Photocells - Measuring Resistance

[chromesphere]

Hi all,

just ordered some photocells (for a tremulus lune build) from futurlec.  Before heatshrinking the photocell with the led, i thought i would test them to see what values the thing gives and make sure they are the correct ones, because they are completely unlabelled.  The photocell i ordered was futurleces photocell1 http://www.futurlec.com/Photocells.shtml in line with the specs needed for the tremulus lune.Connected the photocell to multimeter, switching the lights on and off like a crazy person.  Found that the resistance varied from 0 ohm's all the way up to over 10-15 megaohms!  So i'm assuming, 1) you cant measure these like normal resisters 2) They sent me the wrong part or 3) photocell specifications are only a rough guide?

Or something else im missing?  Do i need to worry about the high dark resistance of 10-15+ megaohms?  (The resistance sort of...continues to climb the longer the light is off).

Maybe its just the behaviour of photocells?

Edit: been googling a bit about photocells.  Apparently measuring their resistance is a 'rough science'.  I'll try and use an led that gives the closest results even though i suspect it wont make much diiference.

Any info appreciated!
CS

[JKowalski]

It is really "rough science".

A: The manufacturing process is highly inconsistant.

B: The amount of light you shine on them is hard to regulate since it totally depends on angle, distance, lumens, the directionality of the source... everything! Even if you set up a "test" LED source minute differences in the way you put your LDR's in front of it will skew results.

Every photocell I have tried I have been able to get to 10 ohms to 20+ megaohms with enough light. However, some LDRS are different in that the same light level will give different values of resistance. Though they all can be blasted with enough light to eventually reach peak highs and lows you don't use them that way. If you only have a dim light source you'll probably use a sensitive LDR, if you have a bright source you'll use a less sensitive LDR.

Personally I don't even bother. Usually you can just stick any ol LDR in there and make a few minor adjustments in the light source + circuit to compensate for the finicky part. Say in the tremulus lune, after you put the LDR in it's a good idea to trim (trimmer) the gain of the second stage so that you don't hear volume drop/gain when engaged. Which is why it has the trimmer.

[PRR]

> switching the lights on and off like a crazy person

You really want a controlled consistent light source.

At-home, you are unlikely to duplicate the photo-R factory tests. Even if you have a light meter, you must also match their color spectrum.

And what do you want? NOT "factory spec", but "works in YOUR project"!

What IS your project? A simple tremolo. You only have one LDR. The "average light" off the LED should force the LDR to roughly the same resistance as the resistor it feeds into, which is adjustable 25K-0 (or *100K-0). Even allowing that "0" should probably be above 1K (so the opamps don't strain), that gives 25:1 of circuit-trim to match the LDR's actual light/resistance curve.

You proved it isn't a dead blob of plastic; it does have light-magic. Stick it in, play, trim.

The lower end of typical LDRs is rarely as low as 10 ohms in bright sun; if you read less then your meter is somehow mixed-up.

> Do i need to worry about the high dark resistance of 10-15+ megaohms?

This circuit never lets the LDR go DARK. Or maybe just-barely. It's not relevant.

Yes, the high-end of a good LDR in TOTAL DARK keeps rising to many Megaohms for hours or days. The spec-sheet only promises a certain minimum at a certain time.

In this use you want to be over 100K in much less than a second. This can be tested. It is FAR faster to just build and listen. If the LDR is really "slow", the trem-depth will weaken at high trem speed. This is not a major musical flaw. You can also hit the LDR with a brighter light to driver it to a lower-resistance more-excited state (reduce the LED resistor and re-trim the gain resistor).

[chromesphere]

Excellent, thanks for the info guys!  I'll play around with the trim and see what i get. Sounds like theres no point trying different LEDS.  Might just socket the ldr / led anyway just in case...

Anyway, thanks again for the info!

CS

[R.G.]

Adding to what Paul said:

To do any measurement, you have to have some kind of reference. You can make the reference up, as Farenheit did on that cold day when he calibrated his thermometer to the outside temperature. It helps if the reference you choose is relatable to some major facet of the world/universe, like the freezing point of water under standard earth-normal atmospheric pressure which is how we got the freezing point of water to be zero Centigrade/Celsius, but if you only want to match things to be equal, that's not a necessity. You just need something that's mostly constant.

So make a reference. One good one you've already tumbled onto is dark; the absence of photons, pretty much. Another is the light emitted by this here ... (picks up an LED out of a baggie of LEDs) LED. I dub thee Sir Reference LED. The world of my LDRs will be measured relative to YOUR light output at a current of 1.00000ma as measured by this here cheap meter I bang around in my toolbox most of the time. I (quickly!) construct a circuit which, when powered by a 9V battery produces 1.0000ma through the LED.    Life is good.

I have no clue what the light intensity out of my Reference LED actually is in photons, lux, spectral distribution, irradiance. whatever. It's exactly one Reference LED.
So I set up my reference LED emitting its photons into this here length of black rubber tubing. I hold the LDR to be tested on the other end of the rubber tubing, sticking a piece of black foam rubber window insulation over its leads to seal out back lighting, and measure the resistance of the LDR under the illumination of one Reference LED. Note: don't use conductive foam for this...  

Over short periods of time, the output of the LED will approach constant if you keep the current reasonably constant. At least, it will vary much less than the LDRs do. You can match any reasonable number at any current you pick. You could even take the data, and plot the curves of resistance versus LED current.

Want to measure LDR speed? OK, rig the LDR in series with 47K to a 9V battery. Watch the voltage across the LDR on a scope and turn the LED on and off. The LED switches in microseconds or less. The LDR rise and fall times can be read right off the scope face.

[petemoore]

  There may be times when a faster sweeper is needed.
  Also, lower current through the LED might be a desired quality for this LDR app.
  Very low resistance [easy to bump up the bottom stop of R value, series a resistor here...oh yes, I like the ones that do the limbo.
  Very high resistance is fine, but "what for?-if ever needed", series 2 LDR's.
  If your LDR is too high all the time, slap a resistor across there...that'll bring it down, especially the high to really high value periods.