Ian Fab

Once the board is milled and the components are picked out,...

Wed, 07 Mar 2012 02:16:42 -0500

Once the board is milled and the components are picked out, it’s a fairly straightforward soldering job. In the final picture you can see both of the completed boards side by side: fabISP and HelloEcho.

Once the design is finalized, you mill it out in the same way as...

Wed, 07 Mar 2012 02:13:03 -0500

Once the design is finalized, you mill it out in the same way as the fabISP was done, and then it’s just a matter of soldering all the components in place. This time around it was a bit trickier since I couldn’t go the whole “match colors and shapes” route since there was no example board or “correct” schematic to copy, which meant I had to riffle through these folders full of components and kind-of-sort-of figure out how to read the labels.

The first step in the project, besides downloading and...

Wed, 07 Mar 2012 00:36:41 -0500

The first step in the project, besides downloading and installing software, was to download the schematic for the Hello Echo board and import it into the EAGLE software.

EAGLE, which stands for Easily Applicable Graphical Layout Editor, is software which allows you to download libraries of manufactured electronics parts and arrange them in a schematic like pieces of a puzzle. The program will connect all the components to one another according to your specifications, and then with a little manipulation it can generate the layout of the milled out connections.

Basically you arrange all of your components on a virtual board, then establish the connections between them, and because the program already has the specifications of all the parts you’re using, it will generate the correct shapes in the correct sizes to define mounting points for all of the components on the board. Then you move the mounting points around to your liking, and draw all of the connections between them. At this point you can generate an image file which can be output to a CNC milling machine to carve out the copper plates which make your board.

Week 7 - Electronics Design Our assignment this week was a...

Wed, 07 Mar 2012 00:25:24 -0500

Week 7 - Electronics Design

Our assignment this week was a continuation of the processes we covered in week 5 when we milled and “stuffed” our own circuit boards to build a micro controller called the fabISP.

This time around our assignment required us to actually design the circuit pattern to be milled out rather than using a pre-existing design downloaded from the web.

Well, kind of. Actually we did something halfway in between. We were given a pre-existing circuit design and we were tasked with modifying the design to include an LED and a button. The name of the board we were starting from is the “Hello Echo” board, and my understanding is that we’ll be using it in future assignments alongside the fabISP.

At the moment, though, I have no idea what it does or how it is supposed to work.

All the images I took of this project are from the assembly (or “stuffing”) phase, but the majority of the steps are done on the computer using a design program called Eagle.

More on that in the next post. Here is a video of my workspace, including the milled board before stuffing.

Fab Academy Week 5 - Electronics Production Troubleshooting,...

Mon, 20 Feb 2012 18:32:06 -0500

Fab Academy Week 5 - Electronics Production

Troubleshooting, de-soldering, following instructions

After scanning back through the instructions I found the entry about the solder bridges. I set up my soldering station a second time, and bridged the two areas specified on the diagram. Then I connected the FabISP back to my computer and ran the steps again. It worked this time.

The final step, once the FabISP had been programmed, was to remove the solder bridges. Unfortunately this is not something that I got to see a demo of in class so I pretty much had to guess how to do it. I had overheard other students talking about using a woven rope of fine copper wire to “wick” the solder off of the board, but had not seen it demonstrated. I should mention at this point that I was working on this project fairly late at night when no-one else was in the lab, so there was nobody available to ask about this.

After some fudging around with the copper braid and the soldering iron, I managed to remove a little bit of the solder. It didn’t look pretty, and probably a lot more of the solder could have been removed, but it was enough to eliminate the bridge so that the two areas of copper were no longer connected. Good enough for me.

I was initially stressed out about this project because it seemed very technical and involved, plus this week brought with it many outside obligations and deadlines such that I did not expect I would have time to finish the FabISP. However once I realized that it was not necessary to actually *understand* any of the concepts or processes I was using, it became a simple matter of following basic steps.

Fab Academy Week 5 - Electronics Production Installing...

Mon, 20 Feb 2012 18:32:01 -0500

the fabISP plugged into my computer

I actually repositioned it

so it would look like the web photo

Fab Academy Week 5 - Electronics Production

Installing Firmware

The next few steps involved installing and running specific software on my computer from the command line. Fortunately the instructions worked exactly as they were supposed to and there were no hiccups.

To install the firmware, I had to connect a second device to my computer, which would in turn be connected to the Fab ISP. The device I used was a black box with a 6 pin connector called a “USBtiny programmer”.

From there, it was just another series of steps involving typing various codes into the command line on my computer. This was supposed to allow a program called “make” to install firmware on the FabISP. Everything worked fine up unitl the last step, when I received an error message.

Well, it turns out my matching of colors and shapes didn’t serve me all the way through this project after all. Remember earlier when I mentioned the solder bridges? They are used to load the firmware, and then once the firmware is loaded, they are removed from the boards Well, since all of the example boards are *finished* boards, they of course do not have the solder bridges on them anymore, which meant that when I replicated them in the apelike fashion that is my wont, the board I built did not have solder bridges. Which in turn meant that my board in its initial form could not be loaded with the firmware, which in its own turn lead to the error message I received.

So: FINE. I’ll READ.

Fab Academy Week 5 - Electronics Production Soldering The second...

Mon, 20 Feb 2012 18:31:57 -0500

Fab Academy Week 5 - Electronics Production

Soldering

The second step involves soldering all the components onto the board. This is called “stuffing” the board, and the placement and selection of components is determined by the schematic available on the Fab Academy website. There are transistors, capacitors, resistors, diodes, a crystal, an actual chip called “attiny44a”, and a mini USB header. All of these pieces are soldered onto the board using tin solder. In addition to the components, there are two solder “bridges” which temporarily link different parts of the boards to each other. These bridges are used to load the initial firmware onto the board, and then must be removed afterwards in a second soldering operation. More on this later. Maybe.

The actual soldering method is pretty easy provided you prep your workspace properly. One thing that helped me was to raise the work surface closer to eye level by setting a small footstool on top of the table and using it as my work surface. The soldering is done with a tin-based solder and an electrically hearted soldering Iron, set to 700 degrees Fahrenheit. The method that worked best for me was to dip the end of my solder wire into the flux gel so that there’s just a tiny bit on the tip, then touch the heated element of the iron to the area of copper I wanted to solder to heat it up. While the copper is heating up, I tentatively tap the end of the solder wire onto the surface so that some of the flux flows off of the solder wire and onto the copper surface. As the copper heats up, the flux liquifies and spreads across the copper. When the copper is at just about the right temperature, the flux will start to smoke a tiny tiny bit, which is when I touch the tip of the solder wire to the surface a second time and if I’m lucky a tiny bit of it will flow onto the copper. If it doesn’t flow, I can touch it directly to the tip of the soldering iron, and it will usually bind partially to the copper and partially to the soldering iron. I then remove the soldering iron and use a damp sponge to wipe off any excess solder still clinging to the tip.

Once this initial dab of solder is in place, a component (such as a resistor, capacitor, etc) can be laid down on top of it and held in place with tweezers while the soldering iron is placed onto the connection point which raises the temperature of the solder and causes it to flow out from under the component and bind the component to the copper.

I don’t know anything about electronics, so this whole process quickly turned into a game of matching colors and shapes. I looked at the sample boards and the images on the website, and then rummaged through our local Fab Lab’s box of components until I found pieces that looked like the ones in the example. Then I soldered them into the corresponding spots on my own board. Once everything was in place, I plugged the Fab ISP into my computer’s USB port. The fact that no smoke issued from the device, and no error messages appeared on my computer, were indication enough that I was at least somewhat on the right track.

Here is the board after milling

Mon, 20 Feb 2012 18:31:52 -0500

Here is the board after milling

Fab Academy Week 5 - Electronics Production Part 1 - CNC...

Mon, 20 Feb 2012 18:31:46 -0500

Fab Academy Week 5 - Electronics Production

Part 1 - CNC Milling

The first step to making this board was to mill out the actual circuit pattern using a CNC milling machine. The material we used consisted of a layer of copper laminated onto a layer of nonconductive material (plastic I think). The CNC milling machine is a small rotary bit toolhead which can move on the x, y, and z axis. The unmilled laminate board is mounted in the bed of the machine using double-sided tape. To make the circuit pattern, you download the design from the Fab Academy website, load it into the machine, and run the machine. The small spinning mill bit cuts through the copper layer and into the inert material below, carving channels in the board and eventually leaving only a series of interconnected lines of copper on the surface. These lines of copper will conduct current between the various elements on the board.

I’ve sped up the video of CNC milling and also applied the “Sci-Fi” filter in iMovie to make it seem more futuristic.

Fab Academy Week 5 - Electronics Production This week we had a...

Mon, 20 Feb 2012 18:31:31 -0500

Fab Academy Week 5 - Electronics Production

This week we had a very specific assignment with very specific steps to follow. To be perfectly honest, I don’t actually understand what we made or how it works, but I suspect it will play into future lessons somehow. We built a device called the Fab ISP Programmer, which to my untrained eyes and brain is a circuit board with a mini USB port on it.

The wooden hinge in action

Wed, 15 Feb 2012 09:25:36 -0500

The wooden hinge in action

Press-Fit Construction, 3.5th attempt After a full day of work...

Wed, 15 Feb 2012 09:25:32 -0500

Press-Fit Construction, 3.5th attempt

After a full day of work at my secret nondescript office job where I manipulate files in an actual filing cabinet while listening to audiobooks, I decided I would ride the rush-hour train from Boston back to Providence for one last crack at the press-fit assembly project.

Upon arriving at AS220, however, I was dismayed to discover that the laser-cutter was already occupied, and would remain so until 9pm. The final train back to Boston leaves at 10pm, so I was faced with the prospect of sleeping in the lab, which I wasn’t even sure I’d be allowed to do since I had not been given my key yet and would have no way to lock up after everyone left. The person operating the laser cutter indicated that she might finish earlier than 9pm, though, so I went out for a coffee and pondered my options.

Upon arriving back at the lab, I decided I would at least stay until 9, just in case the cutter freed up at an earlier time. As an experiment/last-ditch act of desperation–and as a way to keep myself occupied whilst waiting for a turn with the laser–I pulled out Press-Fit Construction Attempt #3 (which didn’t cut all the way through), and decided I would just cut the damned thing out by hand.

I laid the half-cut pattern on top of a cutting mat, and using an Xacto knife and a ruler, I went over all the lines of the design multiple times until the blade passed through to the other side of the wood. I had to be extra careful on the “joints” so as to not split the wood with the pressure of the blade.

After about an hour of handwork, The components could be popped out of the wood panel. With trepidation I gingerly gave the joints a tentative bend. They seemed stabile enough to bend the full 90 degrees, so I went ahead and attempted to assemble all the parts. Miraculously, they fit together pretty well, and I was able to get the whole box together without breaking everything. Even the glass panel fit handsomely in its frame.

I’m calling this project “done”.

There are a *lot* of things I’d like to modify about this box. I’d like to figure out a way to cut everything cleanly on one pass. I’d like to modify the back a little bit so that I can integrate a wall-hanger into it. I’d like to add an extra layer to the front so that the glass is held in a more stabile way. I’d like to paint the whole box, or better yet try assembling it out of acrylic. I think it would look neat in white acrylic. It would also be nice to modify the spacing and placement of the tabs to perhaps create a nice visual pattern at the front of the box. The tabs could fit together better as well and need some tweaking to better accommodate the thickness of the material.

Also, if this is going to be a real display case, there needs to be a good way to open and close it. Probably a final version would have to have all the pieces glued together, so there’d need to be some kind of access panel or something. Then again, maybe not. How cool would it be to have a display case that snaps together and can be fully disassembled, packed flat, and then reassembled again?

I’d like to do a bunch more boxes using this design as a jumping-off point. The boxes don’t need to be square. They could be hexagonal, or even round. This current box is as-yet unsuitable for displaying my sculptures, especially since some of them occupy a more rectangular space.

Press-Fit Construction, 3rd Attempt SO I did a 3rd attempt. For...

Wed, 15 Feb 2012 09:25:28 -0500

Press-Fit Construction, 3rd Attempt

SO I did a 3rd attempt. For this attempt, I duct-taped the corners of the board down, because the panel I was working with this time seemed to have a slight warp to it.

My settings:

speed: 8%

power: 100%

Frequency: 1450Hz

Didn’t cut through. Anywhere. The whole thing was only a partial cut. Which is odd because during the previous attempt, even after the first pass of the laser, *most* of the components were all the way cut out (there were just a few areas hanging on). Not this time, though. I’m chalking it up to inconsistency in the composition of the material. Otherwise I have no idea.

Once this cut is done, I will have to go home. I am out of wood and my train back to Boston leaves soon. I have one more idea to strengthen the joints that I’m cutting, but I doubt there will be time to try it out before the project is due on Wednesday.

Right now, all of the cutting is being done such that the vertical lines which perforate the joints and make them flexible are also running in the same direction as the grain of the wood. This is necessary because all my wood panels have the grain running crosswise rather than lengthwise, and they only fit into the cutting bed one way. If I’m willing to waste an entire panel, though, I could run the wall-strip diagonally across the entire thing, which would probably give it enough structure to withstand the bending even after the excessive charring which is necessary to make the laser cut all the way through.

Or I could just wait until we get a new, *level* bed for the laser cutter. Yeah I’m a poor craftsman. Watch me blame my tools!

But the fact of the matter is that the major benefit of a laser cutter, besides its speed, is that it is supposed to be a precision instrument which can cut materials into very specific shapes with a high degree of accuracy. Right now, it doesn’t do that, and it severely limits the types of things you can do with it. Now I don’t know if a new bed will fix all of this. I’ve been told that the lens is not being maintained as it should be and as such there may be an additional focusing issue to contend with. This is a public shop, after all, and like all public workspaces, maintenance of equipment is severely neglected at best and more often than not it doesn’t happen at all, so at a certain point it’s just going to turn into an exercise in adapting to a sub-optimal set of tools. Still, I hold out hope that things will get at least marginally better soon.

Press-Fit Construction, 2nd Attempt Other people needed to use...

Wed, 15 Feb 2012 08:06:05 -0500

Press-Fit Construction, 2nd Attempt

Other people needed to use the cutter, so I while I waited for it to free up again, I modified my file a bit. I changed the front and back panels to match edge to edge with the outer walls, and I also changed the assembly method to something more like meshing tabs rather than little rectangular projections which go through holes on the panels (This was necessary in order to allow me to have the walls meet up with the front and back panels).

Once my turn at the cutter came around again, I ran the file:

speed: 8

Power: 100%

Frequency: 1425 hz

This time, I had a plan to compensate for the uneven bed. I figured that since the material is higher on one side and lower on another, that I would just run the cutting operation twice, but in between the first and second operations, I would refocus the laser onto the far right side of the material.

Normally before you run the laser cutter, you calibrate it by focusing the beam on the upper-left hand corner of the material. If the beam is too high or too low, it won’t cut all the way through, and it will also cut a much wider line that it ought to. I figured I could run the job once, then refocus the beam on the far right side of the material, then run it a second time. sure, the cuts would be a bit wider than I’d like, but at least this way they ought to cut all the way through.

To save time, the only area I cut twice was the perforated strip which makes up the walls of the box.

Didn’t work. The extra pass with the laser burned too much of the wood and rendered the whole thing so brittle that it snapped as it bent. Three out of four corners were actually OK initially, but after repeated bending back and forth, only one joint (the one closest to the upper-left corner of the material) held up. Additionally, the tabs did not fit together. This was partially due to the extra charring of the wood, and partially due to the fact that I had to shrink the file down at the last minute to get it to fit on the piece of wood I was using. This shrinkage didn’t account for the thickness of the wood and threw the ratios of the tabs off. (I knew this would happen but decided to go ahead anyway because it would take a long time to fix and I mostly just wanted to test the joints anyway).

Back to the drawing board. Which in this case is still Adobe Illustrator. After another hour of tweaking I fixed the tabs so they will hopefully fit together better, and I shrunk the file size down a bit so everything would fit on one 23”x11.5” panel (not to be confused with the 24”x12” capacity of the cutter, nor the labeled specifications of material you buy, which will be sold as 24”x12”, but will not come in those dimensions.) I then fixed the proportions of the tabs to accurately reflect the thickness of the material. I also adjusted the curvature of the corners on the front and back panels to more accurately reflect the way the joints bend.

Our assignment this week was to produce a...

Tue, 14 Feb 2012 16:34:33 -0500

Our assignment this week was to produce a “Press-Fit” construction project using the laser cutter to pierce out parts which fit together without glue or fasteners, using tabs or dovetail-style joints.

A while back, I made a series of sculptures using the same processes I use for Zoa Chimerum Jewelry. They went into shadowboxes which I bought at Ikea and modified. Unfortunately they were water-damaged during an outdoor art fair when a sudden storm came through. I did my best to repair them, and they look OK, but I’ve been wanting to replace the boxes outright so this assignment seemed like as good an excuse as any to make my own display cases.

One of the sites in my RSS feed posted a link to Dutch design house Snijlab, which has discovered a way to make flexible plywood curves using a laser cutter. They’ve been using the technique to make snazzy notebook holders: http://www.snijlab.nl/?page_id=358

I decided that I would like to incorporate these curved corners into my shadow box design.

Fortunately for me, the Oomlaut Lab has already adapted these bendy corners to make an Arduino housing: http://oomlout.co.uk/?p=434, and they were kind enough to post their templates online: http://oomlout.com/ADBB/

I downloaded their template and removed most of the extraneous stuff like the bolt holes and what have you (remember the assignment calls for friction fit only).

I also had to modify the widths and heights of the various tabs to reflect the thickness of the material I was working with. (My plywood is .2” thick). This is important otherwise you won’t get a good fit when you assemble the parts. I did all of this using Adobe Illustrator and it was finicky and annoying. I had to manually click and drag a lot of individual points and I wasn’t able to find an easy way to make the kinds of changes I wanted in any kind of “batched” way. Probably other programs are better suited to this and I’d like to learn how to use some of these programs for projects like this in the future.

After the cut pattern was configured the way I wanted it to be, I sent it to the laser cutter and burned all the lines through the wood.

These were my settings:

speed: 8

Power: 100%

Frequency: 1325 hz

Unfortunately, we still have not replaced the bed in our laser cutter, and as such there is a focusing issue. The laser beam did not cut all the way through to the other side in a lot of places. For the back and front panels, this isn’t such a big deal because you can sometimes still force the pieces out of panel and then trim off any excess with a sharp knife. For the perforated curves, though, this is a HUGE problem, because if the cuts don’t penetrate all the way through, the hinge won’t bend, it will snap.

Which is exactly what happened with all but one of my corners.

Frustrating, but I wasn’t expecting to get it right the first time. With some tweaks to the settings, hopefully I would be able to get the cuts to go all the way through.

Laser Etching on Wood I took some images from the drawing series...

Mon, 06 Feb 2012 13:21:00 -0500

Laser Etching on Wood

I took some images from the drawing series I did for The Sketchbook Project last year, and tried converting the scans into plain black and white. I then imported the images into illustrator and used the “live trace” feature to convert the raster image into a vector drawing. I had to play around with the tolerance settings a bunch before the conversion looked right and didn’t remove too much detail. Then I found some scrap wood leftover from someone else’s laser cutting project, and I set the machine to etch the design onto the surface of the wood.

The results are pretty nice. I’d like to try making a few more of these and maybe putting them up on etsy. They would possibly look even better if I rubbed pigment into the etched lines in the wood, to make them contrast more. Another neat idea would be to paint the wood in a layer of white paint, and then once that coat of paint has dried, paint the wood in a layer of black. Then have the laser etch through the black to reveal the white underneath. I’m not sure how precise you’d have to be with your settings to make this work, though. Perhaps it would work better to use a completely white material under the black paint, such as white acrylic? Better yet - what about using a reflective metal sheet with a layer of paint over the top, and then etching the paint away with the laser? Then you’d have shiny metallic lines glinting through blackness. This would look really nice with polished brass or something like that. Assuming of course that the paint and/or laser don’t tarnish the metal, that is.

One problem with this batch of etchings is that the laser actually etches the entire surface of the wood instead of just where the lines are. The machine is still reading the white background as a tone, albeit a very light one, and so if you look closely at this etching you can see a series of fine points across the surface of the white space. It’s no big deal in this case, but it was unexpected and I’d like to have control over these sorts of things. I think this problem could be eliminated by using photoshop’s magic wand selection tool to actually *delete* the white background before importing the image into Illustrator. At the time that I was working on these, though, my installation of Photoshop had been corrupted and so Illustrator was the only part of the Creative Suite that actually functioned. I’ve got it re-installed now, though, so I’m going to try and see if I can’t get these etchings to look even better.

Toasting bread with a laser

Mon, 06 Feb 2012 12:59:32 -0500

Toasting bread with a laser

Toast of the Town Yeah so the bed on the laser cutter is going...

Mon, 06 Feb 2012 12:54:45 -0500

Toast of the Town

Yeah so the bed on the laser cutter is going to be replaced at some point. In the meantime I don’t want to waste too much time and energy on precision projects involving expensive materials (I wasted a whole bunch of acrylic earlier), so I’ve decided to laser etch some bread and turn it into toast.

I’m kind of obsessed with visual puns, so I’m taking images of city skylines and maps in an attempt to produce literal “toast of the town”.

This was another exercise in tweaking settings and playing with variables. I found that my initial etchings were way too faint, and I was able to correct this with a combination of adjusting the contrast on the starting image, and increasing the power on the machine.

Because of the rough, porous texture of the bread, image resolution was severely degraded. I tried flattening one of the pieces, which helped improve image clarity, but not to such a degree that I’m going to keep doing it.

I did a bunch of overhead maps and one NYC skyline. I think the skyline came out much better. The maps are pretty much unrecognizable.

I think the NYC skyline (and other city skylines) have potential. The result could be substantially improved by making the image into a vector silhouette rather than a raster image. This would give a bold, graphic look rather than a photographic appearance, which would help to make the image more recognizable on the low-res surface of the toast.

I’m going to try another session of laser-toasting soon, and if I can get it to look the way I want it, I’ll start offering Toast of the Town on my etsy site, because - why not?

There’s an attachment for the laser cutter which allows you to etch onto the surface of round objects, so if Toast of the Town is successful, you can look forward to Gender Rolls as a followup.

The laser cutter at as220 has damage to the honeycomb structure...

Mon, 06 Feb 2012 12:20:54 -0500

The laser cutter at as220 has damage to the honeycomb structure of the bed which your cutting material sits on. The result is that the bed is not a consistent, level surface. Instead it is a kind of wavy, hilly terrain with a few patches of flat surface and a few patches of severely damaged surface interspersed with vaguely undulating areas in between. As a consequence of this, any piece of material large than a few square inches will only touch the surface of the bed at a few points, and will not lie completely level. What this means is that the material will be at different distances from the laser in different spots, which means the beam will only be in focus part of the time. When the beam is closer than expected, there is a risk of burning, and when the beam is further away, there is the possibility that the material will not be cut all the way through. The only way around this problem is to crank up the power of the beam much higher than it would otherwise need to be, which will ensure that you cut all the way through in most cases, but will also degrade the resolution of your cut as more material is burned away in the process.

Laser Cutting and Laser Etching I had some masonite boards which...

Mon, 06 Feb 2012 12:10:26 -0500

Laser Cutting and Laser Etching

I had some masonite boards which I had previously painted black with the intention of using them as display platforms for my jewelry at craft fairs, and I decided I would try and cut them into identical squares with the Zoa Chimerum logo etched onto them.

The surface etching worked really well - the laser removed the paint to reveal the brown masonite underneath, which gave a nice, subtle visual effect.

The actual cutting, though, didn’t go so well. My first test piece cut just fine, so after double-checking my settings, I went ahead and set the machine to cut 8 more pieces. Unfortunately, the laser did not cut all the way through the material with most of the other pieces. Some sides of the squares were cut cleanly, and others seemed to only be cut about halfway through. Even after calibrating and recalibrating the machine, I was never able to get everything to cut cleanly before I ran out of material to experiment with.

The reason for this, probably, is because the surface of the laser cutter bed is not even. I’ll describe this further in a new post