Anyone tried to replace the CdS photosensor?

[kl122002]

I have few old light meters and cameras that have dead CdS sensor. I am thinking to give it a try to replace it with a new one.

Anyone got experience in replacing the sensor?

 

[koraks]

Are these even being made anymore?

You could replace a CdS sensor with something else, but you'd have to make some additional adjustments as well to get a similar response, and the spectral response of a more modern silicon sensor will definitely not be the same (which is likely to be a good thing).

Personally, I wouldn't bother and just use a handheld meter.

 

[Paul Howell]

I dont know of a source, maybe someone has tired to 3D print? But, what are the chances that a blue silicon sensor from a modern digital camera can be somehow fitted and matched to voltage of an older camera?

 

[Bill Burk]

The best course these days is to get a parts camera. Do the work necessary to bring one of the cameras as close to working as possible and then take the CdS cells and maybe the entire circuit from the camera that has the best electronics.

Using the same camera model for parts gives the best looking repair.

You can sometimes use generic CdS but the result is always a little janky.

 

[4season]

Anyone got experience in replacing the sensor?

Not yet, but I have a Konica S3 which is otherwise ready for use, save that the original CdS photocell has shorted, resulting in low resistance under all lighting conditions. A part search through USA sources including Mouser and Digikey revealed a small number of available parts with peak sensitivity within the visible light spectrum, with a physical size small enough to fit the existing space, IIRC it was literally just one or two obvious possibilities.

 

[Chan Tran]

Are these even being made anymore?

You could replace a CdS sensor with something else, but you'd have to make some additional adjustments as well to get a similar response, and the spectral response of a more modern silicon sensor will definitely not be the same (which is likely to be a good thing).

Personally, I wouldn't bother and just use a handheld meter.

CdS sensor (photo resistor) are being made plentiful today. The problem is to find out the characteristic of the original cell to find the right replacement. Replacing with different type of cell means to redesign the entire circuitry (which may not be a bad thing as it can be made more accurate). Personally I just use the camera without the meter. I have quite a few hand held meter but I wouldn't use them either. It's more fun to use camera without meter.

 

[koraks]

CdS sensor (photo resistor) are being made plentiful today.

RoHS restricts the use of Cd in electronics. In the EU, these devices can no longer be sold, legally.

 

[Chan Tran]

RoHS restricts the use of Cd in electronics. In the EU, these devices can no longer be sold, legally.

I just check the RS Online website and they list about 10 different models. On Amazon they have a bunch. I still see them used in many cheap/modern devices.

 

[Paul Howell]

RoHS restricts the use of Cd in electronics. In the EU, these devices can no longer be sold, legally.

Why is that?

 

[koraks]

Like I said, Cd-containing electronics are banned in principle within the EU. There are likely exemptions (i.e. no feasible alternative) and I suppose distributors cater to parties who are entitled to these exemptions. Also, this is EU-legislation and apparently, the US hasn't moved in this direction yet. Similar to the dichromate situation.

I still see them used in many cheap/modern devices.

I haven't seen them used in permitted devices in the EU of recent manufacture. On the other hand, there's a plethora of light sensors available that fill this functional gap, but they're not snap-in replacements for older cameras. Hence my initial remark.

Why is that?

Probably because there's very little that can realistically be done to prevent the cadmium from ending up in landfills, become airborne as waste is being incinerated etc.

 

[Chan Tran]

Like I said, Cd-containing electronics are banned in principle within the EU. There are likely exemptions (i.e. no feasible alternative) and I suppose distributors cater to parties who are entitled to these exemptions. Also, this is EU-legislation and apparently, the US hasn't moved in this direction yet. Similar to the dichromate situation.



I haven't seen them used in permitted devices in the EU of recent manufacture. On the other hand, there's a plethora of light sensors available that fill this functional gap, but they're not snap-in replacements for older cameras. Hence my initial remark.


Probably because there's very little that can realistically be done to prevent the cadmium from ending up in landfills, become airborne as waste is being incinerated etc.

Simple devices like this

Intermatic Dusk to Dawn Light Control Stem and Swivel Mount Thermal Photocontrol K4721D89 - The Home Depot

This Intermatic Dusk to Dawn Light Control Stem and Swivel Mount Thermal Photocontrol is ideal for the exterior, landscape and pedestrian pathway lighting.

www.homedepot.com

and these

You can tell by the look of the sensor that they are CdS.

 

[koraks]

Simple devices like this

Yes, I see. They're technically illegal to sell here in Europe.

 

[kl122002]

Are these even being made anymore?

You could replace a CdS sensor with something else, but you'd have to make some additional adjustments as well to get a similar response, and the spectral response of a more modern silicon sensor will definitely not be the same (which is likely to be a good thing).

Personally, I wouldn't bother and just use a handheld meter.

I dont know of a source, maybe someone has tired to 3D print? But, what are the chances that a blue silicon sensor from a modern digital camera can be somehow fitted and matched to voltage of an older camera?

Yea there still places that make them and they are sold online like replies here. I think it is just because other countries has no restrictions so the production just shifted to these places.

 

[Paul Howell]

Yea there still places that make them and they are sold online like replies here. I think it is just because other countries has no restrictions so the production just shifted to these places.

So, how would one know which to buy, or buy one that matches the voltage the camera is designed for. Say 625 battery then figure out a way to integrate into the camera? What about size?

 

[qqphot]

I've been starting to think about a battery powered modern light sensor solution to replace a selenium cell in old meters that used them, purely for aesthetic purposes. I have a couple of the old Minox meters that I'd like to see usable again.

What i'd really like to do is a digital solution with a modern very precise sensor and a microcontroller a2d channel generating the right current for exactly the meter deflection needed across the full range of light values. It could be made very small, and if I could find somewhere to stuff it inside the housing the main problem remaining would be having to open the meter to replace the battery.

 

[kl122002]

Tha

So, how would one know which to buy, or buy one that matches the voltage the camera is designed for. Say 625 battery then figure out a way to integrate into the camera? What about size?

That is the exact area I am studying now. So far I think the resistors ( variable) might also need to be replaced with a new one, rather than using the old one. The older circuit is designed to work with 1.35v or mercury based circuit.

 

[Chan Tran]

To build a meter circuit I would use this sensor. The hard part is how to attach wire to it. It's so tiny

APDS-9007

Ambient Light Sensor

www.broadcom.com

 

[koraks]

The hard part is how to attach wire to it.

You don't. It's far easier to make a PCB for it and then solder the part onto it in a reflow oven or with a hotplate. I use the latter.

Btw, this is the spectral sensitivity of the sensor:

To me, it looks lacking in red and especially in blue. Compare it to CdS (which isn't optimal to begin with!):

Thinking along the lines of @qqphot's idea: AMS have a pretty wide range of sensors, many of which have several color channels you could balance at your heart's content. If there's going to be a microcontroller, such sensors (which interface easily with I2C) are easy to implement and the A/D stuff is handled far better than you could realistically DIY (been there, done that!)

 

[Chan Tran]

You don't. It's far easier to make a PCB for it and then solder the part onto it in a reflow oven or with a hotplate. I use the latter.

Btw, this is the spectral sensitivity of the sensor:
View attachment 359456
To me, it looks lacking in red and especially in blue. Compare it to CdS (which isn't optimal to begin with!):
View attachment 359457

Thinking along the lines of @qqphot's idea: AMS have a pretty wide range of sensors, many of which have several color channels you could balance at your heart's content. If there's going to be a microcontroller, such sensors (which interface easily with I2C) are easy to implement and the A/D stuff is handled far better than you could realistically DIY (been there, done that!)

My primary problem is that I don't know how to inteface with an i2C device and most of the AMS are i2C devices. There are some with linear output but the linear output is not good for light meter ( actually linear is good for light meter but just not good for exposure meter) because you have to use a log amp. I like this Avago sensor because the output is logarithmic.

 

[koraks]

Ok, too bad; good luck.

PS: it is technically possible to solder leads to a QFN package, but I do NOT recommend it. It's mechanically totally unstable and the odds of going batsh*t crazy trying to do it are very real. I know...

That's a TCA8418 keypad controller; it's about 4x4mm, the pad pitch is 0.5mm. Yours will be more difficult to handle since it's more fragile and a little smaller, even though the pitch seems a little bigger.

 

[Chan Tran]

Ok, too bad; good luck.

PS: it is technically possible to solder leads to a QFN package, but I do NOT recommend it. It's mechanically totally unstable and the odds of going batsh*t crazy trying to do it are very real. I know...
View attachment 359458
That's a TCA8418 keypad controller; it's about 4x4mm, the pad pitch is 0.5mm. Yours will be more difficult to handle since it's more fragile and a little smaller, even though the pitch seems a little bigger.

I know that using a PCB is a good idea but I can't find premade PCB. I have never etched one although I was thinking about it.

 

[koraks]

Etching a PCB isn't too hard, but these days, you can get PCB's commercially made to your design quite affordably. Check places like JLCPCB and PCBWay. Their quality is impossible to beat with DIY techniques.

 

[Bill Burk]

I know that using a PCB is a good idea but I can't find premade PCB. I have never etched one although I was thinking about it.

Just take any old board or flex connector having the right pitch (my latest project was taken from a Zip drive)

When I hack-replaced CdS in an ES-II the calibration was off the scale and I had to insert resistors to get to a calibration range. Then to make me madder... The center weight pattern wasn’t the same, I think it went skewed to a side. I think I switched the cells in from a different body and maybe I swapped the board too

 

[ic-racer]

I have a few meter projects on-going. These involve using photodiodes and trying to drive galvanometers with Arduino MWM output and using tiny photovoltaics for driving galvanometers on their own.

To find out more about tiny sensors I analyzed the sensor in a popular tiny meter. The sensor has 4 feet and is intended for surface mount, however, it is mounted straddeling the PC board, with contacts on both sides of the boards. Not an easy task to remove and replace. It got pretty ugly replacing it; getting solder blobs on all 4 contacts. I actually feel there is no need for anything this small. All CDS cells were larger and certainly selenium cells are much larger, so there should not be a need for something so tiny.

 

[ic-racer]

DigiKey lists about 70 CDS cells: