I've always wanted to try terrestrial night photography with an actively cooled CCD imager and a regular SLR lens. The combination of a regular(wide to an astrophotographer) field of view with the incredible noise characteristics of a professional imager seems quite appealing. I recently acquired an SBIG ST-2000XCM CCD Imager for astrophotography, so this dream is only one adapter away from reality. At only 2 megapixels, the resolution might be a tad low, but the light collection and low dark current are way beyond any dSLR, even the low noise Pentax K-5. SBIG makes an adapter for their ST-2000 taper to fit Nikon and Canon lenses, but not for Pentax. I could purchase the SBIG to Canon adapter and use a Canon to Pentax bayonet adapter, but that would increase the flange distance and prevent me from focusing on infinity; rendering the setup totally useless for outdoor photography. Instead I machined my own adapter to go directly from the SBIG Taper to the Pentax K-Mount, properly spacing the lens according to Pentax's flange distance specification.
Front and back views of the completed adapter. The first step was a rough Solidworks design.
The design focused on the internal shape and making sure the Pentax ring would fit, I didn't even include the locking mechanism. Initially I planned to leave out any sort of lock, and let the spring detent hold the lenses in. In hindsight I should have incorporated the lock into the original design, when I did decide to include it I spent more time than I should have trying to visualize how it would work while the part was in the mill. It would have been much quicker if I had pre-thought this out and been able to model a few different designs. Thankfully the solution I ended up choosing works quite nicely, and is much more robust than the prone-to-failure locking mechanism on the Pentax K series extension tubes.
I started with 3" Aluminum 6061-T6 barstock, turning it down to 2.5" OD on a Delta-Rockwell 11" lathe. 2.5" was selected to match the OD of most Pentax lenses, purely for aesthetic reasons.
Next came the SBIG Taper. This is 1.84" OD, and interfaces with a matching female bore on the front of the CCD. A small angled lip was machined to give the set screws in the CCD something more to grip. To my knowledge, this taper is used on all of the ST-7/8/9/10/2000/4000 series cameras.
The part is cut off on the bandsaw, to be turned around in the lathe and grabbed by the taper. This will allow facing the other side, and boring out the ID.
Rough bandsaw cut, ready to be cleaned up.
Back in the lathe with the front side faced. Center-drilling before drilling a pilot hole, then boring to 1.02" ID.
Drilling out to 5/8". This was the last drill used before heading for the boring bar. The drills I used were 1/4, 3/8, 1/2, 9/16, 5/8. Those steps are probably smaller than necessary, but I like to minimize load on the bits and keep them sharp.
Boring to 1.02" ID.
Threading to 96 TPI to minimize internal reflections that reduce contrast in the final image. I intentionally run this without coolant and make the cut in one slow pass to induce a minor amount of chatter, leaving a rougher, less reflective surface.
Boring out the larger cavity to make room for the Pentax K Mount Ring, and any parts of the lens that may stick out beyond its bayonet. This is bored to 1.95" ID.
After chamfering the internal and external edges, all the lathe operations are complete. On to the mill.
The rotary table was already set up vertically, so I decided to cut the grasping grooves on the side first. I used a 1/2" ball nose end mill with a .05" depth of cut. To avoid running into the jaws I placed 3 sets of grooves spaced 50 degrees apart evenly 3 times, resulting in 6 grooves. The spacing is 50 degrees, 70, 50, 70, 50, etc.
After moving the rotary table to its horizontal position, the internal shoulders for the Pentax K-mount and brass tension ring are milled. I decided to mill these instead of boring them on the lathe to ensure perfect concentricity with the circular bolt pattern for the ring. By milling them the work doesn't need to be re-chucked between cutting the internal shoulder and drilling the holes, whereas moving it from the lathe to the mill could induce a 2-3 thousandth discrepancy.
Preparing to cut an internal baffle again to minimize reflections. This is probably not necessary, but I wanted to try cutting an internal undercut slot with a T-slot cutter on the rotary table.
Cutting the internal baffle. It worked pretty well, although the edge finish on the bottom side isn't as smooth as I would like. I was careful and took 4 passes to get the depth I wanted, I imagine this created the poor finish. I should have advanced the cutter a few thousands downward in the Z direction and made one last pass to clean up that edge.
Drilling the circular bolt pattern for the Pentax K-Mount ring. I would have liked to use metric screws like the ones used on Pentax, but I don't have any metric taps that small. I went with Stainless #0-80 flathead screws. While these work, I would have preferred Phillips head to maintain some sense of originality(Pentax cameras use Phillips on the mounting ring), but the local hardware store didn't have any #0 Phillips screws. The drill chuck on my mill only goes down to 1/16", so to drill the 3/64" holes necessary to tap a #0-80 holes I had to clamp the tailstock chuck from my Cowells lathe into the mill chuck. It's not an ideal setup but it does work. The holes were center drilled first, then drilled out to 3/64", tapped, then a 1/16" bit was advanced .0125" inches into the hole to remove a burr that might prevent the ring from sitting flat. This is particularly important because a small 1-2 thousandth burr could cause the image plane to not be parallel to the sensor, making it impossible to focus an entire field at once. All four steps were completed on each hole before advancing the rotary table to make sure each tool entered the hole centered.
Tapping the #0-80 holes. Typically I'll put the mill in neutral and chuck the tap, advancing the feed and spinning the chuck by hand to start it perfectly aligned and straight, but I didn't feel comfortable doing that with such a small tap. This concludes the bulk of the work on the main body. The next part to make is the pin that locks the lens on to the mount. I made this part out of 3/16" brass C360 barstock on a Cowells 90CW watchmaking lathe.
The brass is turned down to 1/8", and the end is turned down to 1/16" for .15". The smaller diameter protrudes from the Pentax ring .034", locking the lens in to place.
After flipping it around in the chuck, a shoulder is machined to hold the spring in place. The spring was harvested from a small RadioShack pushbutton.
Pin, spring, and mount in place to test. When the lens is inserted the pin is pushed out of the way until a small hole in the lens' bayonet lines up. When this happens the spring(on the bottom of the pin, not seen in the photo) pushes it back up, locking the lens in place. Next is to machine a small lever on the outside of the body to pull this pin down, releasing the lens.
With the rotary table back in the vertical position, a small 3/16" by 1/4" slot is machined for the lens release lever. The 3/16" portion is only .02" deep, with the rest of the depth a 1/8" by 3/16" slot. The lever is T shaped(described later), and covers up the other half of the slot at all times to prevent dust and light from entering.
After drilling a hole through the side of the locking pin in the mill, a #1-72 thread is tapped to screw the lever on to the pin.
Back to the Cowell's lathe to make the lens release lever. It's 3/16" OD for 1/8" and 1/8" for 1/8". The 1/8" part protrudes through the body of the adapter and attaches to the locking pin by a #1-72 screw, while the 3/16" part covers the rest of the exposed slot and sticks out for the user to push.
Drilling a clearance hole for the #1-72 screw. After chamfering the lever, all the parts are complete.
All the parts, laid out for assembly. The top two are the Pentax mounting ring, taking from an extension tube. The 6 screws in the bottom right are the #0-80 screws for the ring. Next from left to right is the #1-72 screw, locking pin, lever, and spring. Before assembly, I glass blasted the outside of the body and locking lever.
Masking the areas I don't want to get blasted. I plan to paint the inside of the adapter with antireflective paint, so blasting the inside would only lodge particles in places I don't want to clean for very little benefit. It's surprising how little it takes to stop an area from getting blasted; masking tape is charred by heat from friction with the glass before abrasive forces remove it. Stuffing a small amount of modelling clay in holes that I don't want to get blasted also keeps the glass out, as long as I don't keep the stream on that area for too long. The locking slot can be seen filled with clay in the left image above. An old Pentax K mount cover was used to cover the main opening.
In the blast cabinet. In addition to leaving a pleasant matte finish on the part, glass blasting work hardens the surface, making it more resilient to scratches and scuffs.
The body, after blasting and with masks removed. Despite the matte look the surface is quite smooth, and it won't pick up fingerprints. The locking lever was also blasted to give it a similar look. The last step is to paint the inside with antireflective paint, which I don't currently have on hand. The adapter is usable in its current state, I'll post pictures from the setup as soon as I test it out.