Repairing / calibrating an unknown negative film densitometer

[adrifromhh]

Hi guys,

this is my first post on this forum. I obviously know about it because I have been watching Daves videos for years now Greetings from Hamburg, Germany!

So here's the problem:
Besides the electronics thing, I am a hobby photographer. I shoot on B/W film and process the negatives and prints in my own darkroom. If you want to take this hobby to the next level, or really get the best results from your negatives, you have to make some measurements to know the exact development times for your location (water values can be very different even within the same town). I won't get into detail about this procedure here. The important fact ist, that you have to measure the relative density of test exposures on a negative film. Therefore you expose a series of pictures with known steps in which the film density should increase. Then you "null" your densitometer on a blank part of the film, because the base of different films has a certain density by itself. Now that the blank part of the film reads a "D" value of zero, you go on with your measurements of the series of exposures to get their density, relative to the blanc film.

As I am a poor student, I got a big old 70s densitometer for cheap off eBay. Generally it seems to be working. The problem is, the reading is completely off. I found no single bit of information on the internet, and the manufacturer doesn't seem to exist anymore, so I hope that with the help of this forum we might find out how the electronics work, to be able to calibrate the thing. The circuits aren't too complicated. In the worst case, I would even reverse engineer the schematics to get help on understanding them, but hopefully that's not needed.

What I found out technically:

- The lamp puts light through a small hole, through the film to be measured, onto a photodiode that flaps onto the film from the top.
- The left Knob on the front switches the unit on.
- The right pot dims the lamp, to null the measurement. Let's call it "Pot 1"

- The main problem: Even without anything in the lightbeam, I cannot bring the indicator all the way to the left to read zero.
- A trimpot, which is accessible from the back, ("Pot 2" - 100 Ohm, two wires connected) seems to adjust the amplification of the photodiode. When I turn it all the way to the right (0 Ohm), the indicator nearly reaches zero with the lamp all the way on. You have to darken the beam a lot to make the indicator go up. (little amplification) When Pot 2 is full left (~ 100 Ohm), the indicator doesn't nearly reach zero but goes up very quickly when the beam is darkened.
- I desoldered Pot 2 and replaced it with a higher value one. I managed to get very little amplification and the indicator to zero with pretty high kOhm values. It seems to me that the problem must be somewhere else, because it's impossible to get a useful setting with the given pot.
- There is a crackling / sizzling noise in the electronics when turned on, which alters when the lamp is being dimmed. The board with the huge heatsink gets very hot, so this might even be normal.
- Can the photodiode have gone bad after the years, so you need much more amplification now?

Edit: I forgot to mention that after we manage to get useful readings, I would get a transmission step wedge to really adjust the unit.


I'm just an electronics hobbyist, so I hope the infos that I provided are helpful and someone might be able to help me.

Best,
Adrian

[DaJMasta]

I'm not familiar with the machine, but maybe these will help:
Are you sure it's a photodiode and not a CDS cell or some other light sensor?
Have you replaced the lamp?  Incandescent bulbs definitely age and produce less light over time.
Have you measured any dividing resistors attached to the light sensor and compared them to their markings?  How about ones in series with the adjustment pot?


I don't think crackling noise is a good sign, but a hum is probably fine... is it really crackling?  Do you have a scope to measure power supply fluctuations?  Have you checked any caps for abnormally low resistances across them?

Those would be my first places to consider, at least.

[adrifromhh]

Thanks for your quick reply!
I didn't open the metal case with the detector yet. I'll try to do that tomorrow and see what's actually in there.

Yes I have a scope. I'll try to get some measurements.

The reason why I didn't consider an aged lamp yet is that it was impossible to correct that with the trimpot. I would have thought that the adjustable range should include such drifting of aged lamps. I can try to find a lamp. Let's hope they are easy to find and not too expensive.

[adrifromhh]

The lamp does not have any markings at all and looks very exotic.

I opened the detector. It also looks very interesting and old fashioned. There are two wires coming from it.

I have to go to bed now (It's past midnight here in Germany) but tomorrow I will try to hunt down the noise and measure the power rails.

[Kleinstein]

The sound would be the first thing to fix. Also check the EMF filter cap (gray part near transformer) at the input. If old, there is real chance for them to send out some of the magic smoke - the sound could be a first warning. For a first test one could get a way without it.

The detector really looks strange. Not like a photo-diode at all. However there could also be an optical filter to limit the wavelengths to a reasonable range. Just the lamp and an normal Si diode would give way to much weight to the NIR range, especially if the lamp is dimmed.

The color of the light could give a rough idea if the voltage is about right.

[tronde]

On the PCB with the large heatsink:

The collector connection of 2N3055 seems quite burnt. Is the electrical connection between the nut and PCB OK?

The 470uF/50V capactor (rubber facing up) seems to be leaking.

[adrifromhh]

The detector really looks strange. Not like a photo-diode at all.

The color of the light could give a rough idea if the voltage is about right.

Yes I also thought this large plate could be a filter, hiding the real detector underneath. I don't think I can get further without destroying the assembly.

The color of the light looks pretty white to me when fully on.

To the rest of the hints: I'll try to test all of that today. Thanks so far!

edit:
- The 470 µF cap on the large board is broken
- The connection to the 2N3055 is OK.

[voltz]

Hi,
This sounds like a power supply issue. Backed up by your faulty cap findings.
These units are very old, guessing 70's. Nearly all the electrolytic capacitors would have dried out by now. Consider testing/replacing them all.

If the power supply is of too low voltage or contains too much ripple, the photo detector amp wont operate properly and you wont be able to get a proper null reading. Check the supply rails for correct voltage and low ripple.

The crackling could be one of the electrolytic caps or a dry solder joint to PCB. They can do that when faulty.

The only other idea i can think of is making sure the optical path from lamp to detector is clean from dirt and dust.

Good luck.

[adrifromhh]

I did some measurements:

Image - Transformer output - Nominal voltage
93   -   5/6   -   15V
94   -   5/6   -   15V rectif.
95   -   7/8   -   20V
96   -   7/8   -   20V rectif.
97   -   9/10   -   20V
98   -   9/10   -   20V rectif.
99   -   11/12   -   10V
100   -   11/12   -   10V rectif.

I'm in a 230V country, so mains goes to transformer input 1/4, 2/3 are shorted.
The rectified values were measured across the decoupling cap after the rectifier. I found it strange that output 9/10 breaks down to 7.8V after the rectifier.

Anyway, I will order all caps and probably post again after I replaced them. Should I only swap the electrolytic ones?

[PedroDaGr8]

Replace all the electrolytics, film capacitors should be ok unless they show signs of being bad (burn marks, discoloration, etc.).Those styles of electrolytic caps are very well known to go bad in European gear. Not quite as bad as the bakelite body electrolytics but I've seen many bad ones with that style.

Sent from my LGLS992 using Tapatalk

[adrifromhh]

It might be a heat absorbing filter to protect the film negative. Enlargers had them. This lamp may not be very hot so it is just a possibility.

No, because the plate is on the detector side, after the film

[DaJMasta]

Could you try to figure out what the detector is by just testing a couple bits with a multimeter?  A CDS cell would have a changing resistance with brightness, a photodiode would show up with a diode tester with it's polarity correct, etc.  If you have some magnification you may be able to see in the bottom too (though it may be difficult without bright light and a way to move the arm).

I think the problem is more likely to be the PSU as well, but you may be able to poke around and get a good idea of what the sensor is in the mean time.

[Kleinstein]

Usually it is only the electrolytic caps the go bad. The only other cap one might consider replacing it the EMF one, next to the transformer. If this one makes noise, better remove it.

Checking the detector may be also important. A CdS cell could be aged - up to the point of needing replacement.

[voltz]

Yes, electrolytics only.

[adrifromhh]

A CDS cell would have a changing resistance with brightness, a photodiode would show up with a diode tester with it's polarity correct, etc.  If you have some magnification you may be able to see in the bottom too (though it may be difficult without bright light and a way to move the arm).

The detector shows up as a diode with a multimeter AND has a changing resistance with brightness.

I can not look through the bottom plate, even with bright light.

I ordered the caps today and a 2N3055, just in case because it gets very hot.

[adrifromhh]

Hey guys,

sometimes it will take me a little while to respond, because I don't always have time for electronics while studying

So I replaced all caps and the 2N3055, but it didn't really solve the issue. I didn't have time to do measurements of the power rails yet, but I still can't adjust the meter to read zero and at the same time go up, when darkening the lightbeam.

I managed to open the "detector" and was very surprised: The big round plate that you can see in the photos, is actually the photodetector itself. There are two wires directly attached to it. One from the back, one from the edge.

[helius]

Looks a great deal like a selenium cell.

[Koray]

That is a selenium cell. Under bright daylight it should generate 0.3-0.4V. Selenium cells are notorious for degradation in time. It is hard to find a working one today.

K.

[Kleinstein]

With a a likely bad detector, this calls for a replacement. Chances are one can replace the selenium cell with a Si photo-diode, preferable one with an anti IR filter or human response filter (wie BPW21).

[adrifromhh]

Chances are one can replace the selenium cell with a Si photo-diode, preferable one with an anti IR filter or human response filter (wie BPW21).

You think thats possible? I would be very thankful if you would help me find a replacement part. If that would be useful, I could make a drawing of at least the schematics connected directly around the photodiode.

Selenium cells are notorious for degradation in time. It is hard to find a working one today.

I read a few topics on selenium cells now and it seems that not all of them degrade over time. In many cameras, the problem is with the corroded contacs. In my case, this seems very unlikely, because the wires are soldered directly on the cell.

[tronde]

... and the manufacturer doesn't seem to exist anymore,..

A long shot, but I would try to contact them.

http://www.firmen-finden.com/firma/gunther-engelbrecht-gmbh-egling

[adrifromhh]

A long shot, but I would try to contact them.

http://www.firmen-finden.com/firma/gunther-engelbrecht-gmbh-egling

Hey, thanks for the advice! I actually sent them a letter, which was returned but I didn't have that phone number by then. I'll try that one tomorrow.

[Kleinstein]

As a replacement sensor one could try a simple photodiode first. Electrical it should be reasonable similar to a selenium cell - just two pins and thus two directions to try. If this works out, one might go for a more expensive diode (e.g. BPW21) with filter.

I would expect something like a transimpedance amplifier around the detector. So one side of the detector connected to ground or a fixed voltage and the other towards the - input of an OP, with a resistor+parallel capacitor in feedback.