The Camera

I chose an old Instax camera as the donor for this project is because it includes some parts that would otherwise be very hard to fabricate. In particular, the film developing and ejection mechanism needs to remain completely light tight, while guiding the exposed sheet of film through a set of rollers and out of the camera. It may be possible to 3D print something like this but why bother when used Instax cameras can be found very cheaply in charity shops (US: thrift stores) and auction sites. I found this one being sold for a few pounds for spares or repair. It turned out the only problem was slight corrosion on the battery terminals. A quick scrub with some contact cleaner and an old toothbrush got everything working again. I shot a pack of film just to confirm the main mechanical parts were functioning reliably. 

Initially I hoped to retain as much of the original functionality of the camera as possible. Obviously the film handling mech would be essential but I would have liked the exposure counter to remain functional and perhaps the flash too. This proved difficult. All functions are controlled by a small microcontroller, including the flash and film handling mechanism. At power up the microcontroller starts the motor to extend the lens assembly and waits for a pair of switch contacts in the lens to close, indicating that the lens is fully extended. The timing is quite critical and results in an error if not perfect - this renders the camera unusable until it is restarted. The flash is fully metered and surprisingly complex given how basic the rest of the camera is. However, it turns out the accurate flash is quite important given the characteristics of the film - more on this later. Unfortunately both the flash driver and exposure counter are integrated with the microcontroller and so some significant reverse engineering would be required to retain them.

Making use of the film ejection system was much simpler. Film is ejected from the cartridge by a small lever which pushes the top sheet of film just far enough out of the cartridge for it to be grabbed by the metal rollers. The sheet then passes through the rollers which squeeze the developing chemicals out of the pouch and across the film as it exits through a set of flaps in the top of the camera. The whole thing is driven by a single motor which also drives the lens: the motor spins one way to drive the lens and the other way to drive the eject mech and a ratchet and cam arrangement direct the torque accordingly. A limit switch closes when the eject cycle starts and opens as soon as it completes. It was therefore very straightforward to wire an NO push button switch in parallel with the limit switch, in series with the motor to create a rudimentary eject control. Pushing the button starts the motor which immediately closes the limit switch - this keeps the motor running until the film has been rolled and fully ejected then cuts the power. 

The camera's existing conductive rubber type shutter button wasn't really suitable for switching the motor so I substituted in a sub-miniature tactile switch in the original position. This saved drilling any unnecessary holes in camera and risking light leaks and is very convenient in use, provided I remember that the "shutter" button now only ejects the film - it doesn't actually expose the film of course. 

The Pinhole

The optimum size for the pinhole depends on several factors. Intuitively a smaller pinhole seems preferable in terms of the amount of detail that can be recorded on the film. However, the smaller the pinhole, the less light reaches the film and the longer the time required for exposures. Worse still, diffraction starts to become significant which decreases the effective resolution. The Mr Pinhole website has a calculator for working out the optimum size, taking into account the distance between the pinhole and the film (often erroneously referred to as "focal length", a term which would be correct if using a lens instead of a pinhole). Given the approximate position the pinhole was going to be on my camera I calculated the optimal diameter to be 0.3mm. As I don't have anywhere near the skills or equipment to make this accurately I sourced this part ready made. 

To make way for the pinhole and new shutter I removed the entire lens assembly. This is quite straightforward to do and doesn't require much explanation. However, removing the lens leaves two gaps in the lens turret which I patched up with some black silicone to prevent light leaks. Once this is done you are left with a camera with a big hole in the front. 

The pinhole assembly is mounted on the camera using a Cokin 49mm filter adaptor - this is essentially a simple plastic flange with a 49mm lens thread which is attached to the front of the camera with more black silicone. From this point, anything with a 49mm lens thread can be screwed onto the camera body - I felt this would give some much needed flexibility for experimentation. 

The assembly itself is made from an old 49mm filter. I removed the glass (a lens spanner is useful for this) and replaced it with a piece of 2mm brass sheet cut to the same diameter. In the centre of the brass I cut a 6mm hole. The 0.3mm pre-cut pinhole is CA glued to the back and the electronic shutter on the front, using more black silicone. I found the shutter mechanism on Ali Express for a couple of pounds - various different styles are available. 

The Shutter Controller

The shutter mechanism uses a small solenoid which requires about 5v to operate. Entering the effective f stop (f 180) and film speed (ISO 800) into an exposure calculator app gives shutter speeds of between 1/8 second in bright sun to 30 seconds plus in more gloomy conditions. While a simple switch / battery arrangement would be fine for the slower speeds, clearly I was going to need some help in full sunlight. My solution is a very simple analogue circuit based around the classic 555 timer. A rotary switch selects the speed from 1/8 second up to 30 seconds. Timing isn't wildly accurate given the limitations of the design and the tolerance of components I had to hand but seems to work well enough for the job. I also included a "B" switch for exposures longer than 30 seconds, which can be done manually with the aid of a clock. 

The Viewfinder

As the newly hacked camera has a far wider field of view than before, the old viewfinder now only captures a fraction of the view that the camera "sees". To correct this, I've mounted (using yet more black silicone) a wide angle adaptor of the type sold for mobile phones. The optical quality is fairly poor with poor contrast and a lot of barrel distortion but nevertheless, it now covers approximately the right field of view and is a useful aid to composing shots. 

Tripod Mount

Whilst it may be possible to hand-hold the camera at the fastest (1/8 second) shutter speed, I don't have an especially steady hand so this was always going to be a tripod-mounted camera. Disappointingly, Fuji did not include one so I fabricated a mount by brazing a 1/4" 20 tpi nut to a piece of brass. This piece was attached to the bottom of the camera with yet more black silicone.