I’m using a single 4 inch telescope mirror here. Setup is not particularly hard but is a little tricky mostly in terms of getting the alignment just right. There are also 2 mirror setups that can be used. Here’s a picture of my mirror mounted behind my printer.
Here’s a quick link describing the technique in more detail.
Here’s another example I did using a simple power resistor as the heat source:
I’ve been wanting to try this for awhile on a 3D print in progress. Getting a good angle is a little tough because the camera has to be pointed right at the mirror in order to pick up the differences in refractive index present in the air being stirred around.
Sometimes you can get a good deal on ebay for a telescope mirror, which is where I got this one. At 4 inches, it’s a bit small but still allows the principal of Schlieren photography to be demonstrated.
I probably said something besides oops at the time I had submitted an STL file to one of the 3D printing rapid prototyping services online, and had not checked the STL file that my CAD software had output. What I had intended was to make an array of parts interconnected by some webbing so they could be split apart. Everything looked fine in the CAD software, because I was looking at the assembly in the native CAD format. What I had not realized was that the “Save As STL” didn’t go so well. I was in a hurry and submitted the file for quote, placed the order, and what I got back was this:
Those few little strips of plastic cost about $250! Wondering what the heck happened, I called the vendor and had them email the file back to me. Sure enough, they printed according to the file they received. I finally traced it back to a problem saving the file properly so that all of the components in the assembly ended up in the STL file.
Most good vendors would look at a file like this and ask if I really wanted to print some little pieces of garbage before just happily printing it an accepting the money! I’d worked with this place before and they’ve been pretty good. Ultimately I can’t blame anyone but myself for not checking the STL file before ordering.
So be sure to check those STL files by loading them back into your CAD/CAM software to avoid surprises!
Remember Lightscribe? If you do, and you’re like me, you probably bought a stack or 2 of Lightscribe disks after Lightscribe was introduced, printed a few, and then put them on the shelf. It’s a pretty neat thing though and this post I’ll describe how you can do some optical encoding with Lightscribe by making a cheap and fairly precise optical encoder disk.
If you’re not familiar with Lightscribe, it’s a product/technology that allows you to “burn” your CD labels onto the back of specially treated CDs, DVDs, etc. I’ve been intrigued by thoughts of using it for other purposes, such as this. Now it’s true that you could also simply print a label and stick it on a CD and not bother with lightscribe, but there are certain advantages to Lightscribe. Labels can stain, be applied off center, peel off, possibly even cause balance problems at high speeds. Lightscribe will add no mass to the CD and allows precise printing. You can even print multiple passes with good registration between passes which is interesting in it’s own right.
What is an optical encoder disk you might ask? An optical encoder disk is a disk of material that is often made of glass for precision rotary position sensors for the purpose of precisely determining the angular position of a rotating shaft. It works by putting light and dark patterns or stripes, dots, etc. onto the disk in such a way that optical sensors can detect the light and dark stripes generate a binary number that directly report the position at a given time. Suffice it to say that the light and dark patterns don’t really even have to be visible to the naked eye, but adequate for an electronic sensor to assign a “0” or “1” logic levels to a particular color, shade, reflectivity, opaqueness, etc. The more bits you string together the more steps per revolution and precision you can get from your encoder.
Lot’s of information about rotary encoders is available on the web, such as this Wikipedia entry:http://en.wikipedia.org/wiki/Rotary_encoder
Many many many types of clever optical encoders have been devised, Some are linear, non linear, travel in straight lines, or complex motions. The image below shows a CD that I made using an encoder pattern that I modified to fit onto a CD. I started with the binary encoded disk image I found here: http://www.qsl.net/in3otd/roten.html There is also a Gray code encoder image available. There are certain advantages to using Gray code that are touched on in the Wikipedia article mentioned above. I’ve using the binary coded one here for illustration because it’s pretty straightforward to read visually. You can draw an imaginary line from the center to the edge and easily determine the “1” and “0” bits and directly read the position. Otherwise you’d need to convert the gray code to a number using a process that’s described in this presentation: http://www.wisc-online.com/Objects/ViewObject.aspx?ID=IAU8307
I had to open up the hole in the center to accommodate the printable area on a Lightscribe disk, otherwise and important part of the pattern would be left out. Towards the center of the disk are the “most significant bits” of the encoder and toward the edge of the disk lies the least significant bits. As you can see, when this disk spins, the transitions at the edge will happen more frequently than the transitions toward the center. Each successive “ring” toward the center will have transitions that are half the rate.
Here’s a picture of the CD I burned the encoder image onto. Perhaps in a future post we’ll rig up a little fixture to demonstrate this in operation.
I think a large share of CD writers being sold today still have Lightscribe capability. I wonder how long that will last, because I doubt many people use it. It is pretty neat though. You may think of other interesting things to do with this technology, other than simply printing a label.