Minolta XG-M: Adjusting the shutter

[Andreas Thaler]

At the end of last year I tried to solve a shutter problem on my Minolta XG-M. It got stuck when winding and the shutter blocked:

Minolta XG-M: Shutter problem resolved but be careful

I checked the functionality of my XG-M to see whether replacing the four electrolytic capacitors was successful in the long term, see https://www.photrio.com/forum/threads/minolta-xg-m-replacing-the-electrolytic-capacitors.203831/ After a few releases, the mirror remained up and the LED...

www.photrio.com

I was able to solve the problem, but I also worked on the two adjusting screws on the bottom of the camera for adjusting the shutter tension. These are supposed to be responsible for setting the running speed of the two shutter rollers.

So I assume that the shutter now needs to be readjusted. Specifically the running speed and possibly also the shutter magnet.

I will find out how to do that.

Now that I have a way of measuring the shutter with the Reveni Labs Camera Tester, developed by PHOTRIO member @Reveni-matt this project makes sense.

This will be an interesting journey into the depths of an electronically controlled camera. I am very excited to see what happens.

Stay tuned

 

[ic-racer]

Many times in an electronic camera, the first curtain speed (transit time across the frame) is specified*. After setting that value, the shutter gap is determined by releasing the second curtain electronically.

*For example 11 milliseconds to travel 32mm.

 

[Andreas Thaler]

Many times in an electronic camera, the first curtain speed (transit time across the frame) is specified*. After setting that value, the shutter gap is determined by releasing the second curtain electronically.

*For example 11 milliseconds to travel 32mm.

Thank you!

I played with these two locking wheels.

I think they determine the pulling force of the wound rubber band for the two shutter blinds.

That would mean that both have to be brought to the reference speed (hopefully) specified in the service manual.

The distance between the blinds behind each other is a matter for the electronics (magnet - setting via potentiometer).

Right?

 

[ic-racer]

For best results the whole mechanism would be disassembled and cleaned, but for just experimenting, determine which wheel is for the opening curtain (see diagram below). Then use your new shutter tester to adjust the wheel so your transit time is 11 milliseconds for the first curtain. You might have to measure the space between the sensors on the shutter tester to see if it is 32mm.

Usually it is a metal spring in a tube that pulls the curtains across. The first curtain adjuster would be the one toward the back of the camera.

Now that you have your shutter tester, you might want to get any old camera with cloth shutter and replace the curtains (re-gluing the curtain to the drums, re-timing the gears and setting the spring tension). Aftrer you do that you will know everything about how cloth shutters work!

 

[Andreas Thaler]

Thanks, I'm already thinking about it.

Do you know how to adjust a Copal S shutter or similar shutters? They're usually hidden in the camera.

 

[Andreas Thaler]

View attachment 374674

OK, then there is the additional complication of getting there.

The best way to practice this is on a Canon T90

 

[ic-racer]

I may have had the arrows wrong on the last image.

 

[ic-racer]

Sometimes you can peek in and see the adjustment wheels...

 

[Andreas Thaler]

Thank you for the briefing, that has activated me.

I will familiarize myself with the topic using the XG-M and the Reveni Labs Camera Tester will accompany me.

I'm looking forward to it

 

[Andreas Thaler]

Curtain Travel Times

From an issue of the SPT Journal for the Minolta XG-M:

SPT Journal: March-April 1983

Society of Photo-Technologists Journal

learncamerarepair.com

This information applies to older camera testers, whose sensor distance is 32 mm and is converted to 36 mm for the Reveni Labs Camera Tester.

This means that the curtain travel time, which I will set as the target value, is a little longer, namely around 12 ms:

Thanks to @Reveni-matt and @Gregory_Nolan for the explanation!

 

[ic-racer]

I had to do something similar to find the actual distance between the sensors of the Rolleiflex tester. Rollei only gives the travel time, not the distance.

Rollei EKZM Shutter Tester

Apparently, when Rollei developed the focal plane shutter for the SL66, they had focal plane shutter tester called the EKZM. When they later developed the SL35 focal plane shutter 35mm camera, there was an adapter plate to use the EKZM shutter tester on the SL35 camera. The service...

www.photrio.com

 

[Andreas Thaler]

As I have read, the shutter rollers speed up as they move. This means that the set shutter speed becomes shorter across the film window.

To compensate, the second roller would then have to move more slowly.

This means that the same moving speed for both rollers cannot be the goal.

Or does the shutter have a compensation mechanism?

 

[Reveni-matt]

As I have read, the shutter rollers speed up as they move. This means that the set shutter speed becomes shorter across the film window.

To compensate, the second roller would then have to move more slowly.

This means that the same moving speed for both rollers cannot be the goal.

Or does the shutter have a compensation mechanism?

Do you have more info on that?

 

[Andreas Thaler]

Do you have more info on that?

2.14 PERFORMANCE FACTORS FOR MOVING-SLIT SHUTTERS The two basic types of moving-slit shutters, horizontally and vertically traveling, share some problems. The first one has to do with the acceleration of the slit as it races across the film aperture. Since the effective exposure time is given by the slit width divided by the time it takes the slit to travel its own width (a simplification, as we'll see later), how can these shutters make even exposures if they speed up as they travel? The answer: by widening the slit during the travel

Norman Goldberg, CAMERA TECHNOLOGY
The Dark Side of the Lens, Academic Press: San Diego, 1992, Kindle edition

 

[Reveni-matt]

Norman Goldberg, CAMERA TECHNOLOGY
The Dark Side of the Lens, Academic Press: San Diego, 1992, Kindle edition

Thanks, I have this book, I'll read the section tonight.

 

[Reveni-matt]

Norman Goldberg, CAMERA TECHNOLOGY
The Dark Side of the Lens, Academic Press: San Diego, 1992, Kindle edition

I reread the section multiple times, I don't see a mention of diffetent total travel times. The velocities change (we know this, it's clear in the Tester's velocity view) but it doesn't say they don't travel in the same period.

"It takes each curtain about 18ms to travel the 36mm length of the format, but the velocity is not constant..." Goldberg says.

Velocity varies? Yes. Slit width varies? Yes. Travel time? I think the same. Effective shutter speed and exposure also remains the same because the slit width × velocity = same exposure time.

1mm slit x 10mm/ms = "10"

10mm slit x 1mm/ms = "10"

5mm slit x 2mm/ms = "10"

Etc

 

[ic-racer]

As I have read, the shutter rollers speed up as they move. This means that the set shutter speed becomes shorter across the film window.

To compensate, the second roller would then have to move more slowly.

This means that the same moving speed for both rollers cannot be the goal.

Or does the shutter have a compensation mechanism?

You can see it here. Explained below in post #18,

 

[Reveni-matt]

You can see it here. If the first and second curtain had the same travel time, then the slit width would be the same across the film plane.

They do have the same travel time, just offset.


I did the math in Excel to find the distance travelled for two curtains, the 1st released 1 second before the 2nd. The distance between them steadily grows by the acceleration (I used 9.81m/s/s here). multplied by the time difference. Curtain 1 will always keep moving ahead of curtain 2 if they always keep accelerating forever. And from the table we can see, every distance point is reached by curtain 2 exactly 1 second after curtain 1, so the exposure time is always 1 second for any location on the film plane.

Curtain 1 travel time should equal Curtain 2 travel time. If Curtain 1 equals Curtain 2, then exposure at any location on the film plane will be the same. The exposure time will be equal to the time delay from Curtain 1 release to Curtain 2 release.

 

[Andreas Thaler]

Fascinating!

Thank you both @Reveni-matt @ic-racer

What does this mean for practical use, what should I keep in mind when trying to set up my XG-M?

From what has been said, I conclude that it is a matter of getting both shutter rollers to the same speed.

Then I look at how the electronics determine the shutter speed, i.e. whether, for example, 1/125 s set on the shutter speed button is actually 1/125 s.

Now I would like to know which times I should set.

Is it enough to set the fastest shutter speed, e.g. 1/1000 s, and a slower one, e.g. 1/4 s? Will all the other times then be set too?

 

[ic-racer]

They do have the same travel time, just offset.


I did the math in Excel to find the distance travelled for two curtains, the 1st released 1 second before the 2nd. The distance between them steadily grows by the acceleration (I used 9.81m/s/s here). multplied by the time difference. Curtain 1 will always keep moving ahead of curtain 2 if they always keep accelerating forever. And from the table we can see, every distance point is reached by curtain 2 exactly 1 second after curtain 1, so the exposure time is always 1 second for any location on the film plane.

Curtain 1 travel time should equal Curtain 2 travel time. If Curtain 1 equals Curtain 2, then exposure at any location on the film plane will be the same. The exposure time will be equal to the time delay from Curtain 1 release to Curtain 2 release.

Nice! Thank you for figuring it out!

I should have know that as I did some automotive race engineering in the 1980s. Two equal cars leaving a corner, one after another, will get farther apart down the straight doing the same lap times.

 

[Andreas Thaler]

Don't forget this and the other projects with the Reveni Labs Camera Tester, I've taken on too many projects again

 

[Andreas Thaler]

Curtain-travel times and shutter speeds: First use of the Reveni Labs Camera Tester

After having familiarized myself with the Reveni Labs Camera Tester, today I want to adjust the horizontal focal plane shutter on my Minolta XG-M.

The speed of the two curtains will not be correct, as I had already played around with the adjustment wheels for the curtain tensions.

I take the curtain-travel times from the SPT Journal:

SPT Journal: March-April 1983

Society of Photo-Technologists Journal

learncamerarepair.com

For the tester, I convert it from 32 mm to 36 mm sensor distance:

=> Target curtain-travel time is 12.4 ms.

I measure the shutter speeds per sensor (L, M, R) and the curtain-travel times (C1, C2) for 1/1000 second.

1/60 s

1/4 s


The tester optionally displays the shutter speeds as fractions, which I am more familiar with.

Since I did not photograph the corresponding displays, I convert the shutter speeds myself and use the values of the middle sensor (M):

=> 1/4.4 s, 1/54 s and 1/685 s


The result shows that the values for 1/4 s and 1/60 s are close to the target values. At 1/1000 s the shutter speed is clearly too long although the curtain-travel time is ok.

Since I have no experience and would like to gain some here, I will try to set the curtain travel time on the camera for 1/60 s in the first step.

The aim is to get both curtains to have a curtain-travel time of 12.4 ms. Currently they are running too slowly.


Making settings on the XG-M

To reach the two adjustment wheels for the curtain tensions, I remove the base plate and desolder the plate with the battery chamber.

Then I connect the disconnected power supply to the camera via cables.

I have to be careful not to short-circuit the plus and minus. There isn't much space on the circuit board.

Here are the two ratchet gears for setting the curtain tensions.

To adjust the lower one, I have to loosen a crocodile clip.

Since both curtains move too slowly, more tension is needed here, which is transferred to the curtains via rubber bands. So I have to turn the ratchet gears until they are locked in place.

Of course, I turn the wheels in the wrong direction by releasing the catches. Now the shutter won't open at all.

The patience game begins

---

First conclusion

• Working with the Revenu Camera Labs Tester is the easier part.
• Setting up the XG-M requires some effort. But that doesn't surprise me, as the Minolta Repair Manual describes some setting procedures where test devices are connected and resistors are soldered into the circuit. The bread and butter of service technicians in the early 1980s
• Interpreting the measured values requires experience. It is about optimization, not about achieving 100%.
• If the curtain travel times are correct, the shutter speeds should also be OK. These are generated electronically and I didn't mess around with the potentiometers responsible for this when I replaced all the electrolytic capacitors. The factory settings should therefore be retained.
• This is not for the impatient. So we'll have dinner now and continue tomorrow

 

[monopix]

"If the curtain travel times are correct, the shutter speeds should also be OK."

You can't guarantee that. The mechanics can also cause errors in the shutter speed, especially at high speeds.

 

[Andreas Thaler]

"If the curtain travel times are correct, the shutter speeds should also be OK."

You can't guarantee that. The mechanics can also cause errors in the shutter speed, especially at high speeds.

Thanks, so in the next step I have to play around with the potentiometers.

I wish I didn't have to because this is probably micrometer work

 

[ic-racer]

Yes, usually counter-clockwise tightenes the springs in most cameras!!

Also, note that your RIGHT side, at 1.25ms, is not going to be affected much by the curtain tension. The readout is telling you the second curtain is being released 0.25ms to late.

Also, note that the upper time limit for 1/1000 (per ISO 516:1999(E)) is 1.59ms, so the camera is OK by ISO standards.

So, if you are working with tolerences smaller than what ISO specifies, you might want to check the calibration your tester with some camera you have that you purchased new.