Leica M3 shutter speed measurement

[DaveM3]

Hi everyone,
I built myself a simple IR emitter/detector rig to measure my shutter speeds some time ago. The shutter open time is measured simply using an oscilloscope, so a pulse is observed as the shutter fires. I then measure the pulse width. The detector is placed right behind the open back of the camera and I usually test without a lens fitted. This is accurate for slow times but I am beginning to wonder if this is flawed for the higher speeds where the shutter time relies on precise timing of the leading and trailing curtain to produce a slit which traverses across the film plane.

After a recent service, the measurements on high speed settings appeared slow, 1/1000 measures more like 1/370 i.e. 1.4 stops slow although it had previously tested at 1/700, so it has definitely changed. The readings are very consistent as I would expect since the camera is well serviced.

I am beginning to wonder though if this simple method of testing is flawed - before I go guns blazing back to the service person.

I do question the testing/calibration done though because the shutter is most definitely 1 stop out at 1/8.

measurements v dial
1 2 4 8 15 30 50 60 125 250 500 1000

1.0 1.8 2.9 4.3 13.9 20.2 45.5 47.8 108.7 196.1 303.0 370.4

Cheers,
D

 

[x-ray]

Do you have a film camera with an electronically controlled shutter. It would be more likely to be accurate than a 60 year old shutter, serviced or not.

 

[Erik van Straten]

I would say that only 1/1000 is far off, too slow. But this is common on old Leica's. Usually it is around 1/750.

Far more important is the evenness of the exposure. You do not want that one side of the image is too bright or too dark.

Erik.

 

[a.noctilux]

How about the size of the detector ?
Narrow slit takes some time to travel across the detector.

Without lens, the detector would receive the light not parallel to it.
The rise/fall of the pulse is to be considered also, half or top of pulse ?.

So the method is not really valid as is.

 

[DaveM3]

How about the size of the detector ?
Narrow slit takes some time to travel across the detector.

Without lens, the detector would receive the light not parallel to it.
The rise/fall of the pulse is to be considered also, half or top of pulse ?.

So the method is not really valid as is.

Yes that's exactly what I thought. Now the question is, how to measure it properly (and easily). I have heard mention of the CRT trick but I don't have an old TV - perhaps I can find one somewhere.

 

[yanchep_mike]

Hi Dave,
i ave build a few shutter speed testers, i came to the conclusion the best way is to use a focused laser diode (cheap on ebay) and the tiny beam is interrupted by the curtain movement very quickly and that time is not a real problem. With that setup i can do reliable up to 1/4000, the smaller the laser beam the better.
cheers mike

 

[DaveM3]

Hi Dave,
i ave build a few shutter speed testers, i came to the conclusion the best way is to use a focused laser diode (cheap on ebay) and the tiny beam is interrupted by the curtain movement very quickly and that time is not a real problem. With that setup i can do reliable up to 1/4000, the smaller the laser beam the better.
cheers mike

Hi Mike, that's a great idea. Would you care to share the device part numbers or description? are we talking visible light and simple photo-diode detector?
Thanks
Dave

 

[yanchep_mike]

Dave, just run of the mill visible light laser diode search for something like laser diode red in copper or so, i tried various and they all work.
This is my last one

I started with CRO and then used a MCU to make it easier.
Not difficult, good luck.
mike

 

[DaveM3]

That's pretty amazing!

 

[Rob-F]

How about placing a narrow vertical slot behind the shutter, between it and the detector. You could take three measurements, with the slot positioned left, right, and center to see the uniformity of exposure. It might let you take the measurement with your existing hardware.

 

[Rob-F]

I would say that only 1/1000 is far off, too slow. But this is common on old Leica's. Usually it is around 1/750.

Far more important is the evenness of the exposure. You do not want that one side of the image is too bright or too dark.

Erik.

I think the Leica shutter speed tolerance is around 15% or 20% or so, isn't it? 1/30 looks to be off by maybe 30% or so and 1/15 by perhaps 25%. 1/50 and 1/2 are within 10%, and 1 second is right on.

 

[Erik van Straten]

I think the Leica shutter speed tolerance is around 15% or 20% or so, isn't it? 1/30 looks to be off by maybe 30% or so and 1/15 by perhaps 25%. 1/50 and 1/2 are within 10%, and 1 second is right on.

It also depends on what you are doing. If you make slides, 30% off is significant, but for B+W it is of no importance.

Erik.

 

[DaveM3]

How about placing a narrow vertical slot behind the shutter, between it and the detector. You could take three measurements, with the slot positioned left, right, and center to see the uniformity of exposure. It might let you take the measurement with your existing hardware.

Also not a bad idea, I might try this.

 

[monopix]

Putting a slot over the sensor is a good idea but it will reduce your output. Maybe not a problem if you're using a 'scope - just increase the sensitivity. But using a laser as well will give you a higher output due to the concentrated source. I use a cheapo add-on laser gunsight as it has a useful clamping block attached. And a pinhole over the sensor made from ally foil. The pinhole works for both vertical and horizontal shutters.

Sent from my SM-G925F using Tapatalk

 

[DaveM3]

Measurements with the setup shown in the photo (1/1000 setting):
Left: 2.5 ms
Middle 2.6 ms
Right: 2.4 ms

Now to mention it to the serviceperson (it was CLA'd recently)

 

[rodinal]

You can get some ideas from this thread :

https://www.rangefinderforum.com/forums/showthread.php?t=135793&highlight=arduino+shutter