LED Brightness - mcd, limiting resistors, etc.

[tonyv77]

Just working on a quick (or so I thought) "Tube Screamer on Steroids" mod to my SD-1. Sounds good but the I went to replace the LED with a 5mm green and noticed it wasn't very bright. I lowered the limiting resistor value (R30) without much luck (went down from 3.9k to 1.0k). Then I looked at the package and saw "30 Mcd". I was pretty sure this meant millicandela as Megacandella is not a measurement. So I swapped in a yellow 5mm LED (http://www.taydaelectronics.com/leds/round-leds/5mm-leds/yellow/led-5mm-yellow.html) with little to no difference. Starting to think something is up here. Shouldn't a LED with a rating of 2000-3000 mcd be brighter than the stock Boss? I thought I was going to have to INCREASE the limiting resistor.

 

[jtcnj]

Dunno, I usually use 4.7k on Leds on builds, never modded an SD-1.
I put a 5mm green one in a TS-5 replacing the small stock one but the point was to make it stand out more.

Intuitively I think a larger resistor would affect the brightness as it lowers the current.

 

[tonyv77]

Dunno, I usually use 4.7k on Leds on builds, never modded an SD-1.
I put a 5mm green one in a TS-5 replacing the small stock one but the point was to make it stand out more.

Intuitively I think a larger resistor would affect the brightness as it lowers the current.

Sorry mistyped. The green replacement LED wasn't bright at all.

 

[Bristlehound]

If you know the forward voltage and forward current of the LED from the spec sheet, then use this calculater to get the Current Limiting Resister.

If you don't have the data, then click the question marks and it tells you how to approximate the forward voltages and current depending on colour.

 

[tonyv77]

If you know the forward voltage and forward current of the LED from the spec sheet, then use this calculater to get the Current Limiting Resister.

If you don't have the data, then click the question marks and it tells you how to approximate the forward voltages and current depending on colour.

Cool, I get 390 ohms. Source of 9v, Forward voltage of 2v, current of 20 mA.
Weird. I hear of pople putting in 10 and 15k ohm resistors when they replace their LED. Maybe this is for the 10,000 mcd or more LEDs. Not sure what the specs were on the stock Boss 3mm red LED.

Many thanks!

EDIT: My supply may not be a full 9v. Should I measure the voltage at the LED connection?

 

[Bristlehound]

I wouldn't worry. Sounds like a 330R resister is the right one.

 

[jtcnj]

Sorry mistyped. The green replacement LED wasn't bright at all.

No you were right, my response was not clear.
Sounds like you are on the right track.

 

[reddesert]

Cool, I get 390 ohms. Source of 9v, Forward voltage of 2v, current of 20 mA.
Weird. I hear of pople putting in 10 and 15k ohm resistors when they replace their LED. Maybe this is for the 10,000 mcd or more LEDs. Not sure what the specs were on the stock Boss 3mm red LED.

Many thanks!

EDIT: My supply may not be a full 9v. Should I measure the voltage at the LED connection?

If you put in the maximum rated current into this calculator, you'll get the lowest possible resistor and run the LED as bright and hot as it can go. Any more and you're likely to burn it out.

People often use 1K - 10K resistors because they don't want the LED that bright, and it lowers the current draw of the pedal. An overdrive circuit might draw only 10 mA or so, so if the LED is drawing 20 mA, you're using relatively a lot of power just to run the LED. This mattered more when people used batteries or modest wallwarts.

 

[tonyv77]

If you put in the maximum rated current into this calculator, you'll get the lowest possible resistor and run the LED as bright and hot as it can go. Any more and you're likely to burn it out.

People often use 1K - 10K resistors because they don't want the LED that bright, and it lowers the current draw of the pedal. An overdrive circuit might draw only 10 mA or so, so if the LED is drawing 20 mA, you're using relatively a lot of power just to run the LED. This mattered more when people used batteries or modest wallwarts.

I went with 680 ohms. Just about doubled. Seems to work out. I socketed R30 (been swapping too much, solder pad starting to lift) and tried 390 and 680. 390 made my cheap 300 mA wallwart "whistle". 680 seems fine. I think I'm drawing about 7-10 mA.

 

[tonyv77]

Still learning, more questions: there is a diode attached to the limiting resistor (D2). It's obviously not for clipping, so what it's purpose?

 

[NotAnotherHobby]

First, a resistor attached to a diode or an LED is a means to limit the current going through the diode. You put a diode across the mains of any power supply that provides more current than the diode is rated to handle, and you'll fry it. And LED's - as a rule of thumb - run around 20 - 50-ish mA. They have a minimum forward conduction current, I believe, and an "optimal" current. This is the reason why silicon signal diodes are these tiny beads of glass with two leads coming out of them, and you'll sometimes see power supplies with rectifier diodes as thick as one of your fingers: because one can handle more current than the other. But when put into a circuit, they'll try to pass as much current as is available. Hence the reason why you limit what goes through them.

And yeah, I've seen one of those huge barrel diodes go before. It cleared out a guy's office, the fumes were so bad.

Second, you are correct that the mcd is an intensity measurement. The Boss requirements for their LEDs are probably not uber-critical, but I'm sure they use a type that is probably pretty bright. That being said, there is LOTS of variation in LED intensities these days, moreso when I was actively working in electronics. Even with the intensity rating, I've found that certain LEDs are more visible than others. Maybe it's a biological thing, or a personal thing. A "bright" green LED to me is tolerable, but one of those stupid blue LEDs in the same intensity range will have me seeing spots for an hour.

(I was researching at one point using UV LEDs to make an EPROM eraser, or something else that required UV light; maybe something to do with making PCBs. Never went down that road, though.)

Third, it's not really well known, but the brighter you run an LED using straight DC current, the shorter the lifespan of the LED. LEDs do indeed burn out, and the more current they pass, the faster this happens. It's not like it'll happen in hours or days, but it will happen. So they try to find a happy medium when they figure out what resistor they want in the circuit.

And seeing that you're asking: this is the reason why a lot of circuits that use an LED as an indicator generally tend to do Pulse Width Modulation to control the brightness of the LED. Basically, this means some sort of oscillator or square-wave signal turning on and off the LED at a specific rate. The bigger the "on" pulse-width, or the more often the LED is switched on, the brighter the LED is. This supposedly extends the life of the LED because the thing is not continually lit, and not continually passing current. This probably isn't optimal for your situation, but it is something to keep in mind should you need it. A Boss pedal isn't going to have this kind of circuitry in it because it means more components on the board (and thus more cost), and your standard LED is going to last for years anyways.

 

[Mr. Lumbergh]

Personally I hate a bright LED in a pedal. Clear but not bright. I just want to know if it's on or not, not have purple tracers in my vision when I look back up at the audience...

 

[Bristlehound]

Indeed. I've just got a couple of Caline pedals with retina frying blue LEDs. I'll be swapping those out pretty soon.

 

[NotAnotherHobby]

My former employer made a device that had a tiny blue LED to indicate network activity.

I could never look at the stupid device because of that. Just a glance and I'd be seeing that spot for the next frickin' hour or so.