The shutter speed meter project

Jesito · Expire: 2015-01-07

Hi colleagues,
As you know, I'm behind such a project since quite long ago.
I've been taking some notes and putting them into my server whilst it's under development (it's easier to me for changes), but will be a forum article when finished.
If you want to take a glimpse to the work in progress, please go to

http://watersmurf.homelinux.com/photo/diy/shutterspeed.html

Open to hear your ideas, suggestions, etc.

Jes.

Xpres · Expire: 2014-10-28

Jes - What a great article, thanks for posting. Cool

I'll have to stick to the audio method as an oscilloscope and and the accompanying leap in technical know how is beyond me. But I do like the idea of a laser pointer as a light source - I was using a desk lamp...
For testing rather old shutters, often with only three or four speeds, I wonder if one might not just take readings relative to each other so the absolute result is not important - perhaps using one second as a base?

LucisPictor · Expire: 2013-12-03

That's professional, Jes!

Orio · Expire: 2012-12-04

Hat off, Jes. I would never be able to imagine such a thing, let alone build it.

Jesito · Expire: 2015-01-07

Thanks, colleagues, but this is just "work in progress", maybe it ends in nothing...
But I try to compile all the information I've been able to get, plus my own ideas.
Tonight I expect to try the "oscilloscope/laser" combo, I'll take some pictures.

Jes.

Jesito · Expire: 2015-01-07

Well, well, well...

In the first place, and taking into account that no everybody has an oscilloscope at hand, I've tested the soundcard version. Peter was waiting for this to start building his own one... And this seems to work!.

This is the schematic:

(I used a 4k7 resistor instead the 3k9 one, no problem... Resistors are like film latitude... Wink

).

I built it on a perf board, and since it's a prototype, it has been built the dirty way: Batteries soldered, etc, just to see if it works.
Also not using the intended laser as a light source. I've used a hi-bright LED diode instead.

The layout is critical. Some physical stability is needed to be useable... But for a prototype everything is valid. I used a "third hand" device to hold the board with the phototransistor:

The ligth source comes from a tiltable lamp, cannibalized for this purpose:

The board in place:

Finally, measuring is quite easy: I used Audacity, started recording a track, fired the shutter at 1/300 speed. As soon as I fired the shutter, I stopped recording the audio track. There was a tiny mark that had to be widened a lot to be able to measure the width. By selecting the space in between the two marks, you get the time at the bottom of the screen. It has to be inverted (1/x) to get the speed, that in this case is 1/87,49 far slower than it should be... Not strange, the shutter gets blocked at lower speeds, so it need a deep cleaning and review.

And that's all, folks!

Jes.

Throndor · Joined: 15 Sep 2008 Posts: 157 Location: Ankara / TURKEY

Nice one Jes,
Isn't it simpler to measure the shutter speed with a disc on a motor and a digital camera..

Xpres · Expire: 2014-10-28

Throndor · Joined: 15 Sep 2008 Posts: 157 Location: Ankara / TURKEY

peterqd · Expire: 2014-01-04

Very interesting Omer, thankyou. I believe a rotating disk idea was used when the speed of light was first measured. In the old days of photography, when using a flash meant setting fire to some Magnesium powder, the technique was to open the shutter with the bulb, fire the flash by hand with a spark igniter and then close the shutter. The speed of the ignition of the powder was so slow that the shutter might be open for many seconds. Modern cameras do exactly the same but much faster, the flash sync speed is the fastest speed where the first shutter curtain is fully open before the second curtain starts to close.

Jesito · Expire: 2015-01-07

peterqd · Expire: 2014-01-04

Hi, time for an update on the MFL International Shutterspeed Meter Project (MFLISMP - not quite as complicated as the CERN Project but a much nicer acronym!). At least this is my half of it, I'm hoping Jes will arrive soon.

Following on from the Spotmatic battery problem, Jes and I got talking about the shutter speed meter which he'd been working on. I needed to check the shutter speeds on a few cameras, so Jes's work interested me. There are several sites on the net with circuit diagrams of simple shutter testers using the PC soundcard and audio software. It seemed worth trying.

We started by thinking about a light source. Jes's cheap LED light wasn't giving good results, we needed something more directional and more intense, and Jes hit on the brilliant idea of using a laser pointer! We tried the very cheap Chinese ones on Ebay at first, but they turned out to be poor quality with a very short lifespan, so Jes visited his electronics store in Barcelona and bought some high quality laser elements and very kindly sent three of them to me, along with a few other components and even a bundle of different coloured wires! The beauty of these laser elements is that they require a 3v supply which can be taken from a USB port on the PC. I had an old USB cable handy so I was ready to start the Mark I prototype!

In the meantime, I'd been giving some thought to how to mount the laser and the camera and hold it steady while releasing the shutter. I had an old flat flash bracket with a long slot for adjustment and I worked out a way of mounting this on an MDF base, and by fixing it with a hinge I was able to adjust the camera vertically too, to centre the laser beam in the camera picture frame. Here's the finished test-bed. (This pic is actually Mark II - see lower down)

The laser and the receiving photodiode are mounted in holes drilled in the vertical sides of the contraption, and obviously the holes need to be lined up. This wasn't easy to do, having already fixed and glued the sides. Mark IV is going to have all the holes predilled together through both sides before fixing them. This pic shows the laser beam and the photodiode waiting to be connected to the PCB.

The photodiode acts like a simple switch, allowing current to pass when it is activated by light, so just a couple of wires are needed to run to the PCB. The PCB is just a small piece of perfboard with the components soldered on it, including a 3mm jack socket. There is also a 1.5v AAA battery wired in, not shown in the pic.

The setup was now ready for the first trial. I mounted a camera, plugged in the USB power cable and plugged the signal cable into the PCB socket and the Mic socket of my soundcard, and started recording the audio signal with Adobe Audition. The Audition screen can be set to show the time in seconds to three decimal places, ideal for measuring 1/1000 second exposures. At the same time I ran an Excel spreadsheet so I could enter the two time values taken from the screen and convert these into exposure times.

There was a definite blip for each shutter curtain on the signal trace and the time interval could be measured, but not very accurately. The diode has a slight delay which causes a curve on the signal trace, not a nice vertical line, so it was difficult to pinpoint the passing of each curtain. Another problem was keeping everything steady. Even with the test bed clamped down, holding the camera steady while winding and releasing the shutter is hard enough, but with a loose PCB, a battery and lots of loose wires hanging off it, it was very fiddly. The components needed to be mounted properly. We thought about the wavy signal trace and Jes had another great idea to add a variable resistor into the circuit in order to fine tune the signal. So I went to my local Maplins (like Radio Shack) and bought a box and battery holder, plus a few other components, including a 10k v. resistor, a panel-mounted jack socket and a power switch. All the components could now be mounted inside the box and screwed to the side of the test bed. The piece of perfboard wasn't needed any longer, so soldering everything together was very simple. This is Mark II:

I tested virtually all my cameras using this model, and it gave very interesting results. The additional resistor made a big difference. I could now see why one of the Spotmatics was overexposing, the 1/1000 speed was over 2 stops too slow! However, with just the one signal, I couldn't tell whether one curtain or both were running slow, and whether the exposure was consistent across the frame. I figured with two diodes and two lasers at a measured distance apart, and using the stereo Line-In input into the soundcard I could find out this information, and yesterday I finished Mark III. Besides the extra laser and diode, I raised the camera higher with a deeper section of wood so that it was level and the plane of the two beams was parallel with the shutter movement.

Here is a screenshot of the audio recording I made for the errant Spottie at 1/30 sec: As the first shutter curtain passes each beam it exposes the diode to the laser, the current is connected in the circuit and an audio signal is recorded on the PC. When the second curtain passes it turns off the current and the signal is stopped. This causes a blip on each of the stereo channels for each curtain, so four times can be recorded. The exposure time at each diode is the distance bewteen the two blips on each channel. The travel speed of each curtain can be calculated by measuring the time difference between the two channels.

On this picture, the section highlighted represents the exposure time at the first diode. You can see the start and finish times and the length of the selection in the panel at bottom right. In this case the length is .038 seconds, or about 1/26 sec. so the exposure is slightly too long. The exposure time at the second diode measures 1/23 seconds. This inaccuracy increases as the speeds get faster. At 1/1000 I recorded 1/166 and 1/142 secs - still over 2 stops too slow.

The design curtain travel speed for the Spottie is 14ms across the complete frame +/- 0.3 ms. This equates to a speed of just over 2.5 meters per second. The times I recorded gave speeds of 2.4 m/s for the first curtain and 2.14 m/s for the second. It is now clear how much each curtain tension needs to be adjusted.

The trace has another interesting feature. The light hitting one of the diodes makes a blip on the signal from the other diode. I think this might be due to light entering the black box though each diode, so I am going to try sticking some chewing gum over them! Smile

ludoo · Expire: 2011-12-05

Very interesting. Could you post a list of parts and a schematic for the updated version, with the stereo jack and vr?

Attila · Expire: 2025-11-18

http://www.open.hr/~dpleic/photo/shutter_v2/shutter2.html

I am also thinking on it I found this.

peterqd · Expire: 2014-01-04

The circuit I built is a lot simpler, but it has two diodes and two lasers. I didn't use the small capacitor shown in the PCB picture. The basic diagram is here: my diodes are 24mm apart, not 18.
http://www.kyphoto.com/classics/combinationtester.html

I used a 10k vr similar to this:
http://www.maplin.co.uk/Module.aspx?ModuleNo=6499

and the diodes are like this: (the clear one, not the black IR one)
http://www.maplin.co.uk/Module.aspx?ModuleNo=2242

Jes sent me the lasers, I'm not sure what values they are. They look like this:
http://www.maplin.co.uk/Module.aspx?ModuleNo=11712

You can find the box, sockets, power switch and the battery holder etc on the Maplins site too.

This is the spreadsheet showing the full results (split into 2 for readability.) The bottom row shows the curtain speeds, 1st and 2nd for each speed. See how slow the 2nd curtain is compared to the first, especially at slow speeds! The exposures are quite accurate at slow speeds, where the curtain speeds aren't so critical.

Farside · Expire: 2013-12-27

Excellent work, Peter and Jes.

Arkku · Joined: 28 Feb 2007 Posts: 1416 Location: Helsinki, Finland

I've been successfully measuring shutter speeds up to 1/1000s with a phototransistor (taken from an old mouse) wired to a 3.5mm plug, recorded through the soundcard.

I've been thinking of making a computer-less version of the same using a microcontroller, which would then output the measured time to an LCD display… I could combine this functionality with my camera trigger project, which I want to be able to trigger the camera when a laser beam is interrupted, etc, so the same components would already be there. However, it will be a while before I have the time to do it all, currently I've just written the code for controlling LCDs and such.

peterqd · Expire: 2014-01-04

Attila · Expire: 2025-11-18

Spotmatic · Joined: 18 Aug 2008 Posts: 4045 Location: Netherlands

Some 10 years ago, I bought a home-made shutter speed tester working on batteries. I should dig it up; I haven't tried it for many years. It worked well though the last time I used it. I'll try to post a picture of it here.

Arkku · Joined: 28 Feb 2007 Posts: 1416 Location: Helsinki, Finland