View Full Version : Old Selenium meter resurrection

03-27-2006, 12:53 PM
I just found an old Selenium meter that used to belong to my granfather. It was kind of semi-functional (the needle would move) but it gave wrong readings. So I decided to repair it and use it with my meterless cameras: like my pinhole, Flexaret, Zorki, Smena, etc.

As I openned the meter there were only 3 important parts: Selenium plate, some resistor and a galvanometer. I removed everything but the galvanometer. In my electronics-junk-box I found a Silicone photo cell from a disassembled solar-powered calculator. I put the photo cell in place of the Selenium plate and added a 500 kohm variable resistor (for later calibration). I soldered it all together in series. The galvanometer needle was quite sensitive, so I fine-tuned it with the variable resistor. I set the resistance just high enough so the needle would reach its maximum at EV 15 (enough range for me).

After I put the box back together I placed a sticker over the scale and at various EV in range about 5 to 15 marked the needle position. I also wrote the EV numbers to those marks and so the meter is usable again. No need for batteries, no Selenium getting old... But also poor low light sensitivity (starting at about EV 5), Anyway, a good way to turn non-working collector's item into a working exposure meter.

(hopefully I'll post also some pictures)

Tom Persinger
03-27-2006, 01:09 PM
very cool! thanks for posting. i hope you have some pics of the process, would like to see it.

welcome to f295.

03-27-2006, 06:14 PM
The original selenium probably wasn't worth much in low light, either -- I've got a Gossen Sixtomat (selenium meter) that was probably made in the early 1960s, and it's only good down to about that level (though correctly, what you measure there is LV; EV is calculated from a combination of aperture and shutter speed, with the two having the same values at ISO 100).

If you want to meter in low light, you need either a silicon photorestistor or a CdS cell, with suitable circuitry to convert resistance to a needle reading (and a very stable voltage source, which is why mercury cells were used before they were banned for environmental reasons).

03-28-2006, 01:20 AM
Thanks. I was also wondering if I could make some real exposure meter, sensitive also in low light. It need not be very acurate, say +/- 1EV would be fine. I need something small (the size of Voigtlander VC II) but for less money.

I already have some ideas, but if anyone here has better ideas about such circuits it would be a big help.

I posted a new thread with this question here: http://www.f295.org/DIYforum/cgi-bin/forum/Blah.pl?b-equip/m-1143527112/

03-28-2006, 05:33 PM
The simplest method is to use a CdS cell, galvanometer movement, and any old 1.5 V battery, plus a calibrating resistor, in a "bridge circuit" -- this circuit essentially just compares the resistance between the (fixed in use) calibrating resistor and that CdS cell, the latter of which is proportional to the light falling on the cell (more light, less resistance). The biggest advantage of a bridge circuit for a meter is that it's insensitive to supply voltage -- ordinary alkaline cells, which change voltage quite a bit from fresh to almost dead, aren't usable in many older meters, but they work fine in a bridge type meter; there's also no calibration change from mercury or zinc-air cells, to alkaline, to silver oxide, even though they vary by 30% in voltage.

Your circuit would run from battery, split to the CdS and calibrating resistance (which is usually a fixed and variable resistor in series), through a pair of additional resistances (one on each leg), and then recombine to return to the battery; the meter movement is wired across the two legs between the first and second resistance on each. The diagram for this looks like two triangles base to base, the meter movement forming the shared bases, the CdS cell and adjusting resistance forming the other legs of one triangle, and two fixed resistances forming the remaining legs of the second triangle -- and then both apices connected to the battery.

Of course, suitable component values are left as an exercise... ;)

03-28-2006, 06:25 PM

Though I know what a bridge circuit is I was not able to find a suitable analog V-meter. I think I'm gonna do some shopping for sensitive V-meters :D

03-29-2006, 03:20 AM
Pictures to the original thread post: Old Selenium meter resurrection

Attached files http://f295.f295.org/uploads/expo1_7489.jpg (http://f295.f295.org/uploads/expo1_7489.jpg)

03-29-2006, 03:26 AM
This is what it looks like inside:

the red and yellow wire are from + of the silicone photocell to the galvanometer
the plastic rectangular thing is a multi-turn variable resistor
the black wire goea to a variable resristor
the green wire goes from variable resistor to the galvanometer
the orange and blue were meant to be a bypass of the variable resistor (there is a switch at the end) to increase sensitivity in lowlight, but it doesn't work at all.
Attached files http://f295.f295.org/uploads/expo_open_1_289.jpg (http://f295.f295.org/uploads/expo_open_1_289.jpg)

03-29-2006, 03:26 AM
another view...

Attached files http://f295.f295.org/uploads/expo_open_2_4718.jpg (http://f295.f295.org/uploads/expo_open_2_4718.jpg)

03-29-2006, 03:28 AM
ant this is from the front: you can see the calculator solar cell behind the honeycomb...

Attached files http://f295.f295.org/uploads/expo_front_1_692.jpg (http://f295.f295.org/uploads/expo_front_1_692.jpg)

03-29-2006, 03:29 AM
and with added lenses on the front.

Attached files http://f295.f295.org/uploads/expo_front_2_3266.jpg (http://f295.f295.org/uploads/expo_front_2_3266.jpg)

03-29-2006, 05:04 PM
Nice work! There are actually sources of selenium cells at this late date, but they're so expensive it's not worth using them even to repair the meters in many classic cameras, much less hand-held meters that can be replaced for $5 plus shipping via eBay.

How much calibration did you have to do to make the silicon cell behave correctly? I can see potential issues with the cell having a different light/voltage curve from the original, beyond just a different sensitivity level, as well as different color sensitivity (selenium was excellent because it matched panchromatic B&W almost perfectly, so you could meter through a filter instead of having to remember or read a table for filter factors).

03-30-2006, 04:58 AM
> ImageMaker:

1. first I calibrated the range by adjusting the variable resistor. I set it so that the needle would reach maximum at about LV 15

2. then I would walk aroundour house with a SLR on my neck and the meter in my hand, metering every wall around ;D I simply put a sticker on the original stripes (the stripes between the needle window and the DIN values) and drew my own lines (compare the picture of my meter and the attached picture below, but the spinning dial is in a strange position in the picture below ::) imagine turning it 90o counter-clockwise).

That's all. and even though I tried to calibrate it for range of LV 5-15 in the min. and max. values it is quite difficult to see difference between neighboring lines... so the range is rather a LV 6-14

BTW: when I compare the readings to my other SLR (not the one I used for calibration) they are always in the +/-1EV range, more often in the +/- 0.5EV. Attached files http://f295.f295.org/uploads/meter_metra_3781.jpg (http://f295.f295.org/uploads/meter_metra_3781.jpg)

03-31-2006, 11:19 AM
Aaah, I see it now. It looks like you're using the meter as it was originally set up: set DIN speed at the arrow via the marks on the sector (or your sticket, after recalibration), then read the EV (true EV, already compensated for film speed) from the red scale at the red arrow, or directly read exposure from the aperture and speed scales (including cine frame rates at an assumed 180 degree shutter angle, as is common on meters from this era of the 1950s and 1960s).

I'd be tempted to alter the calibration to make LV 16 (bright sun on a beach or snow) accessible and sacrifice the lower light levels if necessary, since you're more likely to shoot wihout flash in full sun than at LV 4-5. You could also mask off part of the cell to expand the scale, I think -- you'd get less meter deflection at high light, which would let you reduce the calibrating resistance and likely gain on the low end of the scale. Ideally, you should get maximum range with the calibrating resistance about the same as that of the meter coil, but you probably won't get a silicon cell to that level.

06-30-2006, 02:52 AM
My selenium lightmeter can meter values from 3 EV, but values from 3 to 5 ev can be measured, if you measure light putting your meter near object and pointing it to camera. There is special mode for that.

05-22-2013, 06:23 AM
Nice work, and I am happy I have found this thread since I am going to do exactly the same with the cell from a solar powered calculator. Moreover, your meter is Czech product ("Metra" factory in Blansko town, near Brno, Czech Republic). I think it was produced until early 80's. I used to have that one, just cannot remember where it ended...