This is a teaser for what’s in store for this week. This design is an exploration of the use of 3D printer work for creating form and providing structure in plastic. The first components (black) took over 30 hours to print, with the second set (red) half way through its 36 hour process time. While this sounds silly, the machine runs unattended - all that is involved is designing the parts and letting it run, through the night and all day, so the hours are continuous, not restricted to working days. Part of the reason for the extreme build time is that the product is almost solid plastic,with only the voids needed to run wires and contain electronics. This drives the cost of material up to around $400. Sounds nuts, but this design is a research project to see just how far this can be taken. Besides the physical time it would take to create similar forms, using carved wood for example, would be impossible to consider. CNC machining from billet aluminum, wood or plastic could be used, but the process time would not be a great deal different, plus material cost, not to mention the pile of scrap created. CNC machining centers also cost many times that of the 3D printer and requires considerably more space – and cannot run unattended, so the point is moot. If I were to attempt to duplicate the design employing a shell design using welded or soldered materials, I would need to change the shape and invest so many hours in fab and finishing, that the $400 material cost seems outright cheap in comparison, not to mention the time required, all on my clock, since I don’t run 24/7.
In the fixture head created in week 12, the material cost was just $21, and the FDM (fused deposition modeling) process time only 6 hours total. For a one-off design and/or short run, this is certainly reasonable and an indication of what a smaller hollow shell involves.
The video here shows the machine at work, for those who have not seen one. This particular machine is a Dimension 1200es, which can build a volume of 10″ x 10″ x 12″H. What it does is lay down layers of .010″ of plastic by extruding it in a thin line tracing the form. It lays down both model material (in your choice of color for the part), and support material to allow it to create negative space and overhangs. The internal structure of the parts can be solid, semi-solid (small air spaces) or a semi-open structure of webbing to hold the part together. The material is ABS, and can be easily sanded and painted, as well as glued, and is available for this machine can be had in red, black, blue, yellow, green, white, ivory, and gray.
This is one of the coolest tools around – for product people anyway. It can be used to make working prototypes, finished parts, or for forms to be used with other processes, like sand casting, fiberglass layout or vacuum forming.
Since installing it on Friday, I’ve had it running continuously. The printer que, with parts for my projects and customer projects, extends will into the weekend. Along that time, I will have entered more work for it, meaning it is unlikely to see any idle time in the near future. The result is like having a production person making parts while I do other things.
If you are anything at all like me, you are getting older by the minute. This means your eyes are getting less sensitive to light and less adaptable to objects near or far. This aging thing has put to the test every theory I have ever had on lighting design in application – particularly my preference for low ambient levels (low energy consumption and drama), with task or accent lighting located at points requiring greater illumination to see detail, read, or execute tasks with high degrees of accuracy. This is the foundation of my personal affinity for portable lighting, as it allows me to adjust and orient lighting in my spaces to suit current needs, without having to hammer holes in drywall. I like spaces with a lot of small sources, some for visual interest, some for ambient illumination, and others for task lighting. Which leads to the weeks project.
Since jumping into this 52 in 52 project, with all of its soldering and detailing, I’ve needed a task light that is able to produce very high light levels (500+ Fc), that does not push heat at me, that can be located to produce very intense light in a very small spot, or a wide pattern over the work area for general detail work. With this in mind, the weeks design emerged. First, it uses (6) Seoul Semiconductor P4s I have a spool of these, so they will be appearing here and there.) In front of each I decided to use some optics I purchased from a vendor on Candlepower Forums, which generate a 35 degree beam pattern with a nice center focus and excellent cutoff at 30 degrees from vertical (60 degree field angle). I grabbed a trusty LuxDrive 1A driver with a pot dimmer, and it was off RhinoCAD to design the parts.
To orient the head anywhere I want, I adapted a knuckle design from a photography umbrella stand. In this case, it allows me to locate the horizontal arm anywhere I want on the vertical axis. It also allows the horizontal arm to swivel at any angle, as well as tilt at any angle. Finally, the horizontal arm can be located horizontally anywhere along its length. In other words, with a single knuckle and two wing knobs, I can put the light head anywhere I want within the X-Y-Z axis, as well as tilt the head to produce a tight focus (aimed level), to a wider pattern (tilted at an angle). With all of this, I can get as much as 1000Fc on a small area, like when soldering little wires on little pads.
The structural components of the light are stainless steel and aluminum, with a brass counter weight. I left all of the metals raw because I like metal and the way it patinas with time and abuse. The head body is made from ABS plastic, grown on my Dimension 3D printer, allowing me to create a complex housing I could not create using my manual machining equipment. The head houses the driver, conceals the wiring, locates the optics, locates the LEDs (mounted on star boards), and provides a foundation for the aluminum cube heat sinks. We just purchased the printer, which generates rigid ABS plastic parts that are usable as real components. In fact, it was installed on Friday of last week, so this task light had to be designed, printed, machined and assembled in a couple of days. Next week will be a more involved all-ABS design to show what this thing can do. The base plate and some of the first parts for week 13 are in the photo for this product in fact.
As usual, more can be found at Lumenique
This weeks project is a little different and slimmed down. Each of the six tiers contains a Seoul Semiconductor P4 LED, all driven at 700mA with integral dimmer. the vertical array provides a nice pool of light with no visible brightness, while the top of each disc below the sources acts to produce a little indirect light.
Also included this week is my first attempt at making video in the process of building the product. If you have 4 minutes, take a look. I learned a lot from the process, so next attempts will improve as we go.
In the video you’ll see that each part takes several steps, from blanking to finished shape, then for multiple parts it has to be repeated several times. I enjoy the time spent in the shop, as its productive and is done quietly (other than the machine noise). Gives one time to think about what’s next, both on the project in process and the future of things in general.
Next week we should have a new capability on hand, which will change the shape of things to come. Stay tuned!
More at Lumenique.
After last weeks push and work with massive materials (relative to others I have done) I needed to take a break from complex machining and do something simple and straightforward. This small project is based on a request to create a small desktop item that was lower in cost, coupled with some ideas I’ve toyed with for making picture lights, from wall mounted to desktop. This is the first shot at a small desktop lighted picture frame.
To make this low in cost, I dug into a box full of unused phone chargers (from phones that have been long since abandoned) – all with voltage and current characteristics suited to LED application. In this case, the Nokia ACP-7U charger, rated at 3.7VDC, 350mA current is a decent match to a Seoul Semiconductor P4 (3000K 92CRI) mounted on a star board. As a test, I soldered the wires from the charger to the LED and stuck the LED to a heat sink and let it run for two days. The charger was warm, but not hot, and the LED seemed happy enough, so I set it aside and went to work on the frame.
Looking at it now, the frame is a little heavy handed, as was its intent. I had my eye on something more mechanical, so it’s made from solid stock, which is carved out to make a recess for the LED and a reflector insert. The switch is a pushbutton on-off control (no need for a dimmer here.
As a proof of concept and first iteration, it works very well. I like the picture lighted up, and the energy it uses is nothing – even with the charger, it’s less than 4W total.
Now, for those looking at that picture and wondering – yes, that is me on the ladder in the fatigues. This image is from 1978. As a graphic artist in the USAF, I painted the sign, which was placed on a little hut each group on base constructed as part of the annual open house. This was one of many signs, posters, flyers, handouts, and ad inserts I did that year. I look at this and think how appropriate it is that two senior NCOs, a Captain and a Colonel were required to install a 30″ x 36″ plywood sign on a bamboo hut. My fiance at the time (now wife of 31 years) took the picture.
I have some other ideas for little picture frame lights and some new methods of making parts that will make them affordable as well. Future designs will be lighter for those who prefer such things. For anyone interested, visit the the Lumenique Product Center or the Lumenique web site. I can make these in any color, as well as configure them for the photo format of your choice.
Better a little late than never for Week Nine Business had me running elsewhere, so there was little shop time to be had, but its done now at last. This is a monster compared to most of the others. First, its heavy, as the thinnest metal used in the thing is 1/4″ wall solid aluminum. The glass is even 3/16″ thick, clear on one side and frosted on the back. It has a very architectural look and feel, and mass to it. Would look great in a corporate lobby or as a hotel desk accent.
The light source is a Philips Fortimo remote phosphor light module rated at 1100 lumens. The structure is 35″ tall and is 8″ square at the base. Gray wrinkle powder and a special textured finish on the aluminum vertical bars adds some sparkle, while the white reflective rings provide diffusion.
One of the issues I faced on this one is the massive scale of the light module and its related power supply. There is just no way to put these into a small product.This was obviously not the design intent for this light source, but it does produce a nice color at 3000K and 80CRI.
Now on to something a bit more sexy for Week Ten, perhaps something in Ferrari Red.. Oh yeah, bought a neat little camera for making You Tube movies, so will make an attempt to capture some action to be included on future design projects.
As always more on this can be found at Lumenique Week Nine
This week is going to be my first shot at putting the Philips Fortimo light module into a portable product. The design will incorporate the 1100 lumen module at 3000k. The design itself is going to be industro-architectural, with a combination of thick glass panels with exposed edges, textured gray powder coat, white, and hand finished satin aluminum. It’s going to be around 36″ tall and will be heavy, since the glass and metal work are robust, and the scale of the unit is rather on the large side.
To manage heat, I’m incorporating the usual tie to the aluminum structure, as well as a supplemental heat sink in the base.
This should be a great uplight/ambient contributor, as well as a significant presence. Could easily be scale up to a floor lamp model using the 2000 lumen light source – might have to try that later in the year.
03/07/10 - Sunday night update. Have all of the parts made – which has been interesting, since most of them are at the limit of what my machines can handle – and have everything finished and ready for assembly. Problem is, it’s 8:00PM Sunday night and I need to stop to get ready for a road trip Monday and Tuesday. The finished Week Nine project will be done Wednesday, leaving me to jump right into Week Ten over the weekend to get back on track. It’s no problem, just a stretch of available time. I think this weeks effort is going to be interesting, it’s massive and minimalist, yet has a very nice edge to it. so far so good. Also, ordered a camera to film some of the action here to post in future updates and to produce some LED/SSL lessons as well.
I am particularly happy with the way this retrofit and upgrade turned out. It was fun going through a fixture I had made over 17 years ago, revisiting how I made things back then. This fixture is primarily brass and copper, soldered together to make the structures and shapes. This took a lot of filing to smooth corners and requires some patience, as heating metal to make solder joints often causes others to soften and fail, or melt out altogether. This design was also made without a drawing originally, not even a sketch, as is the case in a lot of one-off work I have done over the years.
The upgrade here blends the old with the new. I machined a new lighted section, which acts as heat sink/detail intended to look like a power source. The Two LEDs inside are Bridgelux 400lm arrays, and the 700mA driver is contained in the lower housing, along with a dimmer control.
Originally, this fixture utilized a 50W T2 halogen lamp in a somewhat poor performing and glaring reflector housing. An image of this can be seen on the Lumenique site here. The upgrade removed the old magnetic transformer, and uses less than 14 watts now, a reduction on consumption of 72%, while improving the amount of light generated and reducing unwanted glare.
The glass on this are wing windows from a Porsche 914, sand etched. I once thought these would need some form of support to endure, but find that this was unnecessary, so eliminated it.