stroke your leds, map your ldrs

[duck_arse]

everyone seems to be playing with ldr//led combos lately. even me. I'm driving a led in an opamp-feedback loop with an envelope (more of that soon), and I was testing led brights with a pot and a cro. later I thought it would be easier if I connected to an osc, to see the real ldr response, so I hacked up most of that on the breadboard, and here it is.



the upper circuit is for manual mapping. the 47k//50k pot provides my manual envelope, swinging from 0V to about V/2. I am swinging the ldr across a 47k resistance, so that is the value shown across the ldr, with 47k above the ldr to the + supply. connecting an o'scope, or a voltmeter with an analog display, shows the resistance change for the ldr with a particular led, and for a particlur led with a value ? for R-bright. this is to test the chosen led produces a response from the ldr to the full range of pot rotation, or near to.

the second circuit is just an lfo. it can be used to produce a slow triangle output, or a (slightly faster) ramp output, up or down, to test an ldr's recovery/response. the ramp can be trimmed in with the "slope trimm" pot, or the "ramp diode" can be connected across the pot, either direction. again, with the o'scope connected across the test ldr, it's response will be mapped in a bright blue/green line. I've re-biased the led drive in the bottom circuit to at least attempt to emulate an ordinary envelope, all the way down to 0V.

and that is why we would be using the LM324. I used an LM358 (and an LM741 for half the osc, cause it was there), probably any of the "output down to ground" types will do. this is the scheme I intend using, so it's what I'm testing. with 9V supply, the osc produces about 5V p-p, with 2V offset-under. by the time it reaches the led driver, it looks quite a bit like an opamp//diode//cap style rectifier output.

please note that the ramp diode shorts the frequency-setting pot on the half cycle it is conducting, so that half a cycle goes missing, but for the series 1k. you can increase the 10uF timing cap for longer ramp times, but it produced some weird dc offsetting-effects at the led when I went to a 22uF tant.

this has probably all been done before. over to you for corrections, comments, improvements. thank you for your time and attention.

[Kipper4]

Thanks Duck Arse.
Looks intresting.
"Pot and cro"     Cro?

So your testing the responses or differant led.
Which Lars are we talking about?
Without looking for the data sheets. Did you pick the lm324n because of low power consumption and linearity?

I guess the lfo could be taken further to produce a more sinusoidal wave and still be able to power the led/ LEDs? In the feedback loop?
I wonder if it will tick ?
Sorry for all the dumb questions. Just trying to learn.
Thanks Rich

[armdnrdy]

Did you pick the lm324n because of low power consumption and linearity?

I guess the lfo could be taken further to produce a more sinusoidal wave and still be able to power the led/ LEDs? In the feedback loop?
I wonder if it will tick ?

I would imagine the choice of the LM324 is because four op amps were needed? Also, the LM324 can go rail to rail. Good for LFOs.

An LFO is less likely to tick with a sine wave than a triangle or square wave. Most flangers/phasers use triangle and tremelos use square.

[duck_arse]

"cro" - cathode ray oscilloscope, an old technology, replaced with lcd's. you know what pot's are.

the only ldr's I have at the moment are some 4mm, coarse channel type from rockby. they will get down to about 5k bright. the envelope will go to 0V with no signal, so the leds want to turn off, and that seems to me to means an opamp that will output down to 0V. the lm358 is one such, a dual, I did my testing with. the lm324 is another, I haven't tried, along with the cmos varities. it was also easier to draw with.

the lfo tikk is immaterial. we are just testing leds for brightness, resistor value, and response to the envelope. the osc is a refinement to the manual method, saves a lot of knob twiddle. the ramp is usefull to test the response/recovery speed of your ldr. knowing you have a ramp driving, you can see the lag on the drawn waveform, which comes from the current through the ldr itself. a sine would be less usefull.

the whole idea was to compare one led to another, set some brightness resistors.

[Kipper4]

Thanks Duck Arse.
You know me I think aloud.

[GibsonGM]

Nice setup.  I might have to put this together, DA!

I love my old CRO     I have a Tektronix 2213, serves me well!   

[duck_arse]

have at it, gibson, let us know if it works how you go. this might be a good use for the old moving-coil voltmeter under the bench, too. and you get a good indication of how the distance between the led/ldr affects the ldr outcome.

I probably should have added that this will work equally well with vactrols. I hear they are quite popular hereabouts, I just don't have any.

[GibsonGM]

have at it, gibson, let us know if it works how you go. this might be a good use for the old moving-coil voltmeter under the bench, too. and you get a good indication of how the distance between the led/ldr affects the ldr outcome.

I probably should have added that this will work equally well with vactrols. I hear they are quite popular hereabouts, I just don't have any.

If I get time, I will!

I use the old meter movement from an ancient VTVM as a series ammeter for my pedal projects     But I have a couple of extras here somewhere! 

[Kipper4]

This Is a lot of fun so far I built the test circuit after making a perf layout in DIYLC.
I used 250k trimmers for the R Brights and put wires on them to be able to test the pots resistances.
This could keep me busy for weeks
Thanks DA

[tommycataus]

Thanks DA! Great idea.

[duck_arse]

have you seen any interesting waveshapes across the ldr's, kipper?