Build a shutter tester for Focal Plane shutters - Cheap, Easy & it Works

[Niglyn]

Oh, yeah, I remember seeing that in the write up. It sounds like a nice option.

Perhaps this has been discussed already, but since I haven't read through all of the posts on this forum, nor all of the excellent documentation you provide (I just wanted to start connecting things up as soon as I stumbled across this design) I didn't see it. Are you going to do a future box build around that PCB?

Hi, yes, 3d print files available for a case, works with pcb or expansion board.

Also new today, enclosures for the sensors and Lasers. They can be configured for vertical or horizontal shutters.
(or both if you want to add switches to toggle sensor output)

Also 3d print files for cutting template if using the project box,
bezel for screen, if using project box.

 

[Niglyn]

Newly uploaded to Github,
Full enclosures for both sensors and Lasers.

Sensor enclosure is designed to mount bare sensors without the module.

A small vero-board can be used (which clips into the enclosure) to solder the sensor onto, add the capacitor and wires.
A build guide will be added, waiting for my vero-board to arrive.
(It would probably be ok to use the compete modules, but I have not tried it)

Two flat parts can be used to hold the sensors in place.
Cable entry can be either left or right and cable grips are the T pieces.
Holes are designed for 2mm screws.

Laser enclosure is designed to use bare Lasers, which come with wire tails, so easier to connect than the modules.
Complete modules can be made to fit by bending the laser away from the board.

 

[Gruby104]

Dear Niglyn,
I have a problem with tester. First of all I managed to build frame with horizontal sensors. But I have serious issue with alignment. Despite the fact that I decided to use better lasers that can be seen on the photo. Laser points directly into sensor"lens" but I got information that it's blocked and I don't know what to do. Only if laser points not directly to the lens (for example upper the lens) tester shows that laser can be seen. I tried to adjust shutter speeds of one of my cameras and it's nearly impossible to do it. The results are really inconsistent if it comes to how classic (copy of ) Leica shutter works (hard slit). And now I struggle to make even one reading cause of this error. And there were two times when whole tester crashed. Please help. Maybe I am wrong with something but this is my first project with ESP 32 ever (I learn how to repair old film cameras )Cheers!
Photos: https://drive.google.com/drive/folders/1JYlTyYjR-IeasL_swDsMy5MfkJgbeMOl

 

[Niglyn]

Dear Niglyn,
I have a problem with tester. First of all I managed to build frame with horizontal sensors. But I have serious issue with alignment. Despite the fact that I decided to use better lasers that can be seen on the photo. Laser points directly into sensor"lens" but I got information that it's blocked and I don't know what to do. Only if laser points not directly to the lens (for example upper the lens) tester shows that laser can be seen. I tried to adjust shutter speeds of one of my cameras and it's nearly impossible to do it. The results are really inconsistent if it comes to how classic (copy of ) Leica shutter works (hard slit). And now I struggle to make even one reading cause of this error. And there were two times when whole tester crashed. Please help. Maybe I am wrong with something but this is my first project with ESP 32 ever (I learn how to repair old film cameras )Cheers!
Photos: https://drive.google.com/drive/folders/1JYlTyYjR-IeasL_swDsMy5MfkJgbeMOl

Hi, firstly be careful using powerful lasers, you do not want to burn holes in the shutter!

I have never had the firmware in The Shutter Tester crash.

Suggest you temporarily remove the mask in front of the sensors..

Please go into the alignment screen and check the sensors show 'Seen' when the laser is pointing at them & 'blocked' when you cover the Laser.

If it is working backwards, then there is the 'new' old' option to reverse the settings. This is documented in the user guide.

Sensors can be triggered by ambient light and also if near computer or tv screen. I built a little hood to cover mine from sides, rear and above. Just out of a piece of cardboard.

Please let me know how you get on.

 

[Gruby104]

Hi, firstly be careful using powerful lasers, you do not want to burn holes in the shutter!

I have never had the firmware in The Shutter Tester crash.

Suggest you temporarily remove the mask in front of the sensors..

Please go into the alignment screen and check the sensors show 'Seen' when the laser is pointing at them & 'blocked' when you cover the Laser.

If it is working backwards, then there is the 'new' old' option to reverse the settings. This is documented in the user guide.

Sensors can be triggered by ambient light and also if near computer or tv screen. I built a little hood to cover mine from sides, rear and above. Just out of a piece of cardboard.

Please let me know how you get on.

I will make more tests tomorrow but I discovered that adjusting lasers really matters. I mean every laser has got "focusing" knob. I made the tests and when laser is focused on the shutter it increase up to 1/450 from 1/350 on the same shutter speed without any adjustment. Also I managed to get 1/14200 of a sec. on Zorki 4 haha I got an error and I couldn't make a photo but it was something like SS<micoss100 or something like that. Then tester restarted

 

[Niglyn]

I will make more tests tomorrow but I discovered that adjusting lasers really matters. I mean every laser has got "focusing" knob. I made the tests and when laser is focused on the shutter it increase up to 1/450 from 1/350 on the same shutter speed without any adjustment. Also I managed to get 1/14200 of a sec. on Zorki 4 haha I got an error and I couldn't make a photo but it was something like SS<micoss100 or something like that. Then tester restarted

Hi, the distance is only 6 inches, so powerful lasers are not required and you could be overloading the voltage regulator on the ESP32.

The error you see is The Shutter Tester reporting that a very fast speed was recorded. Often caused by PWM lights, computer or tv screen.

 

[Gruby104]

Hi, the distance is only 6 inches, so powerful lasers are not required and you could be overloading the voltage regulator on the ESP32.

The error you see is The Shutter Tester reporting that a very fast speed was recorded. Often caused by PWM lights, computer or tv screen.

So my plan for today would be attaching lasers that I got from AliExpress and making cover from ambient light. Btw the lasers that I use now have power of 5mW. I will test it on every camera that I have.

 

[Gruby104]

Dear Niglyn, I managed to create cover for sensors with light proof black foil. With that I separated every single sensor from reflections from other lasers and also I added big cover that covers whole sensors compartment. Furthermore, I glued sensors to their connectors protect against unwanted movement. As I said earlier I used stock lasers and issue with alignment has gone. I tested It with Pentax S3 that has much more stable shutter than Leica type. Despite the fact that there was no capping effect (I was checking the shutter in slow motion) and I had perfect alignment I have errors that 2 sensor cannot be recorded. Also it seems that tester is unstable cause I managed to regulate the shutter to almost 1/1000 and after the shut down the tester and and then has shown 1/500 maybe this is my mistake but I get you know that. I haven't tested all my cameras yet and I'll do it in the next week. Cheers! Photo: https://drive.google.com/drive/folders/1JYlTyYjR-IeasL_swDsMy5MfkJgbeMOl

 

[Niglyn]

Dear Niglyn, I managed to create cover for sensors with light proof black foil. With that I separated every single sensor from reflections from other lasers and also I added big cover that covers whole sensors compartment. Furthermore, I glued sensors to their connectors protect against unwanted movement. As I said earlier I used stock lasers and issue with alignment has gone. I tested It with Pentax S3 that has much more stable shutter than Leica type. Despite the fact that there was no capping effect (I was checking the shutter in slow motion) and I had perfect alignment I have errors that 2 sensor cannot be recorded. Also it seems that tester is unstable cause I managed to regulate the shutter to almost 1/1000 and after the shut down the tester and and then has shown 1/500 maybe this is my mistake but I get you know that. I haven't tested all my cameras yet and I'll do it in the next week. Cheers! Photo: https://drive.google.com/drive/folders/1JYlTyYjR-IeasL_swDsMy5MfkJgbeMOl

31, 32 , 23

Hi, please look at page 31 / 32 of this thread to ensure you do not have one of the faully batch of sensor modules.

If you look at page 23, there is a photo showing a simple cardboard cover I made for the sensors.
The rear of the sensors also needs to be covered if they are affected by stray light.
My tester sits right in front of the compute monitor, hence my little cardboard cover.

I would suggest you dispense with your mask.

 

[Gruby104]

31, 32 , 23

Hi, please look at page 31 / 32 of this thread to ensure you do not have one of the faully batch of sensor modules.

If you look at page 23, there is a photo showing a simple cardboard cover I made for the sensors.
The rear of the sensors also needs to be covered if they are affected by stray light.
My tester sits right in front of the compute monitor, hence my little cardboard cover.

I would suggest you dispense with your mask.

I covered the whole sensors. I put out the main cover in order to show you how did I separate sensors. I also have bought new sensors from my local seller and I will try once again

 

[crispernaki]

Hi Niglyn:

I got things up and running yesterday! Once again, let me say you have put together an awesome project. I'll send better pictures once I get around to bolting everything down, but here is the current state:

I'm going to build up a second set of emitter/detector boxes to have both horizontal and vertical capabilities available. The emitter/detector boxes you designed work really well! Aiming the lasers wasn't too hard, but I've been pondering how you might make aiming even easier. Have you considered leveraging the fact you are 3D printing, and how you could implement a ball/socket in place for each laser? This way, the lasers can be in place and wired up, with the enclosure being closed and mounted, prior to aiming. The lasers could then be aimed precisely from the outside, and locked in place with a bit of glue.

I can sketch something up if my description isn't very clear.

Also, it would be helpful to indicate on the inside of enclosure print which detector is #1 and which is #3. Although I do see the comment above about reversing the order in the settings. COOL!

I still haven't read through all the posting in this thread, so forgive me if this has been discussed already!

 

[Niglyn]

Hi Niglyn:

I got things up and running yesterday! Once again, let me say you have put together an awesome project. I'll send better pictures once I get around to bolting everything down, but here is the current state:

View attachment 416827

Hi crispernaki, thanks for your positive feedback and comments. I always welcome these as it helps to improve the product.


Only yesterday, I uploaded a better box for the Lasers. Rather than the Lasers being mounted in the lid, they are now in the main box.
I am also using Lasers that are loose and not mounted on a module. These are much easier to work with.
(Details of these will be included in the new build guide - coming soon).

With both boxes mounted on a base, it is now much easier to align the Lasers.
Still a bit fiddly I know, a blob of hot-glue on the Laser and having to hold it until it cools.

I think a ball & socket might be too fiddly to print & being behind the Laser, multiplies any movement in use.
It might be better to have the front of the Laser fixed and the rear moved up/down/left/right.

I can see how to do this, however it starts to get complex & I don't think 3d printers are accurate enough, so extra precision drilling & reaming would be required. Hopefully the new box and hot-glue will work satisfactorily.


It is possible to use one box for both horizontal and vertical sensors (theoretically). A 2 pole two way toggle switch is required, fitted to the sensor box. Then the outer two horizontal and vertical sensors can be wired via the switch, to select horizontal or vertical.

The Lasers could also be switched in the same way. All five could be left on, but I am concerned about the current loading of the 3v regulator on the ESP32 board.

I need to slightly modify the 3d printed T piece for the Sensor box. It works fine for horizontal sensors, but obstructs the legs of the top vertical sensor. It only needs a hole added, for the sensor legs to protrude.

This week, I have been making a new testing frame using the new 3d printable sensor/Laser boxes and photographing each step, which will be used for a new build guide.

I might even make a video ) Now I have got the hang of making them. I have a few tubes, demonstrating my shutter alignment jig & other camera repair tools..

 

[crispernaki]

Great idea with the DPDT switch!

 

[ltbphoto]

OP requested a link to the expansion board that I ordered - I got my ESP32 from the same page. The one I ordered and received was the "expansion board" not "breakout board"

https://www.aliexpress.us/item/3256808232780563.html?spm=a2g0o.order_list.order_list_main.45.4a2918028tULEA&gatewayAdapt=glo2usa

 

[Niglyn]

HI everyone,
Thought I should report on an issue I have found.

Having purchased new sensor modules & separate Lasers (just with wire tails, not soldered to a module) for the new 3d printed sensor & Laser boxes.

However, was getting the S1&M only>2 error


Have finally today, found the resolution.....

The new sensors were struggling to work from 3.3v (they are meant to be 5v devices)

The solution is simple, feed the sensors & Lasers from 5v rather than 3.3v.

Now the purists will say this is wrong. As the ESP32 is 3.3v device, all inputs should be 3.3v.
Whilst this would be best practice, the input pins are 5v tolerant, so are fine.

Build guides will be updated & I can now get on with adding the guides for the new 3d printed sensor & Laser case.

 

[crispernaki]

Interesting. I believe I was seeing the same issue, then.

And the purists are silly.

 

[Niglyn]

Hi, I replied to a DM, with the reply below. Thought it also worth posting here.



Measuring Curtain speed is easy. Service manuals do not specify an actual speed, but the time taken for the blind to move a certain distance, often 32mm.
So all one needs are two sensors spaced at 32mm, or what ever spacing is specified in the camera service manual.

When the blind passes the first sensor and it is 'seen' the internal uS clock time is recorded.
As the blind passes the second sensor and it is 'seen' and again the internal uS clock time is recorded.
Subtract the larger number from the smaller and you have the time it took the curtain to travel 32mm.

As we are measuring just one blind and it is going in one direction, there is no issue with hysteresis or sensor size.

For the second blind, it is exactly the same, but the 'blocked' value is used.

Curtain bounce is also very easy. Just monitor how many times sensor 2 is 'seen'. Any more than once & it is bounce.
If it is never seen, it is capping.

However measuring 'shutter speed' is far more complex.
First of all the term 'shutter speed' is a misnomer. The shutter moves at the same speed regardless of setting. It is the time that the second curtain is released in relation to the first and the resulting slit width between the two curtains, which determines exposure, incorrectly called 'shutter speed'.

Because two curtains are used, the issue of sensor width comes into play.
A sensor of 2mm diameter may report 'seen' when 0.5mm has been uncovered by the curtain.
However, when the second curtain then travels to block the light, the sensor may have to be covered 1.5mm before it looses enough light and switches to 'blocked'.
Thus the sensor has added 1mm to the slot width.

At flash sync speed, say 1/30s the blinds will have a slit of 36mm so ratio of sensor to slit is 1:36
Now move to 1/500s and the slit is 2.25mm. Giving a ratio of 1:2.25
So here the sensor adds almost 50% to the reading.

So, we are trying to measure the width of a single photon, which has no width, with a sensor with a width of 1mm.
This issue has been known about since the invention of the focal plane shutter. It is documented and discussed in journals going back to the 1960s.
Kosmofoto.com has some good articles on the subject & I have posted the links in the Photrio thread.

This is what the ISO has to say on the subject, ISO 516:2009

1 Scope​

This document provides a uniform basis for determining the exposure times for all types of shutters used in still cameras and contains suitable definitions of the terms used.
It specifies the exposure-time markings for all types of shutters and their tolerances.
The characteristics of all types of mechanical shutters, which are mounted in still cameras and affect the control of exposure, motion-stopping ability and synchronization with a photoflash light source are also defined.
The tolerances specified are the target values for the shutter performance that can be expected to give good results. They are not intended for application as a general inspection standard in controlling the performance of mechanical shutters, since tolerances may vary with the feature and price class of camera tested.
Test methods are described for routine manufacturing testing and quality control. These test methods require access to the focal plane of the camera and can therefore not be applied to assembled digital still cameras.

2 Normative references​

There are no normative references in this document.

3 Terms and definitions​

For the purposes of this document, the following terms and definitions apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:

• — ISO Online browsing platform: available at https://www.iso.org/obp
• — IEC Electropedia: available at http://www.electropedia.org/

3.1
front shutter
any shutter in the vicinity of the lens
Note 1 to entry: The front shutter can be in front of, behind or between the lens elements and can consist of rotating discs, rotating slats, sliding blades, oscillating blades, etc. Programmed shutters are also included.
Note 2 to entry: The common characteristic for the front shutter is that the entire picture area is exposed almost simultaneously.
Note 3 to entry: When the shutter and diaphragm are located too far apart, both exposure and shutter speed may vary at different points in the picture area.

3.2
focal-plane shutter
any shutter in the vicinity of the focal plane
Note 1 to entry: The focal-plane shutter can consist of fixed or variable slit curtains, rotating discs, sliding blades, etc.
Note 2 to entry: The essential feature of the focal-plane shutter is that the picture area is exposed incrementally, in such a way that the time required to expose the entire picture area is greater than the exposure time of any one point.

3.3
effective time
te
best measure of the amount of light falling on the picture area
Note 1 to entry: Effective time is defined by the following formula:

Note 2 to entry: At any point on the picture area, te is generally the same for the entire picture area for front shutters when vignetting is not severe. For focal-plane shutters, te will vary with w and vc. The formula in Note 1 to entry can be approximated with the below formula for convenience in measurement:

(focal plane shutter)

The formula in Note 2 to entry can only be applied under the condition of w ≥ ds/A.

3.4
exposure time
teo
effective time measured at the centre of the picture area

3.5
total time
to
the time for which any given point in the picture area is exposed to light
Note 1 to entry: At any point on the picture area, to is generally the same, or almost, on the entire picture area for front shutters.
Note 2 to entry: For a focal-plane shutter, however, to is dependent on w, A, ds and Vc. The curtain displacement to completely expose one point becomes w + ds /A, which can be converted to to, if the velocity is known, using the following formula:

Note 3 to entry: This formula can be inexact in the presence of vignetting.
See Figure 1.
Figure 1 — Total time for a focal-plane shutter

a	Taking lens.
b	Curtain.
c	Focal lens.

3.6
shutter efficiency
η
ratio of effective time to total time
Note 1 to entry: The shutter efficiency is given by:

3.7
fluctuation of exposure time
p
the value of pis determined by the following formula

where x and σ are the mean and standard deviation of the values of five successive measurements

3.8
ratio of two adjacent exposure times
q
ratio of the mean values of two adjacent shutter speed settings obtained from values of five successive measurements
Note 1 to entry: The ratio is expressed by the following formula:

Note 2 to entry: teo (n) and teo (n + 1) are the exposure times of two adjacent shutter speed settings represented by (n) and (n + 1).

3.9
non-uniformity of exposure
r
characteristic which may be found during any single exposure due to lack of coincidence with the principal plane (front shutter) or to variations in curtain velocity or slit width (focal-plane shutters)
Note 1 to entry: Such non-uniformity is expressed as the ratio of the maximum and minimum effective time found by exploring the picture area, and is derived from the following formula:

3.10
overall time
T
elapsed time for exposure of all points in the entire picture area
Note 1 to entry: For front shutters, T = to.

3.11
photoflash synchronization delay time
td
time interval from the initial closing of the shutter synchronization contacts to the moment at which the shutter element moves to the specified position (see 5.2)
Note 1 to entry: For details of ignition circuits of synchronizers, see ISO 10330.

3.12
X contact
synchronization contact for an electronic flash unit
Note 1 to entry: The contact closes while the shutter is fully opened to enable reception of the reflected light from the object through the aperture of the lens or for total illumination of the camera aperture. The X contact can sometimes be used for the M or MF class of photoflash lamp at the slower shutter speeds.

3.13
M contact
synchronization contact for M class of photoflash lamp

3.14
FP contact
synchronization contact for FP class of photoflash lamp
Note 1 to entry: This contact is provided only in the focal plane shutter and can be used for M or MF class of photoflash lamp at the slower shutter speeds.

 

[Niglyn]

Hi, was in communication with a user who suggested a graphical display would be good.
They had hooked up an oscilloscope to The Shutter Tester sensors & shared photos of it.

Not to be outdone, I have added an oscilloscope screen to The Shutter Tester.
To access this screen, press the Blue encoder button, to toggle though the different screen layouts.
The firmware can be found in the beta folder.

The photo below shows a test at 1/30s.
Each trace (blue, green, yellow) shows each of the three sensors.
When the trace drops, the sensor is 'seen' and then raises when it is blocked by the second curtain.

The trace runs from left-to right, in the same way an oscilloscope trace does.
(viewed from the rear, curtains will travel right-to-left)

We can see the yellow trace, sensor 2, is low (seen) whilst blue trace, sensor 1 is also low (seen),
thus showing both sensor 1 and 2 are uncovered by the blinds at the same time, thus flash sync is available.

The bottom photo shows the same camera with a shutter speed set to 1/500s
The seen/blocked of each sensor can be seen. The never overlap, so only part of the film is exposed as the slit moves across the gate.
We can also see the space between S1 and SM is greater than that between SM and S2.
Indicating the second curtain is moving faster and catching up with curtain 2.

 

[koraks]

Hi, I replied to a DM, with the reply below. Thought it also worth posting here.

Hi @Niglyn, would you be OK with it if we duplicated the two posts above into a new thread and made it a sticky on this forum? We believe it would be a great resource for future reference. The potential 'downside' is that you may keep receiving questions about it for years to come - but of course, there's no obligation for you to keep responding if at some point you decide it's no longer in your interest to do so. Thanks in advance!

 

[Niglyn]

Hi @Niglyn, would you be OK with it if we duplicated the two posts above into a new thread and made it a sticky on this forum? We believe it would be a great resource for future reference. The potential 'downside' is that you may keep receiving questions about it for years to come - but of course, there's no obligation for you to keep responding if at some point you decide it's no longer in your interest to do so. Thanks in advance!

Hi @koraks , yes no problem.

 

[crispernaki]

Hi, was in communication with a user who suggested a graphical display would be good.
They had hooked up an oscilloscope to The Shutter Tester sensors & shared photos of it.

Not to be outdone, I have added an oscilloscope screen to The Shutter Tester.
To access this screen, press the Blue encoder button, to toggle though the different screen layouts.
The firmware can be found in the beta folder.

The photo below shows a test at 1/30s.
Each trace (blue, green, yellow) shows each of the three sensors.
When the trace drops, the sensor is 'seen' and then raises when it is blocked by the second curtain.

The trace runs from left-to right, in the same way an oscilloscope trace does.
(viewed from the rear, curtains will travel right-to-left)

We can see the yellow trace, sensor 2, is low (seen) whilst blue trace, sensor 1 is also low (seen),
thus showing both sensor 1 and 2 are uncovered by the blinds at the same time, thus flash sync is available.

The bottom photo shows the same camera with a shutter speed set to 1/500s
The seen/blocked of each sensor can be seen. The never overlap, so only part of the film is exposed as the slit moves across the gate.
We can also see the space between S1 and SM is greater than that between SM and S2.
Indicating the second curtain is moving faster and catching up with curtain 2.




View attachment 417741


View attachment 417742

Yay!

 

[crispernaki]

Hi koraks, yes no problem.

I was going to try and suggest the same kind of thing. There are currently 36 pages in this thread. They contain a lot of excellent historical content, but it's a lot to sift through when coming at this from scratch. This is such a wonderful project, and there is a lot to grasp. The summaries in github are wonderful, but if most folks are like me, they are going to look at both github and the thread. Having said all that, it's also tricky to have the same information listed in multiple places. Nothing worth doing is ever easy!

 

[Niglyn]

Yay!

Have added flash indication to the scope display, so one can visually see where the flash fires in relation to the curtain openings.
uploaded to Github in the beta directory.

Currently one has to be in the scope screen before a test, else the results will not be valid.

Toggling to the scope screen after a test using the blue button will show erroneous results.
This will be corrected shortly.

 

[Niglyn]

I was going to try and suggest the same kind of thing. There are currently 36 pages in this thread. They contain a lot of excellent historical content, but it's a lot to sift through when coming at this from scratch. This is such a wonderful project, and there is a lot to grasp. The summaries in github are wonderful, but if most folks are like me, they are going to look at both github and the thread. Having said all that, it's also tricky to have the same information listed in multiple places. Nothing worth doing is ever easy!

Hi, In theory, everything to build The Shutter Tester should be on Github.

The thread here is for discussion & help. I often post updates here, so those having already built the tester can update if they choose.

I do need to go though the build documentation to make it a bit clearer, as there are now options for for expansion board or pcb,
project box or 3d printed case etc.

As always, I welcome suggestions & comments.

 

[Niglyn]

Below are a couple of photos showing the new oscilloscope screen with flash trace.

The white vertical line indicates when the flash is fired in relation to the curtain timing.

The first photo shows the line appearing when both sensor one, blue top trace) and sensor two, bottom yellow trace are both seen, indicating that the blinds are fully open, exposing the whole film-gate to the flash.

The second photo shows the line after sensor 1, blue line, has been blocked, caused by the second curtain starting to close.