Thursday, April 10, 2014

The perfect coffee maker

TL;DR: This thing makes amazing bitter-less coffee, cost only $150, and was built in one weekday evening.

This contraption is known as an Oji slow drip cold brew coffee system.
You fill the top with ice, and a bit of water.
You fill the center with coarsely ground coffee, and a glass filter disk.

You set it so the top container drips 1 drop of water per second, and the second container so that it drips 1 drop of water per two seconds.
This allows a pool of water to form in the middle beaker, maximizing extraction surface area.

Because you use ice instead of hot water, far less tannins and oils are extracted from the coffee. It also takes 10 times as long to extract the caffeine and coffee flavors out of the ground coffee beans; this is why the brewing process takes roughly 8 hours.

There are lids/filters on each beaker to prevent dust/hidden nasties from contaminating the brew during the 8 hours it takes to brew the coffee.

8 hours may seem like a long time, but if you start it before going to bed, its ready when your ready to drink coffee. Another cool aspect of cold brew is that it doesn't go stale (unless you heat it, at which point it will go stale if you don't drink it).

Because the coffee doesn't go stale you can supposedly age it and develop the flavors even further.

About the build: These setups normally cost at least $500 for a 12 cup system. I pulled this off for $150. I purchased the replacement parts from Yama (the beakers are uniquely suited for the task of brewing up to 12 cups of coffee). I purchased the glass filter (just wash, never needs to be replaced). I already had the carafe. Not pictured are the stopcocks for the top and middle beakers which are just glass valves for precisely adjusting the flow. I bought them on ebay for $20 each. I am still waiting for them to arrive so I rigged some vinyl tape stopcocks for the first brew.

I drew the design in cad, laser cut templates, and cut it out on a scroll saw. I sanded, primed, stained, lacquered, oiled, glued, and mounted in a few hours last night. 

Success metric:
  It needs to be as practical as a normal coffee maker in terms of cleaning, ease of use
  It needs to produce a noticeably better coffee than a coffee press.

The coffee is amazing. I am not a coffee snob. The purpose of building this was for the fun of building it. I expected to not be able to notice any difference in taste compared to other coffee makers being a non-coffee aficionado... I was super wrong. There is almost no bitterness. It is pleasant to drink black (I am a creamer guy). Everyone who has tasted it agrees unanimously it exceeds expectations in terms of taste (my wife was pretty skeptical about this project).

Tuesday, April 1, 2014

Exoskeleton for the girl with a broken ankle

My sister broke her ankle rock climbing. The crutches were annoying her and I thought I could come up with an engineering solution to the problem.

The straps are laser cut acrylic. I thermoformed them to the shape of her leg. On the inside of the straps is non-slip foam padding, and on the outside is industrial strength velcro.

The hinges have bearings in them. The general idea is you slip your leg in, wrap velcro around the straps, and your ankle hovers a few inches above the ground. A shoe riser goes on the other foot, and you walk with a cane.

The device worked as intended, but proved to be cumbersome to wear for long periods of time. It is also still easier to get around on crutches. The upside is, it did allow her to walk for short periods of time without crutches. This was version Mk I. We did three iterations before moving on to other projects

Thursday, March 27, 2014

Highly detailed wood inlays

 I had an unscripted maker day recently. The day consisted of making various random things. These are some of the wood inlays we created on my laser cutter. They turned out pretty awesome! We also did a bunch of welding, wood working, etc. But I wanted to show off some of the cool laser cut projects.

Wednesday, September 11, 2013

Open gantry update

We have been working for some time to get approval from Google to release the files for the completed open gantry. It looks like that should be done soon, and that we will be able to release open gantry to the world. Here are some updated photos. The CNC is completely operational and cutting wood and cardboard. This is out 1:4 scale model. The total cost of this version came in at roughly $350. We have done the research and it looks like the full scale 9ft X 5ft version will cost $450 excluding the cost of router bits. The next steps for us is to organize the source files, create a bill of materials, write detailed assembly instructions, and get the word out.

We would love to work with some Alpha testers. We will provide detailed instructions and guidance for cutting and assembling the CNC. Anything and everything we can do to help we will. If you want a large CNC for ~$350-450 this is a great option.

Sunday, June 9, 2013

Open Gantry: Open Source CNC designed specifically for producing wikihouses Part 1

Open Gantry:
Ken Sherrif, Alistair Milne, Tim Hatch, and I have been working on an open source, self replicating cnc designed for producing wikihouses as a 20% project. Our goal is to produce a full size version that a user can feed in sheets of plywood, that will spit out ikea style flat pack houses that slot together and do not require construction experience or power tools to build. We are shooting for building these as kits an order of magnitude cheaper than  the lowest cost competing cnc. is a very cool idea, but it occurred to us, that no one has a CNC with an 8X4 ft cut envelope. Upon researching the cost, the cheapest option out there is the blackfoot at $3200 for a kit. We have done our research, and although the prototypes will cost closer to 400-500, we think that if we build lots of kits, we can get cost down to $300-400. We are sacrificing precision, and lifespan in order to get the cost down to this level. It is designed to be shipped as a kit to a disaster zone along with a couple pallets of plywood, and operate for a few months. If daily operation/precision was ones goal it would be wise for that person to use a higher quality router, and use some screws.

We are going to open source everything. Check out our progress on github
Next week our plan is to install the electronics, and get it moving, and perhaps tackle the z axis.

Tuesday, May 14, 2013

$10 + 10 minutes = lots of fun

This was a cheap, fun, and quick project. I don't think I have ever made anything so quickly, which resulted in so much fun as this. I started this 3D print before I left for work yesterday. It was ready and waiting for me when I got home from work. Two $0.99 3/8 dowels and some surgical tubing was all that was needed to complete the build. 

It has a range of ~50 meters. I used a binary search strategy to find the optimal angle for releasing the shot.

Monday, April 22, 2013

Some success

This weekend I took a loooong drive in search of snow. I had busted my ass every night this past week to get my V2 prototype built so I could test it Saturday. I worked out as many bugs in the system as I could prior to driving up to the snow. I also brought all of my tools with me. I got there, and had to do some work to strengthen one of the paddles which was too fragile. The fix worked great. Upon the first test, I found that under resistance the paddles pushed closer to the board by a half inch, and the paddle's teeth actually became snagged on the rear end of the board preventing them from spinning. The intense torque caused the weakest link in the system to fail... The gear box. One of the gears pretty much shredded. I adjusted the device position so it couldn't snag the rear of the board but the damage had been done. However there was some success. Even with the shredded gear tips, it was able to get traction, and move a rider on the board. The biggest unknown in the design was whether I could keep traction. The rest can now be engineered to perfection. I have determined that I need a better way to test the device. If each iteration takes 12 hours to test, I will never finish this thing. I am looking into buying/renting a wood chipper. I figure I can put many bags of ice through the wood chipper to create a snow like consistency in my back yard. This should allow me to rapidly iterate at home where I can print new parts, test them, and work out the bugs quickly.
Some other lessons were learned. A spring (not pictured) works great to keep the device retracted. The board is easy to carry around like a normal snowboard. The device does not get in the way in any way, shape, or form. Once deployed it stays deployed. If traction is a problem, the pull strap works great to dig the teeth deeper into the snow. The strap is not a good way of deploying the paddles into the snow. Manually pushing them down works much better. Perhaps with a better design the strap could be more effective. The weakest link is in fact the gearbox. My plan is to make a mold, and cast gears twice as thick (1" thick) in a mixture of the hardest resin I can buy, mixed with small fibers. I think it would also be wise to bring a spare of each part. Another lesson learned is that the mounting system where you remove the binding, and slide the plate under the binding, and strap it down (as pictured) is not a good system. The firmer you keep everything mounted to the board the better. Any give in the system has negative results. I ended up just epoxying the device to the board, and was very happy with the results. Although it will not be easily removed. Perhaps I can epoxy some sort of mount, which the device latches onto instead. V3 will be very similar to V2 with a few tweaks to make everything more practical. I also think some work is needed to achieve my stretch goal of being able to drive this thing up a 20 degree slope.