Basic Electronics Help--Building a Shutter Speed Tester

bvy · Jan 10, 2010

I want to build Dead Link Removed -- it's a quick and dirty little device to test the shutter speed of some of the cheaper film cameras I have. The precision I'm after kind of stands in contrast to the whole lo-fi photography trend, but humor me. If nothing else, it will be a fun project for my son (the wannabe roboticist) and me to work on.

I'm no electrician and the instructions are sketchy, so I need some help picking the right parts. Here they are with my questions:

phototransistor (will infrared work?)
5K ohm resistor (half watt or quarter watt? 4.7K seems standard.)
mono plug
switch
battery holder (2 AA? What about 9V? How do I know which type will work?)
project case

Also, do I need a particular gauge solder?

Thanks for the help!

tkamiya · Jan 10, 2010

I would imagine, the type of photo transistor must match the type of light source you'd be using.

As far as resister goes, 1/4 watts 4.7Kohm is sufficient.

Based on the picture, the power source is TWO AA batteries. So, you'd need a holder for two double-A batteries.

You'd need a resin core solder for electronics work. You can buy a small roll or ones that come in small tube at your local Radio Shack. Don't worry about gauge.

Other than the photo TR, you can buy everything at a Radio Shack.

tkamiya · Jan 10, 2010

I wouldn't put too much trust in something like this. Following this instructions, one is trying to measure a rise in DC voltage with something (sound card) that doesn't respond to DC voltage. Long story short, it might give you some indication but you have no ways to know if the results are accurate.

pnance · Jan 10, 2010

Most any resistor of 5K ohm will work, if max amps flow (assuming the photo transistor is a short) the current would only be .0006 amp , and wattage used would be 0.0018 w (1/18000) - I'm not aware of a resistor with that small of a wattage, so about any 5k you can get will work. The resister in photo appears to be 1/4w or smaller.

Am skeptical about any phototransistor working however. The phototransistor chosen should have a rise time shorter than the shortest desired measurements. To measure 1/1000 s one would need at least 1/2000 minimum rise time.

There is nothing wrong with testing different phototransistors however, using a camera with a known good shutter (consistent negatives, using all combinations of shutter speed/aperature) will allow you to test the tester.

Of course infrared will work, however, you need to insure that infrared exists in the light source. Probably not from a visible light LED.

Using 9v instead of 2 AAs (3v) would triple the current (probably not a problem) but why not use the 2 AAs as designed. Using a potentiometer instead of the fixed resistor would allow you to adjust to get the output in the desired range, but is a needless complication, use the 2 AAs.

Solder? If you can solder with it, it is the correct gauge. However, do NOT use acid core solder, resin core only. When doing really small work, I tend to use smaller solder gauge than when soldering large items. Using large gauge on small items can tend to oversolder the joints, using small gauge on large items can use an awful lot of small solder.

pnance · Jan 10, 2010

Actually there is nothing wrong with this device, there are many sites recommending this type of system. The only problem is getting a proper phototransistor and the software - Audacity - which used to cost $$ but now is free from SourceForge. I've been meaning to build one myself, but unfortunately I've been lazy, other than aquiring a few phototransistors.

Laurent · Jan 10, 2010

I have something much simpler and, I think, more reliable...

It uses a photo transistor and a few resistors, and plugs on the parallel port of a PC. I'm not the designer of the thing, and don't know how it works on a "modern" PC (I still have a brand new 386 to run the program under DOS).
It's reported to be usable under XP, but I've never tried it (I no longer have a PC under XP with a Parallel port).

To make matters worse (unless you speak French), the documentation is in French (it was written at the University of Strasbourg).

Anyway, here's a link. http://www-physique.u-strasbg.fr/~udp/logiciel/chpchtml/index.htm

I've planned to try the same under Linux, will keep you posted if I have any success.

Ian Grant · Jan 10, 2010

tkamiya said:
I wouldn't put too much trust in something like this. Following this instructions, one is trying to measure a rise in DC voltage with something (sound card) that doesn't respond to DC voltage. Long story short, it might give you some indication but you have no ways to know if the results are accurate.

You can test the slower speeds another way and compare the accuracy to the suggested device. This will work best with Leaf chutters rather than Focal plane, Just record the shutter being tripped and measure the time the shutter was open in Audacity or Sound Forge.

This is what the sound of a 1933 Compur looks like on 1 second:

So if that was compared to a similar light output file in Audacity you'd be able to see how accurate the two are to each other up to about 1/25th / 1/20th second. After that the light system will be the more accurate.

Shutter speed tolerance is supposed to be ± 15% of the marked speed for acceptable accuracy.

Ian

Steve Smith · Jan 10, 2010

An infra red photo transistor would be fine. The voltage it's running from isn't too important so 2 x AA as specified is o.k. as would be a 9 volt battery.

The value of the resistor isnt critical either. Anything from 1K to 47K will be good enough.

Steve.

tkamiya · Jan 10, 2010

No doubt something like this will display something, and it's entirely up to the individual to use or not use devices like this. But, the fact remains, many, if not most, sound cards do not react to DC voltage. When DC voltage is applied, such as square waves like this with rise and fall time, it will react to the AC component of the rising edge and falling edge. How the device will measure such input largely depends on the sound card themselves. There is no telling what the rise and fall time is and how it relates to the actual signal without having some reference. The relative error will be magnified in shorter shutter speeds.

The article ends, the author wants to CLA his shutter based on this result. I wouldn't come to this conclusion.

Comparing with known-good device will give a reference point. But, still, unless the sample itself was calibrated by something known and recently, one is checking the relative accuracy, not the absolute accuracy.

bvy · Feb 20, 2010

Okay, I'm revisiting this project and have another question.

I'm looking at electronics at a place called Mouser.com. The number of phototransistors they have is baffling. Relative to the device I'm looking to build, are there some specs I should be looking at specifically -- related perhaps to light/dark current, rise/fall time and peak wavelength?

Thanks for any help at all...

mr rusty · Feb 21, 2010

You don't need audacity

Reaper www.reaper.fm is uncrippled and fully functional shareware. Willdo just the same job.

bvy · Feb 21, 2010

Thank you. Actually I already have Audacity, but I'll check it out.

Still looking for help with the phototransistor (post #10)...

Steve Smith · Feb 21, 2010

Try this one: Dead Link Removed

And look at Circuit 1 at the bottom of the data sheet: http://www.optekinc.com/datasheets/OP506A.pdf

Steve.

tkamiya · Feb 21, 2010

I would make sure the following is true:

1) it has sensitivity at the wave length of your chosen light source
2) it has the reaction time fast enough for your application
3) it has the current carrying capacity necessary for your application
4) it is in a shape that can easily be handled (ie. not a surface mount type)
5) is an NPN type instead of PNP type
If you happen to have PNP type, you just need to reverse the C and E on your schematic.

Since your application is not anything ultra fast, super sensitive, or high current, I would imagine bulk of what is being offered at Mouser would fit your application. But, going beyond making this general statement, it would involve going over data sheet for each which I have no access.

bernard_L · Feb 21, 2010

The basic circuit that you refer to is OK except for one oversight: the loop for the current closes through the sound card input, which is normally high impedance (47k_ohm is standard). Unde those conditions, the response of the circuit might be slower than optimum. You should move the 5k (actually 4.7k wil be easier to find) from where it is on the schematic (replace it by a direct connection), to being across (in parallel with) the two wires that go into the sound card. Other than that, any photo transistor would do provided (as pointed out by tkamiya, it is NPN, as drawn on the schematic; but it is probably hard to find one that is /not/ NPN.

Regarding the more recent comment also by tkamiya
I wouldn't put too much trust in something like this. Following this instructions, one is trying to measure a rise in DC voltage with something (sound card) that doesn't respond to DC voltage. Long story short, it might give you some indication but you have no ways to know if the results are accurate.
That is not a major issue. True, sound cards (at least those I've seen) suppress the DC component. So, the signal you will see in Audacity (or other) will not be a faithful representation of light intensity versus time (especially visible for long shutter times). But that is not the goal. You will still see clearly two spikes: one at shutter opening, the other at shutter closure. Intensities are not accurate, but times are.
Feel free to send me an Email if tha does not make sense. I could also send you screenshots taht illustrate what I wrote.

Chan Tran · Feb 21, 2010

Instead of using the sound card, I used an old oscilloscope (old analog scopes are fine as the phosphor does stay lit long enough for you to see the trace well and even a lowly 20Mhz units are plenty fast for the job). Using an ociloscope eliminate the problem of dc voltage input and response time. Even a slowest 20Mhz scople response time is plenty fast enough for the job.
The problem is still with the photototransitor. I found that they are fast enough to measure a 1/125 time with less than 10% error but with higher shutter speed there is problem with phototransitor response time.
I used a prebuilt sensor from Banner Engineering. It has a response time of 50microsec (1/20000) and uses an opposed fiber optic system of about 1mm. It works well but I can't test shutter speed in auto mode.

BetterSense · Feb 23, 2010

I found that they are fast enough to measure a 1/125 time with less than 10% error but with higher shutter speed there is problem with phototransitor response time.

It is NOT a problem with phototransistor response time. Phototransistors have response times of microseconds. Maybe picoseconds! Your circuit as a whole may have a slower response time however.

The DC blocking is not really a problem. Either the sound card has input coupling capacitors or it can handle the short DC spikes. As already mentioned, the test is still perfectly valid because all you care about is the time between the opening and closing spikes. When I made my shutter tester, I used a capacitor so that I did not send DC to the soundcard.

Practically any phototransistor or photodiode will work. Any silicon photodiode is going to have roughly the same IR response; no matter, they still respond to visible light. I use a flashlight. You can get IR photo detector/emitter pairs at radioshack. The detector is a silicon phototransistor that will work fine.

neallofgran · Mar 11, 2010

I have no idea about it so consult good electrician to solve this problem.

Basic Electronics Help--Building a Shutter Speed Tester

Privacy & Transparency

Privacy & Transparency