Arithmetic or Geometric mean?

Arithmetic, surely (the resistance is proportional to the illumination, not exponential). More important, though, is the COLOR of the illumination, since photon energy determines the interactions that free those conduction electrons. It's possible that the minimum for a given sample is much of that range, depending on the light color.

There's an order of magnitude between the response to IR and blue, for CdS sensors.

Hard to say. CdS (and especially CdSSe) is a truly horrible material, whose conductivity can vary 5:1 depending on its previous history. Its main virtue is a gigantic photoconductive gain.(1) Long ago, it was considered superior in camera light meters because its spectral sensitivity curve is sharply peaked in the blue, making it a good match for orthochromatic B&W film.

Unlike photodiodes, photoconductors are unipolar--there's no PN junction, just a chunk of doped amorphous or polycrystalline semiconductor that functions as a very sensitive light-dependent resistor (LDR).

Cheers

Phil Hobbs

(1) When light hits a photoconductor, it produces carrier pairs that drift apart under the applied E field. If the carrier lifetime (tau) is long, as it is in PbS (milliseconds iirc), it may be much longer than the transit time t_T.

The photoconductor then exhibits gain, so that the charge transferred per detection event is

q = e tau/t_T,

which in most PbS detectors is a big number (thousands).(2) You'd think that this would be great--a poor man's photomultiplier. In some respects it is, but it comes with a huge cost in speed and flakiness.

(2) Perceptive readers may wonder how this can be--the carriers all recombine when they reach the metallic contacts, so even if

tau = 500 t_T,

how can that work? The physics is that the electron and hole don't arrive at the contacts simultaneously, so that there's a net charge imbalance for a bit, leading to carriers being emitted by the contacts to maintain overall charge neutrality.

Maxwell's demon is still on vacation, so this doesn't happen carrier-by-carrier, but rather stochastically. Nevertheless, on average one detected photon increases the current for a time tau rather than T_t. The stochastic nature of the recombination process doubles the shot noise in photoconductors compared with photodiodes, where the recombination is deterministic and only the generation is stochastic.

Helpful info. Thank you!

Thanks, Phil, for the in-depth explanation. Very much appreciated.

Cheers!

John S wrote: ===========

Illegal in any country that follows RoHS rules.

Insane crap IMO cos the amount of cadmium is so minuscule.

.... Phil

Agreed!

John

Aren't there still Roman vintage aqueducts/pipes in Europe that are lead lined?

How does that tally with RoHS?

John :-#)#

John Robertson wrote: ==================

** U tell me - Mr Smartarse context shifter.

..... Phil

Your point was minute bits of Cadmium are enough to get CDS cells banned in Europe, so why aren't they digging up any and all lead pipes? And car batteries...if you catch my drift.

Yeah, I shifted the context slightly, but felt it was on your modified topic enough (you added RoHS) that it still fit...

I use 70 or so CDS cells in my strobe light Rifle Gallery - once we open again later this summer when crowds aren't a big risk - so I was following the discussion avidly.

John ;-#)#

No real profanity. I see Mr. Phil is having a good day.

** I am not liable for retrieving random pieces of your mental flotsam.

FYI: The bans on Cd, Pb solder etc are not retrospective. 98% or so of car batteries are re-cycled in the EU. Pb plumbing has been banned for decades.

What I called you is clear and precise.

..... Phil

I really do laugh out loud sometimes when I read this guy's posts. Truly epic.

10 lux @ 2856 °K produces a minimum of 10k and a maximum of 100k ohms. In a batch of, say 1000 photocells, what would be the most likely resistance value of a photocell chosen at random?

Data sheet provided by Digikey:

Geometric.

The typical value would be about 30K with a 3:1 spread in either direction.

But a *minimum* of 10-100K? That makes no sense.

That's like saying, "Expect the shipment no later than 2 to 4 weeks from today". So if the shipment arrives in 1 week I should call back to complain?

When a company quotes delivery in days, such as 110 days, is that calendar or work days?

There are other old uses for lead. The Parisians are having a problem of lead after the Notre Dame fire destroyed coloured windows in the cathedral, and there is lead dust around the Ile de Cite now.

Nothing to do with the windows. It was the lead cladding on the roof. Many ancient building have lead-clad roofs. Even on my own roof I have some lead sheet parts.

Jeroen Belleman

Back in the '90s there were folks here who offered churches a brand new lead roof in exchange for their old leaky one.

Sounds like Abanazar trolling for Aladdin's lamp, doesn't it?

The issue was that the lead in the roofs had been smelted a century of so ago, so the accompanying radioactive isotopes (produced from decay of uranium, thorium, and so on) had decayed, leaving valuable 'low-alpha lead' for soldering VLSI chips.

Cheers

Phil Hobbs

You are correct. It is an ambiguous comment. Let me fix it... The data sheet specifies a minimum of 10k and a maximum of 100k ohms at

10 lux @ 2856°K illumination.

Thanks for your input.