Anyone replaced a selenium cell?

[Muggins]

I have a Fed Mikron and a Zorki 11 that both have dead selenium cells. Both bought for their looks, TBH, but awkward to use with a single speed of 1/30th without the meter.

I note that Polar Bear Camera are offering solar cells to replace selenium cells - Solar Cell for Light Metering Selenium Cell Replacement [FREE SAMPLE] - but (a) electronics is Not My Thing and (b) these are rectangular cells of uncertain size, and both cameras have a curved cell round the lens.

There is stuff out there on the intertubes about this but if someone here has done it, there's a lot more opportunity to discuss the ins and outs! Is this a reasonably easy repair, or such a drag that I should just stick to the 1/30th?

Thanks!

 

[JeffS7444]

So far I've done a selenium > silicon upgrade on a Ricoh Auto Half and most recently, a Minoltina-S. It's an easy-ish job. You don't need to replicate the size or shape of the original part. In the case of the Ricoh, the replacement was a fraction of the size, yet I seemed to get pretty good results with no other circuit alterations.

But before getting too deeply immersed in the project, might want to verify that you've got a functional galvanometer.

 

[Muggins]

So far I've done a selenium > silicon upgrade on a Ricoh Auto Half and most recently, a Minoltina-S. It's an easy-ish job. You don't need to replicate the size or shape of the original part. In the case of the Ricoh, the replacement was a fraction of the size, yet I seemed to get pretty good results with no other circuit alterations.

But before getting too deeply immersed in the project, might want to verify that you've got a functional galvanometer.

No, it's just the way I walk.

More seriously, will a multimeter do the job? If so yes, if not no. That gives you an idea of my level of understanding of the function of tbe sparks in a camera, though I'm also thinking that I can't make them less functional than they already are so I might add well have a go.

 

[JeffS7444]

will a multimeter do the job?

Sure, set it to measure resistance, then see if the galvanometer needle twitches. The voltage and current from the multimeter ought to be sufficiently low to do this with no risk to the delicate galvanometer coils.

 

[johnf04]

Some years ago, I set up an experiment to test if solar cells from calculators (I bought a heap of them, cheap) would replace the selenium cell in a Zenit E. I found that galvanometers from different Zenits had different characteristics - they had a series resistor to match cells to meters, and these resistors were of different values. I also found that the output curve from a selenium cell was different to the output curve of the silicon solar cell. I still have the experimental setup, and I may revisit this one day.

The setup to measure the silicon cell voltage (The cell is mounted where the selenium cell would go)


The connections for the camera's galvanometer and the silicon cell extended so a variable resistor can be put in the circuit.

 

[BernardL]

I also found that the output curve from a selenium cell was different to the output curve of the silicon solar cell.

+1
Voltage versus light is different.
Voltage versus current is different.
Possibly some combination of series and parallel resistors could allow nominal performance at both low and high light levels.
Plus, the spectral sensitivity is different: silicon is more red and IR sensitive.

 

[JeffS7444]

I'd welcome more detailed information on doing this properly, with the caveat that it's not clear to me how I'd gauge performance characteristics of my stash of silicon PV cells or an expired original part except in a very general way.

In the case of the Minotina-S, my silicon replacement is about the same size as the original selenium, and as expected, my readings under the sorts of daylit conditions I'd likely encounter were off by at least two stops. So I experimented with improvised neutral density filters (silvery gray metallized plastic cut from an antistatic bag) to reduce it's sensitivity until I arrived at something which "seemed pretty close" when used in combination with the camera's stock egg-crate grille and lens array.

Ricoh Auto Half was even easier: At the time, I was still learning to hand-cut crystalline PV cells, and wound up using a chunk maybe 1/3rd the size of the original. By eyeball guesstimates, I thought the camera's response to changing light seemed within the ballpark, so I went with it. I've since run a few rolls of color negative film through the camera, and have been pleased with the results.

 

[Valkir1987]

I'm experimenting with Osram phototransistors at the moment. When exposed to light, they output current like photodiodes and even leds. They can only be connected in parallel for this purpose. The advantage is that they take up very little space.

Gossen Bisix 2, replaced the cell with two phototransisors. I think it needs one or two more to give proper readings in lower light conditions. The problem is the linearity (or non linearity) compared to an original selenium cell.