Can we discuss the proper method of testing the values of an LDR?

[kimelopidaer]

Hello diy.

So the datasheets for various photoresistors show min and max values of resistance for different ldr's.
Reading around, I see that one method of measuring the ldr's min/max yourself is to bring your multimeter outside into regular daylight and read the resistance value, then you cover the ldr to find the max resistance.

I wonder how "covered" the ldr needs to be? I find that most of my ldr's read in the megaohm range when in complete darkness. Is that what you want the conditions to be when measuring max resistance? I ask because when I am outside, and cover the ldr, let's say with my hand or heavy paper, the resistance goes up but usually not nearly as high as it would be measuring in a total absence of light. Still a good bit of indirect light bouncing around there.

Would anyone care to comment from personal experience?
I have a couple of ldr's from a grab bag that work much better for my needs than the others that were included. I have a feeling it's due to a higher max resistance, but I need certainty before I try to order more of the same.


Regards,
K

[.Mike]

I've never heard of using the sun to test an LDR beyond just making it sure it is working. While this isn't rocket science, the amount of light that hits your particular spot on the planet changes every day. Results from the winter will be different than results from the summer. Plus, the spectrum of light the sun puts out is much wider than an LED. Not a good test.

Your better bet is to create a scenario that closely matches how you expect to use the part. In order to make sure that test results from today are accurate relative to test results a year from now, you could, for example, use a certain LED at a certain amount of current, completely shielded from outside light.

A big factor in LDRs is response time, not just resistances. LDRs are fast to to react to light, and much, much slower to react to dark. Different compositions react in different ways, and parts of the same model number but a different batch can be widely varied.

Check out the datasheet for the NSL-32. They list some test conditions, like rise time (how long it takes to get to 63% of final conductance after applying 16mA) and decay time (how long it takes to get to 100K after removing 16mA). It might give you a few ideas.

Mike

[Earthscum]

Remember that your LDR is just a resistor. If your LDR has 10MOhm dark resistance, and goes down to  10k when you shine an LED with 5mA current lighting it up, but you only need a dark resistance of 100k, why not just use a 100k resistor in parallel? Then you also soften the "knee" if it has one (some LDR's have an apparent break-over point where the resistance changes drastically for a small change in light... I've found this to be more a function of the LED than the LDR in self-rolled vactrols, but the resistor works the same trick).

In a wah I'm building, I shaped the LED to distribute light in a (somewhat) controlled manner. It uses a physical gate, like Morley wahs use, to cut off the light (you're probably familiar with this arrangement). In case you're interested, it uses 2 LDR's, one over the other for a strange arrangement. Once I got the light scatter working good, I used an LM317 as a current regulator so I always have a constant light instead of varying with battery drain. It's all completed (the gate mechanism, anyways) and I put a cover over the side of the gate to keep any light leaking into the case from disturbing the LDR's.

Aha... found the pic... It's important to maintain a stable environment for your LDR's, unless you're making your own Vactrols... then just roll 'em and test 'em. 9V power, 1k resistor on the LED side, read your resistance. 2x 1k, read your resistance. 5k, read your resistance, 10k, read your resistance. Make a little chart and it will show you a curve, the real-life curve of your new Vactrol! Convert it to mA by measuring the voltage each time across the (total) resistance, divided by the (total) resistance, and map that out, and now you have a Current to Resistance chart.

[Ronan]

>I have a couple of ldr's from a grab bag that work much better for my needs than the others that were included.

Sounds like you have already worked out some sort of testing apparatus. They can only be "tested" with a known quantity light source, that "light source" is not easy to reproduce accurately for all pedal builders. Therefore, without laboratory quality calibrated light sources, the pedal builders "test" their LDR's with whatever means works for them. That is how I see it. Every individual LDR I have purchased differs from the others in a batch, the datasheet can shed some light on how close they should be

[R.G.]

I've thought some about this.

The way I came up with was to adopt an LED as a "standard" light source. It's not particularly laboratory-grade, but the same LED generates the same light at the same current every time. They change slowly over time, but for an LDR tester, there's not a lot of power-on time, so it's probably good enough as long as it generates enough light for you, and it's uniformly spread over the LDR. The rest of the test rig allowed testing the resistance at fixed current levels on the LED (and hence, fixed light levels), including some maximum, some intermediate range, and zero current, which is zero light emission. A metal tube that's reflective inside helps with light mixing, and prevents outside light from poisoning the results. Light covers allowing fitting the LED into the tube at one end and the LDR at the other completed the light fixture. The LDR fitting has to block outside light and not be difficult to get onto/off the tube. I used black insulating foam to avoid changing the measured resistance.

There are two tests for an LDR that we can manage OK. One is the raw resistance-versus-light curve, and the other is the time response. The resistance curve is easy. Just hook on an ohmmeter (we'll designate  your meter as the officially-accurate standard meter) and rig up some way to force measured currents through the LED. That generally requires some current-setting circuit and a meter, possibly a second meter. The current-setter can be as simple as a pot and a switch (to get truly zero current), and in a pinch you can use the same meter to measure current, although that compromises accuracy. You set the current, write down the current value, then write down the resistance at that current for a range of values up to the LED's maximum.

LDR speed is another thing. For this you need an oscilloscope and some way to electronically switch the LED current on and off. You make the LDR part of a voltage divider and measure the voltage across it with the oscilloscope. Then you use the electronic switch to turn the LED on and off and watch the voltage rise and fall across the LDR. This tells you how fast the LDR will change. This is important because we generally need the fastest LDRs for musical stuff. Some LDRs take seconds to change.

There are other important things that might be tested, but are difficult to test. LDRs are color-sensitive and temperature sensitive. It's hard to do much with those in a quick-and-dirty amateur setup, although you can simply use a thermometer and make sure you're not testing at 110F one day and 32F the next.

[Earthscum]

LDR speed is another thing. For this you need an oscilloscope and some way to electronically switch the LED current on and off.

I didn't use my scope for this one (and kind of forget about it, to be honest). I had actually rigged up a test oscillator on my breadboard. I  just plug the LDR in, fire up the circuit, and measured the frequency and voltage using my DMM. I used schmitt triggers, nice square, fairly fast rise and fall, faster than the LDRs I had by  far). I can't remember the math right at the moment from lack of coffee (not hard math, either... something like voltage divided by the frequency since I had the cap as a constant or something), but it was fairly accurate as far as I could tell, at least for me finding the fastest ones and slowest ones in my bunch. I had a black rag I'd throw over with a couple test leads hanging out.

Anyways, it worked great the one time I sat down with it (and I threw them all in the same bag within a week, of course).

[bonaventura]

i would second Mark's suggestion and test it in its working sorroundings.

just to share,i find it odd that here in this part of the world,sellers dont rate the ldr they sell. there is just one kind of ldr and they come with three leads.

when i told them that i needed 2 Mohm dark resistance they gave me the take it or leave it look.

surely enough the one i bought clipped the compressor i was building.

since my DMM couldnt read the dark resistance,i placed it as voltage divider and work out its dark resistance.turned out that it was ard 80Mohms or over.while with the led on,it was abt 5k.

so needless to say,alot components were replaced. but after several tries i finally said to myself, done!

boxed it in a plastic box, shielded with aluminum foil as adv in RG's site (thanks for the tip!).

i couldnt find any info abt 3 legged LDR,but it behaves just like LDR (to my knowledge anyway) and once i figured out the parameters i got it working.so im quite happy (i modded it twice after i said done  ).

[kimelopidaer]

Those wah pictures are fantastic. Now I anticipate etching some special function pcb's like you made for various control elements in the wah.

I figured out through trial and error that I can't correlate an ldr's physical size with its resistance properties. I assumed this.

I set up an LED testing rig to get as close as possible to the actual conditions. That meant, for me, identical color and current.
I have a few packages of cadmium sulfide photocells that I picked up at RadioShack.

In a simple tremolo style circuit I am playing with, the ldr is in the signal path and is modulated by a flashing led, which is made to pulse by an oscillator. Certain ldr's appear to give the tremolo pulse more definition. The less suitable ldr's induce a weaker effect. The chop is not as strong.
For the sake of repeatability, I hope to identify the characteristics of the ldr's in my successfully operating circuit.

I tested my radioshack photoresistors and 6 out of 15 "work" the way I need them to.
I think that it might hinge on rise and fall time, since most of the ldr's in tested hit the megaohm range when dark.
I have some assorted photoresistors coming via mail, and the component numbers of these will be known. Perhaps i'll know more then, if I don't find an oscilloscope by the side of the road first.

Regards,
K