solder the cable to the board.

The cable I had on hand had very thing conductors and not many of them. The cable I was using before was much better but had already fatigued and broken off some wires at the board. These thing conductors were sure not to last very long. So I decided the quickest solution was to add some mechanical support. See the piece of clear acrylic along the left hand edge of the board. This clamps the plastic part of the cable to the board and prevents movement (and fatigue) of the conductors as they enter the board.


In the close up you can also see a heat sink on the right hand edge.  This is for the transistor that controls the mash motor. This motor can draw up to 5 amps at stall, so the transistor can generate a little heat.

The other free standing TO-220 transistor is for the mains water solenoid. This only draws about 300mA so no heatsink required.



The orientation of the board was chosen so that I could get to the back side without removing it from it's mounting. This should facilitate easy mods or additions.

 3) Level probes in the pot.

Conductive level probes will be used to determine if there is enough liquid to safely power the heating element, and to determine how much water to initially fill the pot with.

The other conductive surface is the pot itself. Therefore the probes need to be insulated from the pot. To do this I used a cheap plastic chopping board.

I cut the corner off and mounted inside the pot.



The problem themselves are 1/8" 316 Stainless steel rod.

 Two screws come in from the outside and hold the rods in place.



4) Mains water solenoid. To be used to fill the pot with cold water at the start of the brewing process.

Decided to mount at the back, low and away from nearly everything. Hopefully this will prevent anything interesting getting splashed or sprayed should a leak develop.


Even tho there is a clamping diode across the solenoid driving transistor on the electronics board, I also included two clamping diodes right on the solenoid. I wanted to keep any EMF and the currents generated away from the MCU.

5) Shroud for the temperature sensor in the pot.

First up was to connect a DS1820 to some cable so that it would slide down inside the thermowell.
 
The a housing around the outside of the thermowell.

6) Lots of wiring. Not interesting enough to have a photo, but pretty much all the electronics and motors are now wired.

7) Adding a tap to the pot.

8) Adding limit switches to the Y-axis part of the crane. No pics so far.