When I started the hobby of repairing old cameras, I soon realised that I needed a shutter tester. I built a single point one that works nicely, in particular when the camera under test behaves consistently. I can test the exposure time in the middle of the frame, then move to the edges. I thought about adding two more measurement points, but I was never motivated enough to do that. I know that I can buy ready-made ones that work well, but where’s the fun?
A while ago, I started working on a Gamma II (https://www.dpreview.com/forums/thread/4755595). The shutter behaved really randomly, with measurements – even in the same position of the frame – being very erratic. I could tighten or loosen the curtain springs with exposure times increasing or decreasing in an unpredictable way. I later found out that this was caused by the springs occasionally becoming detached to one of their anchor points loosing the tension.
I thought that in this situation, even a three-point shutter tester could be of limited use. On the other hand, a CRT monitor could show the behaviour of the curtains throughout their entire travel. CRT monitors are not that common household items nowadays, so I did some online searching and found some on offer on ebay for approximately £20. Interestingly, they were also quite small – 4” diagonal, of the type used in door CCTVs, with side projection – which I thought would help framing the CRT through the camera film opening.
A few days later, I started playing with my CRT. It requires a 12V power supply and two trimmers to regulate the contrast and brightness.
The CRT on the left, with a Pico-based controller on the right in a 3D-printed enclosure
A useful characteristic of this particular model is that when powered up and after regulating contrast/brightness it shows by default a bright background, so it can be used straightaway to visually test shutters! The pictures shown in the following were taken with the following arrangement:
An example of how the CRT appears through the shutter of the Gamma II is show here:
CRT picture captured while activating the shutter of the Gamma II
A CRT tester can highlght several anomalies. The weird stripe on the left of the following picture is due to the second shutter curtain bouncing back on the Gamma II (the reason why the following two pictures show fine horizontal lines will be explained below).
Second curtain bouncing back
Below, the stripe ending with a sharp point (upper left corner) indicates capping: the second curtain reaches the first before the end of the frame opening.
Capping
Being a bit nerd-minded, I started wondering if I could take more quantitative measurements. When looking at the display while shooting a film camera with open back, the CRT appears as curved stripe(s). The CRT horizontal and vertical scanning is well controlled, so vertical extent of the stripe(s) is precisely related to the exposure time, but the stripe apparent height also depends on the distance between the observer – my eye or a digital camera – and the CRT. The solution was to show only one out of three raster lines (could do one of out two, but the contrast was poorer). This required connecting a Raspberry Pi Pico to the CRT input and some programming, but after some tinkering, I had nice stripes on my CRT. An example is shown below. Since every raster line takes 64 microseconds, three lines (one bright, two dark) take 192 us, or approximately 200 us. So, if I can count say 5 lines, that means that the exposure time is 5 * 200 us or 1/1000 s and this can be done for the whole extent of the frame. Even better, this counting can be done relatively easily in an automated way by a Python script.
About 18 stripes vertically correspond to 1/300 s
Some coding later, I tethered my Lumix S5 to my PC and used it to take pictures of the CRT through the camera under test. The tether program is configured to transfer the pictures automatically to a folder on my PC, while another Python program scans the folder for new pictures. When a new picture is detected, it is analysed automatically and the results shown graphically in a window, as shown below.
The left/right edges of the stripe are detected and the horizontal raster lines counted vertically and converted to exposure time (bottom-right panel below). Knowing that the frame is 36x24 mm, also the speed of the curtains can be calculated (bottom-centre panel), showing the acceleration of the curtains through their travel.
Canon VT deluxe at a nominal speed of 1/500 s. The actual speed is approx 1/650 s, very uniform across the frame. The curtains about 5-6 mm into the frame is 0.8 m/s, accelerating to approx 1.4 m/s.
All in all, a very fun project that proved really useful in fixing the shutter of the Gamma II. Happy to share details.
A while ago, I started working on a Gamma II (https://www.dpreview.com/forums/thread/4755595). The shutter behaved really randomly, with measurements – even in the same position of the frame – being very erratic. I could tighten or loosen the curtain springs with exposure times increasing or decreasing in an unpredictable way. I later found out that this was caused by the springs occasionally becoming detached to one of their anchor points loosing the tension.
I thought that in this situation, even a three-point shutter tester could be of limited use. On the other hand, a CRT monitor could show the behaviour of the curtains throughout their entire travel. CRT monitors are not that common household items nowadays, so I did some online searching and found some on offer on ebay for approximately £20. Interestingly, they were also quite small – 4” diagonal, of the type used in door CCTVs, with side projection – which I thought would help framing the CRT through the camera film opening.
A few days later, I started playing with my CRT. It requires a 12V power supply and two trimmers to regulate the contrast and brightness.
The CRT on the left, with a Pico-based controller on the right in a 3D-printed enclosure
A useful characteristic of this particular model is that when powered up and after regulating contrast/brightness it shows by default a bright background, so it can be used straightaway to visually test shutters! The pictures shown in the following were taken with the following arrangement:
An example of how the CRT appears through the shutter of the Gamma II is show here:
CRT picture captured while activating the shutter of the Gamma II
A CRT tester can highlght several anomalies. The weird stripe on the left of the following picture is due to the second shutter curtain bouncing back on the Gamma II (the reason why the following two pictures show fine horizontal lines will be explained below).
Second curtain bouncing back
Below, the stripe ending with a sharp point (upper left corner) indicates capping: the second curtain reaches the first before the end of the frame opening.
Capping
Being a bit nerd-minded, I started wondering if I could take more quantitative measurements. When looking at the display while shooting a film camera with open back, the CRT appears as curved stripe(s). The CRT horizontal and vertical scanning is well controlled, so vertical extent of the stripe(s) is precisely related to the exposure time, but the stripe apparent height also depends on the distance between the observer – my eye or a digital camera – and the CRT. The solution was to show only one out of three raster lines (could do one of out two, but the contrast was poorer). This required connecting a Raspberry Pi Pico to the CRT input and some programming, but after some tinkering, I had nice stripes on my CRT. An example is shown below. Since every raster line takes 64 microseconds, three lines (one bright, two dark) take 192 us, or approximately 200 us. So, if I can count say 5 lines, that means that the exposure time is 5 * 200 us or 1/1000 s and this can be done for the whole extent of the frame. Even better, this counting can be done relatively easily in an automated way by a Python script.
About 18 stripes vertically correspond to 1/300 s
Some coding later, I tethered my Lumix S5 to my PC and used it to take pictures of the CRT through the camera under test. The tether program is configured to transfer the pictures automatically to a folder on my PC, while another Python program scans the folder for new pictures. When a new picture is detected, it is analysed automatically and the results shown graphically in a window, as shown below.
The left/right edges of the stripe are detected and the horizontal raster lines counted vertically and converted to exposure time (bottom-right panel below). Knowing that the frame is 36x24 mm, also the speed of the curtains can be calculated (bottom-centre panel), showing the acceleration of the curtains through their travel.
Canon VT deluxe at a nominal speed of 1/500 s. The actual speed is approx 1/650 s, very uniform across the frame. The curtains about 5-6 mm into the frame is 0.8 m/s, accelerating to approx 1.4 m/s.
All in all, a very fun project that proved really useful in fixing the shutter of the Gamma II. Happy to share details.
Attachments
Last edited:
