CdS cell selection (Autocord just an an example)

[River Mantis]

I accidentally got Autocord from my friend for free. He got it almost for free with some other gear and never used it. But it seems to be heavily used by it's previous owner(s). I cleaned and lubricated almost every moving part and managed to make a new torsion spring for the cocking lever which was missing. The last issue is the meter. It's about -3 stops off despite being overpowered by 1.55V cell. So I'd like to know if it's possible to find a replacement from parts still in production. I see some old school 5.5mm cans for example. I can pick one with desirable resistance and replace the whole calibration network in the camera. But I don't know what to deal with response curve. Let's see. The CdS cell response can be approximated with power function (correct me if I'm wrong). I use to think that the ammeter used for indication is inherently linear. The meter supposed to show logarithm of the sensor illumination. But there is no way to approximate logarithm with passive network and therefore rational function. So how do CdS meters work in the first place? What am I missing?

Update: this paper might help http://www.biasedlogic.com/files/on CdS light meters v0.3.pdf

 

[reddesert]

A question you may ask is whether the CdS photoresistor is actually the source of the meter error. These cells rarely fail at the semiconductor; if anything is likely to go bad it's the connection at the terminal. It is possible that the cause of the problem is a bad connection or a dirty trimpot, etc.

The more or less logarithmic sensitivity is approximated in these meters by a simple (downright primitive) resistor network with only a few parts. I don't know the circuit in the Autocord, but you might find this thread about the Yashica-mat meter circuit interesting: https://www.photrio.com/forum/threads/yashica-mat-124g-meter-ciruit-design-and-adjustment.153186/

 

[River Mantis]

A question you may ask is whether the CdS photoresistor is actually the source of the meter error. These cells rarely fail at the semiconductor; if anything is likely to go bad it's the connection at the terminal. It is possible that the cause of the problem is a bad connection or a dirty trimpot, etc.

The more or less logarithmic sensitivity is approximated in these meters by a simple (downright primitive) resistor network with only a few parts. I don't know the circuit in the Autocord, but you might find this thread about the Yashica-mat meter circuit interesting: https://www.photrio.com/forum/threads/yashica-mat-124g-meter-ciruit-design-and-adjustment.153186/

Of course I reversed the schematic and did some math and modelling and this is why I was so confused. I've added a link to a very useful paper. The answer is no, you can't make a logarithmic response but you can get a more or less straight middle part of the curve. The author didn't say how exactly did he come to the solution but the answer is simple: ignore the meter shunt for the beginning and take the whole series resistance which must be equal to the cell's resistance in the middle point of the scale for you to get the right response shape. I just tried to get a random CdS datasheet, approximate cell's response with a power function (I got 35483*Ev^-0.85 from the plot) and did some modelling and it works. I think it's a good starting point.

 

[itsdoable]

That link above is to my Yashicamat disassembly a many years ago.

My understanding is that CdS cell do degrade over time, and it seems to depend on the type/quality of cell. Some of my Gossen / Ikophot cells are fine, some have lost a stop or 2. The Yashica ones seem to all have degraded a fair bit, it's probably due to the quality/cost cutting at the time of the latter Yashicamats.

I re-calibrated the above circuit to use a 3V lithium battery (which has a reasonable flat discharge curve, thou not as flat a the original mercury) and if you select the correct size, it fits into the same battery compartment, But that cell has degraded beyond the range of using a larger voltage drive (it's OK in bright light, but low light response is poor).

I've been looking for a replacement cell. New CdS cells are rare, and co-axial ones used in the yashicamats are just not available. So I was planning to get single cells, and modify the circuit to accommodate them - it's on my long list of to-do's. I don't want to scavenge a cell from a working Gossen, because those meters are arguably better. In the mean time, using a hand held meter is pretty efficient.

As for the logarithmic response, I though you just used a log meter.

 

[reddesert]

One of my other little pursuits is building electronic musical circuits such as guitar effect pedals, and CdS photoresistors are often used in those (paired with an LED for a type of control circuit). They vary somewhat in min and max resistance (dark resistance) and are not precision devices, so may require some fiddling to rebuild a meter. But they are cheap. Here are some photoresistors from Tayda: https://www.taydaelectronics.com/resistors/photoresistors.html and one from Mouser: https://www.mouser.com/ProductDetail/Adafruit/161?qs=GURawfaeGuDpQ5XPTNqKUw== (if this Mouser gibberish link stops working just search for "photoresistor"). You can also get them on ebay.

My limited understanding is that when a CdS cell "goes off" it's more likely to be degradation at the terminal where the wire is connected to the semiconductor, than an actual breakdown of the semiconductor. The CdS material is unlikely to fail internally like a selenium cell does.

The split cell in the Yashicamat is a PITA. If I had to try replacing that, I would put two photoresistors in series to get the center terminal, but fitting the two cells in the space formerly occupied by one would be a challenge.

 

[gone]

The only thing I accidentally got for free, well, a shot and some pills cleared that right up. And you get a Minolta Autocord. It doesn't seem fair, but after reading the manual, it does mention that life would not be fair. Pretty small letters though.

TLRs w/ meters tend to slow me down anyway, so I get usually get better results using a handheld meter. My 'cords took really good photos, that lens on yours is a killer. In fact, I've never even seen a metered Autocord, in the wild or otherwise. Just looked at one online, wow, those look quite different than the ones w/ no meter.

 

[itsdoable]

...The CdS material is unlikely to fail internally like a selenium cell does....

On a similar note. Selenium cells are also a reliable semiconductor material electrically, but the material has a tendency to oxidize in air, so they are covered with a lacquer that seals them. It is this lacquer that degrades and causes the material to stop being a semiconductor. It seems that CdS's may suffer a similar fate, only it is the wire connection that degrades - probably corrosion/oxidation related.

 

[River Mantis]

On a similar note. Selenium cells are also a reliable semiconductor material electrically, but the material has a tendency to oxidize in air, so they are covered with a lacquer that seals them. It is this lacquer that degrades and causes the material to stop being a semiconductor. It seems that CdS's may suffer a similar fate, only it is the wire connection that degrades - probably corrosion/oxidation related.

I'm about to try this one: https://www.tme.eu/bg/en/details/pgm5549-mp/photoresistors/token/
Also TME delivers faster in EU