Sat 23 Apr 2011
CEB Machine Controller: Fabrication Recursion
Posted by Marcin
We have built the complete automatic controller for The Liberator Open Source Compressed Earth Brick (CEB) press a year ago, and have demonsttrated up to 16 brick per minute brick pressing rates. Now we move on to technological recursion of the open source CEB controller technology – by circuit milling our own control circuits with an open source CNC Circuit Mill instead of buying them off-the-shelf.
The CEB controller circuit is already open source, as it builds on the RepRap and Arduino projects. It is useful to be able to mill these circuit boards ourselves, in case we can’t get our hands on them, or if we want more flexibility. In this particular case, the desirable flexibility would be combining the 2 solenoid driver boards that we currently use in the controller into 1, for a small improvement in the fabrication of the controller box.
It should be relatively easy to build on existing work to come up with a circuit mill suitable for milling our own circuits. Short of developing a CNC circuit mill ourselves, we have identified these available options. There are many others.
We need your help. First, help us identify the best solution for an open source CNC circuit mill, and then help us source the parts so we can build it (we can’t bootstrap ourselves without another CNC mill or router). We’d like to go straight from self-milled boards, as then we are in full control of the technology and we are forced to understand it at a deeper level. SnapLock, perhaps modified for lower cost, is our best choice now, since we have a collaborator who can help us bootstrap this route. Our RepRap in the making is another possibility, but that seems to be a more complicated route. I personally like the super-strapped $100 option of the Mantis 9 machine – with completely bootstrapped electronics and router spindle - but I don’t like the wood frame.
Second, help us generate the toolpath files that can be used with LinuxCNC to mill our own CEB controller board. This means the 2 solenoid driver boards, on one circuit board. And, while we are at it, it would be useful to mill our clone of Arduino. We can then populate these components.There is some cost savings in this, but the main point is to gain mastery of the technology as producers, not mere consumers.
By the way – grand news. The first ever independent replication of our open source hardware is happening. Someone just built a replica of the CEB press controller from the documentation that we have on the wiki and blog. This is history. This was a totally independent effort, not instigated by me:
No related posts.
10 Responses to “ CEB Machine Controller: Fabrication Recursion ”
Comments:
Leave a Reply
Trackbacks & Pingbacks:
-
Pingback from CEB Press: Generation 3 Electronics | Factor E Farm Blog
December 16th, 2011 at 8:39 pm[...] breakout board. Generation 3 is now under evaluation at Factor e Farm and at Creation Flame. Generation 1 electronics included 2 solenoid driver boards borrowed from the RepRap project, and are robust but not [...]
April 23rd, 2011 at 4:55 pm
I’m a little surprised that you’re not choosing to etch PCB’s over milling.
With etching you can do large batches of circuit boards all in one go and I think all the chemicals involved are reusable, not sure on that.
Why have you chosen this route?
April 23rd, 2011 at 10:25 pm
Circuit milling can be done on a multipurpose device that is also used for 3D printing, laser scanning, laser cutting, and other functions. This is consistent with our generalized tool set product ecologies. We foresee custom production of boards, not mass fabrication, as the preferred route in the next economy. This sounds like inefficiency, but as we shift away from mass culture into customization, this can make sense. Etching is sound for larger volumes, but one-off production appears to be easier on the circuit mill.
April 24th, 2011 at 3:59 pm
You don’t like the wooden frame?
In the Canaries a friend of mine is building what he calls the CachaRap – translation some other day, but it’s a dirt cheap machine, with metal pieces built from the metal pieces that let the drawers move inside the piece of furniture. For less than 100 euros, I think. It’s still in developement but it looks easy to build for schools etc.
More details soon.
April 24th, 2011 at 6:20 pm
Wood does not last as long as metal. I think this may fail the lifetime design criterion, so the question becomes how much energy is spent on replacing the frame over time.
On cheap machines – the question is replicability and performance.
April 25th, 2011 at 12:24 pm
I recommend getting in touch with Mikey Sklar of Holy Scrap Hot Springs.
He’s a great guy, based on the few interactions I’ve had with him, and he’s blogged a lot recently about circuit board milling.
Separately, I think his temperature controller circuitry would also be worth adopting: it will plug in nicely to your egg incubator, and might also be useful if you’d like a solar-powered refrigerator for food (adapted from a chest freezer; supremely energy-efficient), or are interested in adding fermentation or sous-vide cooking capability to your agricultural efforts.
April 25th, 2011 at 7:20 pm
Yes, we’re in contact, and recently Tim Corrigan joined us on the power inverter side of the Universal Power Supply.
April 26th, 2011 at 5:12 am
Great news on the replication. Congratulations!
May 19th, 2011 at 7:08 pm
On the choice between etching or milling, I believe you have it backwards.
As someone who actually does “one offs” in my basement or bathroom, or on a friend’s porch, or in the garage, etc., I can vouch for the effectiveness of etching, and with the setup costs and time, it will be much more economical to etch a single or dual-sided board than it will be to do that same board on a mill.
That is not to say that one wouldn’t have reason or need for both, but for exactly 1 of, it is no contest, etching is faster & cheaper. If you are using the laser toner method, it will even be easier & faster to etch for quantities of upwards of 10 or so, and only when you can get full automation and need hundreds or thousands, do you start to gain from the benefits that milling can provide.
(By then, I’ve generally sent the pattern off to a circuit board fab house for production)
Having said that, one will still likely need both, and also need plating capability for any real advance circuit board production, both because higher current portions of boards will need to be “plated up” from the standard 1 oz copper layer to something thicker, but also any vias that needed to be plated through will also require such too. This can only be done in a combined mechanical and chemical fabrication facility, where one will need the use of precision milling & drilling, and also etching & plating.
One will also require these functions to really go beyond just a 2-layer board too. Typical computer motherboards are 4-layer boards, with some of the higher-end or more complex designs going even higher, such as 6 or more layers.
It is good to be able to master milling a board, but that is only a beginning to a full-function electronics fab. Don’t give up on the chemistry part, though
May 19th, 2011 at 7:23 pm
By the way, the “Laser toner” method I referred to, is basically taking your circuit pattern you wish to make, reversing it to make a mirror image, and printing it out on normal paper on a laser printer, but making sure that your pattern is 100% black.
Modern laser toner is primarily plastic, and with such a print out, one can take that and a standard clothes iron, and “iron on” your pattern to your blank board as an “iron on transfer”. Use a good hot iron, and plenty of patience – it may take some practice to get the hang of it, but you really do want the paper to be stuck to the copper blank and good by the time you’re done ironing.
Let it cool, and then soak the stack in water to soften up the paper. You can peel off and/or tear off the paper, and you’re left with your pattern on your blank, in black. Toss it into your etch tank, and swish it around for a few minutes, rinse, and drill it as needed.
Easy to setup, repeatable all over the world, and the patterns can even be mailed to places where having a laser printer might not be a convenient choice. From start to finish, it is faster & cheaper than any milling setup can be, and the drilling parts can even be done with a hand drill and a steady hand, too.
The most environmentally friendly set of chemicals is reusable, and only needs a refresh of hydrogen peroxide on occasion for quite a long time.