Today's options for a digital rangefinder camera range from the obscure (see: Epson RD-1) to the unorthodox (see: PIXII) to the exorbitantly expensive (see: anything Leica sells) - leaving practically nothing for the proletariat.
The goal of this project is to shoehorn the guts of a Sony NEX-5 digital camera into the "working man's Leica," the Canon Model 7 rangefinder. Hopefully, the end product will retain the mid-20th century aesthetics and rangefinder focusing of the 7, while injecting 14.2 megapixels of instant gratification from the NEX-5.
This project stands on the shoulders of creators like Oliver Baker, who have done their own digital rangefinder conversions.
I have a terrible, terrible habit - and that is speeding through a project without making detailed documentation of the process. Sometimes the momentum just builds and rather than pause to snap a picture or two - and just like that, the camera is already done! I did manage to take a few in-progress photos, though:
Some sample photos from the camera, using the Canon 50mm f/1.8 (the only LTM lens I own):
I'm not entirely sure if it's the nature of the lens, but it certainly is not very sharp! It improves a little bit when stopped down, but off-center softness is very evident when focusing on distant subjects. I suspect that the sensor is skewed a bit and doesn't align perfectly with the image plane. The sensor is held in place rather lackadaisically without any real measures to ensure alignment other than relying on the dimensional accuracy of the 3D print. One of these days I'll need to place a depth gauge between the sensor and the lens flange, and try to get it dialed in. But for now, I'm just glad that it works!
Ah, finally - some progress to report. All of the parts and pieces are in at last. The pieces I was anticipating the most were definitely the 3D printed pieces. All were printed in Shapeways' MJF plastic, which I have had numerous excellent experiences with. I used their medium-fast service and I had parts on my doorstep within 2 weeks. I also ordered one translucent SLA part (a replacement frameline window for the Canon 7 - mine crumbled into dust during disassembly), which took a couple more days to print and be processed. If I had printed everything in MJF, I'm sure it would have been even faster.
I've already started the process of putting the camera together, but there have been two major setbacks:
Something in the NEX-5 donor electronics gave up the ghost. I've not been so kind to it (generally manhandling the electronics without any anti-static measures) I think the nail in the coffin was plugging in the battery without properly insulating all of the flying wires and exposed PCBs. A short probably occurred which has sent the camera to an early grave. RIP, you will be missed. Anyway, I bought a new donor camera - there is no shortage of these things.
I bought the wrong freaking size of M1.6 hex nuts. This is quite a sordid tale. First, I wanted some thin M1.6 hex nuts, under DIN 439B standard. Those things are unobtanium, it seems. And when they are available, sellers want an arm and a leg for them ($3-4 a pop! And I stupidly designed this thing needing 26 of them...). I finally found them at a decent price, but the seller waited a full week to tell me that they were out of stock and not planning on backordering any. So, I decided to suck it up and buy some regular M1.6 hex nuts (the DIN 934B standard) and shave them down where necessary (there are only a few places where their thickness is actually important). I decided to go the cheapest possible route and buy some on Amazon Prime, not realizing that they were of a nonstandard size. See, standard M1.6 nuts are nominally 3.2 mm flat-to-flat and 1.3 mm thick. These ones I got on Amazon were 2.5 mm flat-to-flat and 1.5 mm thick. This meant that the hex-shaped features I designed in the 3D prints did not hold prevent them from rotating at all. A copious amount of CA glue was needed to fill in the gaps and hold the nuts in place, which worked half of the time. Another quarter of the time, the nuts would simply break loose under torque, or a stray glob of CA glue would gum up the threads. The lesson here? Always check availability of components before you design around them. I took the availability of a simple hex nut for granted, and it keeps biting me in the ass (hence the title).
There was some success, however, with doing some of the electrical setup for the conversion. First, a bespoke solution is needed to connect the battery to the motherboard. To fit inside the Canon 7, the battery compartment of the NEX has to be rotated 90 degrees and sit just behind the motherboard, as shown below. The original flex connector is not meant to contort in such ways, so I snipped off the end and soldered 3 strands of a flat ribbon cable between it and the battery contacts. It's a tedious and careful job to strip away enough of the polyamide insulation on the flex circuit to expose enough bare area to solder to.
The top flex circuit of the NEX has some very important features: it holds the power switch, shutter button, and a button to enter playback mode. These will all be "broken out" via flying wires to locations easily accessible under the Canon 7's top shell (the exploded diagrams in the last entry may help to explain this). More tedium, but after a bunch of squinting and careful tweezer action, 38 AWG leads have been attached to the proper points on the flex circuit:
Here's the idea. Let's make a digital rangefinder camera by combining the electronic bits of a Sony NEX-5 (the OG version circa 2010-2011) and the body and rangefinder unit pulled from a Canon 7.
I should point off off the bat that this isn't something new - there have been various successful efforts to combine film cameras and digital sensors (see: https://frankencamera.wordpress.com/2014/08/30/the-making-of-frankencamera/). My tenuous claim to originality is that nobody (to my knowledge) has done this to an Leica Thread Mount (LTM) rangefinder, let alone the Canon 7.
Perhaps the best way to introduce this project would be to answer the "whys":
Why build a digital rangefinder camera? I don't know, for fun, I guess? Digital rangefinder cameras exist, and if you were so motivated, you could just buy one off the shelf. However, the cheapest regular production digital rangefinder is probably the Epson RD-1 (which had a heroically long production run for a digital camera, from 2004-2014) - which you would struggle to find for under $1,000 on eBay. All for 6.1 megapixels. On the other end of the spectrum you have the PIXII ($3,300, with... interesting reviews) or the Leica M8 or later, which when new, will cost well north of $5,000. Before being usurped by the single-lens reflex, rangefinders were the norm. These days they are novel and unique, offering a different shooting experience. This project should cost much, much less than any of the aforementioned cameras, and might open that experience to those who don't have wads of cash burning Leica-sized holes in their pockets, and don't mind doing a little bit of tinkering.
Why the Canon 7? Well, there are several reasons:
They are cheap - you can find "for parts" cameras on eBay for as little as $50, and working examples are closer to $150, including shipping.
They are plentiful - Canon made well over 100,000 of these cameras in the 1960's and 70's, so I don't feel too bad about chopping one up.
The viewfinder/rangefinder optics comprise a single self-contained module. This makes it exceedingly easy to work with without fears of screwing up the optical alignment.
Why the Sony NEX-5? Again, a few reasons:
They are cheap - used examples can be found around $100, sometimes less.
They are simple inside - Sony made this camera extremely easy to disassemble, by modern standards.
This camera has been used for film-to-digital camera conversions before, so I don't have to tread through unknown territory. It's been done before, and people like Oliver (maker of the Frankencameras) have done a great job of documenting their work.
Based on others' work, I was feeling pretty confident that shoehorning this would work out. But still, I did a bit of research before buying parts. Repair manuals for both the Canon 7 and NEX-5 are available online (and free, with a little Googling) and are great resources for figuring out how things fit together. The NEX-5 also has vector drawings of the camera in various states of disassembly, which appear to be true to scale - which I used to make some mockups. Both the NEX and 7 are very prolific cameras, so there is an abundance of repair walkthroughs and teardowns showing off their guts - all good fuel for the imagination when considering how things will fit together.
I happened upon a 7 and an NEX-5 on eBay for $65 and $80, respectively. Both were in pretty well-used condition, and the 7 had recently undergone some surgery to fix a dislodged beamsplitter for the framelines... which immediately undid itself during transit. So, I had the opportunity to get familiar with the optical layout of the 7. Nothing a well-placed dab or two of super glue couldn't fix.