LED indicator VS. Battery Life: can it be better?

[earthtonesaudio]

Okay, I want to make my pedals very eco-friendly, and I noticed that the LED indicator consumes more power than the whole effect circuit.  With the idea of making the whole pedal less power hungry, I have been thinking of various schemes to reduce LED current consumption.

-Constant current
-Pulsed current
-???

So then the question is, how many additional components is "too many" when your goal is making the pedal better for the planet?  For example, if I extend battery life, but use 50 extra components and 180 extra solder joints, that's not really a good trade-off in the grand scheme of things.  However, if I can extend battery life with maybe 10 components or less, that seems pretty good.

Bonus question: anyone know of a simple and good way of reducing LED power consumption?

[theehman]

Use a larger resistor to reduce LED current consumption?  I use 1.5K which drops it down to about 5mA.

[Wimpy]

You could use ultra bright led with bigger value series resistor and less current if your effect has "normal" led now. Pulsed or animated leds need some kind of oscillator and may inject noise

[petemoore]

  A Super bright with current limiting.
 Pulse is a great idea except maybe not so good near a circuit, because it may have a strong tendancy to inject pulse from PS into signal path, and I don't know if an LED can pulse faster than you can hear [IOW at a frequency above human hearing].
 I'm not sure how uncomplicated this could be: ramping voltage up and down instead of 'clicking' on/off.

[R.G.]

There are blinking LEDs that do this all on their own. Craig Anderton did an article on this a couple of decades ago. I believe National Semi made them.

You can blink an LED and human vision cannot tell that it's not solidly on if you blink it faster than about 24Hz. It's an effect called the Persistence of Vision. Google "POV"

Zap an LED for a millisecond, do it again 8 to 16 mS later.

There is a real trick in HOW you supply current. If you do it through a resistor or transistor, then there is a significant amount of heat involved. Instead, charge up a cap or an inductor, then discharge them into the LED. This is non-dissipative. I believe that the blinking LED uses the cap version.

But for simple, it's hard to beat the ultra-bright LEDs.

[Mark Hammer]

We've discussed this ad nauseum many times before, but I guess there's a new crop of folks every now and then.

The overall challenge is that of providing sufficient visibility of status at the lowest "current cost" possible.

One tactic, noted immediately above, is to save up small amounts of current over brief intervals and blow one's wad with visible or less-than-visible intermittent discharges.  here, the long-term current drain is decreased but the accumulation of current allows each discharge to make the LED be visible.

Use of superbrights permits smaller amount of current to go a longer way.  By changing from a 300mcd LED to a 6000mcd LED, one can get away with using much less current to obtain the same luminance and visibility.  In some instances, I find I can stick as much as 15-18k in series with the LED to reduce current draw, and still see it reasonably well.  Maybe not from across the stage with spotlights blaring,but certainly in my workshop.

The third tactic is a physical, rather than electronic one: that of installing the LED in such a manner that maximizes visibility per-milliamp-drawn.  So, LEDs inside a black plastic (rather than metallic) bezel, and against a dark surface will permit even small amounts of colour contrast (on vs off) to be visible.  The need for less current to provide the needed contrast lets you use a larger limiting resistor, whether you use a normal LED or superbright.

The overall idea is to create less need for current to provide status indication.

[earthtonesaudio]

I totally agree with the high-efficiency LEDs as a first step.  And good advice about mounting the LED in a way to maximize visibility, thanks. 
I guess I just want to get the average current consumption down to sub-mA levels.  When you have a fuzz circuit that only draws maybe 7 micro-Amps, it just seems so wasteful to have a 3mA LED hogging all the juice.

I've thought about slow blinking, but to me it seems you'd have to use a sine oscillator to keep from having sharp edged waveforms that can get into your circuit.  But the problem with rounded-edged waveforms is that they waste power.

So that leads me to believe that a pulser circuit that operates well above audio is the way to go.  Of course I'd like to do it with a couple discrete components, but maybe one of those Schmitt trigger circuits is better.

[Processaurus]

I got some chinese super bright blue's that I can put a 47K resistor in series with (on 9v) and get plenty of light, lemme see, V=IR, so V/R=I, so 9v/47000 ohms = .191mA.  Crazy.

[mnordbye]

Another one.. size. Haven't tried this, but i believe it's logical that 3mm LEDs use less power than 5mm etc. Anyone know about this?

Magnus Nordbye

[earthtonesaudio]

I believe the size doesn't make any difference until you get up into the high-power LEDs, like the multi-Watt ones that are 10mm across, that sort of thing.  I think the single-diode low-power LEDs are all the same except for the case/lens style.

[mnordbye]

I believe the size doesn't make any difference until you get up into the high-power LEDs, like the multi-Watt ones that are 10mm across, that sort of thing.  I think the single-diode low-power LEDs are all the same except for the case/lens style.

Guess so. Noticed that both my 3mms and 5mms have the same numbers on the inside leads, so they might be identical, mechanicly, actually. 3mms use less case material though! 

Magnus Nordbye

[theehman]

I got some chinese super bright blue's that I can put a 47K resistor in series with (on 9v) and get plenty of light, lemme see, V=IR, so V/R=I, so 9v/47000 ohms = .191mA.  Crazy.

I've never gone that low with my blue LEDs but I do usually run them at 2mA w/ a 3.3K resistor.

[ACS]

My brother is an electrical engineer and built him self a high powered LED light unit for night mountain biking (!)  I remember having this exact discussion with him - and while I didn't understand enough of it to be truly useful here, the gist of it was this:

"There are a few ways to do this, but all involve wasting loads of your battery energy as heat"  He felt the best way to do this was with Pulse Width Modulation (PWM).  He built pretty simple circuit to do this (I think it had a programmable chip at its heart) and sure enough got roughly 6hrs battery life out of a battery that he used to only get 2hrs out of, with the same apparrent brightness...

Hope this helps someone!

Aidan

[Skreddy]

The latest batch of ultrabright blue LEDs I got will burn a hole through your retina if you star directly into them using anything less than a 100k current-limiting resistor.

[any]

I think that a lot of pedals don't need led's, you know when they're on...
Especially circuits like a positive ground Fuzz Face have no benefit at all, I rather build those without led or dc plug.

How about powering the led of a mini solar cell charging a small secondary battery? (like a calculator)

[darron]

Google "POV"

hmm... but DON'T google IMAGE search "POV".... i don't know what it stands for but it's not what you were after!

[Paul Perry (Frostwave)]

It wouldn't surprise me if you could put a led in series with the emitter or collector resistor & get enough light, for "free" in some circuits, maybe tweak a resistor or two to make up.
I must say, having a blinking led at about 5Hz allows you to use a much dimmer led and still have it noticeable.
In fact, even if you have power to burn, it is sometimes hard to tell if a led is on, or just reflecting. With a blinking led, there is no argument.

[geertjacobs]

Stupid idea:
Why not have a mechanical way of indicating the status of the switch?
E.g. a switch with a window displaying a bright colored surface when on versus a black surface when off?

There are these "hand counters" like this that add when you push them.
http://www.brannanshop.co.uk/acatalog/Counters.html

So there must be something that toggles between two statusses (is that english?) when pushed.
If it had extra electrical contacts that would be great.

Just a thought...

[George Giblet]

Obviously you will save power with a bright LED and run it at the lowest current.  Choosing the least brightness required is a good start.

The major evil is powering a 1.7V LED from 9V.  The power dissipated in the resistor is about 4 times that getting to the LED; P_R/P_LED = (9-1.7)/1.7.  So you can gain by feeding the LED using a lossless current source, which is basically a switched inductor.   The cost here is the inductor and in some cases you might need to shield the inductor.

The pulse thing RG mentioned gives you about twice the perceived brightness increase.

So if you use an LED which is twice as bright, and drive it from half the current, then use an inductor based current source, you will reduce the power by 2*4 = 8.  If you use the pulsed scheme you will reduce power by  2*4*2 = 16.

Once you get your LED current below some percentage of the circuit current it becomes a waste of time saving power.  For example a 5mA LED current on a 30 to 40mA delay effect isn't a big issue.

[earthtonesaudio]

Good points, all around.  I guess my main goal is to find a way to get LED consumption down to or below the consumption of the effect.  I have some circuits that draw only a few micro amps, so that's the ideal target.