Driving an LED with the guitar signal?

kforte318 · February 24, 2016, 07:14:53 PM

I want to drive the LED inside an optocoupler with the incoming guitar signal, but I'm having a little trouble. I'm very new to electronics, and at first I simply tried the basic LM386 amp with an LED and a 1k resistor on the output as opposed to a speaker. It seemed to work at first, and then slowly the LED stopped responding as drastically until it didn't respond at all. I then tried the 386 portion of the Uglyface schematic, considering it is doing exactly what I'm trying to. The same thing seemed to happen. What am I doing wrong? Here is my schematic as it stands currently. Sorry for the noob questions.

R.G. · February 24, 2016, 07:33:25 PM

We were all noobs sometime. As long as you're learning, it's good.

The reason the LED died is probably that LEDs do not like to be reverse biased. It slowly kills them for reverse currents to flow. They can only deal with a little more reverse voltage than forward voltage. Generally they're about 1.4 to 3.5V forward to get them to emit light, and maybe 5-7V reverse, which starts killing them. So putting AC on them is a death sentence.

You can do a couple of things. One is to put an ordinary silicon diode across the LED but in the reverse direction. That way, the reverse voltage is let flow by the silicon diode. There are others, but you're trying to do something there that already has limited voltages, so a reverse shunt diode is the first thing to try.

kforte318 · February 24, 2016, 08:14:39 PM

Is there a reason it works on the Uglyface schematic without the second diode? I will definitely try that when I get a chance regardless.

deadastronaut · February 24, 2016, 08:25:02 PM

http://www.aronnelson.com/gallery/main.php/v/chickpea/sound2light2.jpg.html

with a 386.

kforte318 · February 24, 2016, 08:39:08 PM

Would the LEDs not be reversed biased in that case as well?

idy · February 24, 2016, 08:45:05 PM

The "reverse bias" would be held to around .6v by the silicon diode. Voltage above that will turn the si diode on and slip right through.

armdnrdy · February 24, 2016, 08:51:42 PM

Couldn't this be accomplished with a "modified" peak level indicator circuit?

kforte318 · February 24, 2016, 08:53:06 PM

Quote from: idy on February 24, 2016, 08:45:05 PM
The "reverse bias" would be held to around .6v by the silicon diode. Voltage above that will turn the si diode on and slip right through.

Oh! I actually meant in the layout deadastronaut posted.

PRR · February 25, 2016, 01:23:42 AM

> slowly the LED stopped responding

Did this LED actually die/dead/kaput/extinct ?

Or did it stop working while playing, but came back after a rest?

My suspicious is that a cap driving a single diode will always charge-down until current stops. If you let the charge bleed-off, it will work again for a while until it charges-down again. This is a key part of tube distortion (grid-block); and of several voltage multipliers.

If you don't want charge-down...

* Two diodes back-to-back will equalize things.

OR

* Add a heavy bleed resistor.

Both shown below.

Two LEDs is wasteful of your LED budget. Heavy bleeder is wasteful of your battery. In this case, neither "waste" is prohibitive.

Since you are driving 100 Ohms or 10 Ohms, the 2uFd cap seems small. It will also charge-down faster. For full bass in 100r, or short-bass in 10r, 100uFd is a fine value.

The second LED "could" be a plain diode. But the 0.6V versus 1.8V difference of voltage still leads to asymmetry and charge-down. It could be 3 or so plain diodes series (1.8V), but today three plain diodes cost as much as many basic LEDs. The second LED does not have to be exposed, it won't look different from the first one (but does give more light).

A bridge rectifier is another fix, but it adds 1.2V to the threshold voltage, and under 9V power you may not have that much swing. Anyway a Bridge is four diodes which is typically more cost than a cheap LED.

I would suggest Red, maybe Green, LEDs. Blue and White LEDs need more voltage, and a '386 on a 9V supply may not be able to get them lit well.

smallbearelec · February 25, 2016, 02:00:02 AM

Quote from: kforte318 on February 24, 2016, 07:14:53 PM
I want to drive the LED inside an optocoupler with the incoming guitar signal...Sorry for the noob questions.

No apologies necessary! I was wondering what you intended to do with the LDR side of the photocoupler. Turning on an LED momentarily at an intensity that varies with your picking dynamics is what one kind of envelope generator circuit does. A familiar one is part of the Mutron filter. Check out Mark Hammer's article at GEOFEX to see what I'm talking about.

http://geofex.com/Article_Folders/ECFtech/ecftech.htm

kforte318 · February 25, 2016, 08:01:28 AM

Quote from: smallbearelec on February 25, 2016, 02:00:02 AM
I was wondering what you intended to do with the LDR side of the photocoupler.

I am trying to add a function to a tremolo that will modulate the rate and possibly depth based on picking dynamics. I'll definitely check that article out, thanks!

kforte318 · February 25, 2016, 08:05:24 AM

Quote from: PRR on February 25, 2016, 01:23:42 AM
Did this LED actually die/dead/kaput/extinct ?

Or did it stop working while playing, but came back after a rest?

That. It didn't burn out entirely, it just fizzled and would return to a working state after a minute or two. I thought it might have something to do with the cap! I'll test out your solutions and see how it goes. Thanks for the super detailed answer (also rad copy-paste diagram, took me a second to realize you hadn't just very accurately recreated my handwriting).

Kipper4 · February 25, 2016, 08:40:16 AM

Why not go the envelope detector route.
Search for my ldr sweller. That will get you started with detectors

kforte318 · February 25, 2016, 08:54:09 AM

Quote from: Kipper4 on February 25, 2016, 08:40:16 AM
Why not go the envelope detector route.
Search for my ldr sweller. That will get you started with detectors

Is this an accurate schematic? It looks like the original you posted in that thread had been deleted. Also, just for learning's sake, is what I am currently doing technically just a very crude envelope detector?

kforte318 · February 25, 2016, 08:57:49 AM

Quote from: PRR on February 25, 2016, 01:23:42 AM
OR

* Add a heavy bleed resistor.

The heavy bleed resistor got the circuit up and running! Thanks! However, I'm going to explore an actual envelope detector as Kipper suggested I think. For some reason, the guitar signal from the 386 seems to somehow be making it's way back into my main circuit... the modulated rate is working, but the very distorted guitar signal is constantly in the background.

Kipper4 · February 25, 2016, 09:47:25 AM

Also if you look at replays 34 and 57 there's lots to choose from. Take note of what the guys noted too.
Maybe look at the nurse qaucky envelope detector.

Edit if you look at the schematic linked in this thread there's a similar envelope detector in the dod440 . At the top of the schematic.

http://www.diystompboxes.com/smfforum/index.php?topic=113385.0

Post 1

duck_arse · February 25, 2016, 10:20:36 AM

have you chosen a target tremolo yet?

kforte318 · February 25, 2016, 10:30:12 AM

I'll check me out Kipper. Thanks! duck_arse, I do have the tremolo up and running, yeah.

Kipper4 · February 25, 2016, 10:45:01 AM

Which trem are you using though?

kforte318 · February 25, 2016, 11:18:11 AM

Quote from: Kipper4 on February 25, 2016, 10:45:01 AM
Which trem are you using though?

Not any specific one. I'm using an LFO I found a while back (I could dig it up later but am at work now) and a single bipolar transistor optocoupler VCA (also will have to wait until later to find it). I can post a full schematic later as well.

News:

Driving an LED with the guitar signal?