AB box LED is way too bright.

[pinkjimiphoton]

Speaking of magic smoke, I recently learned the hard way that if you are using a pot, put some other value in series in case you accidentally turn the pot to the minimum. Both LEDs and pots can be destroyed otherwise if you are using a 1 amp 9V power supply. I am using a scrapped resistor sub box from work now. Much better.

now ya gotta custom taper!!! its not a problem, its a feature!!!

and burned out shorted led's make great clippers if ya drive 'em with a tesla coil

lol

[Mark Hammer]

one side of my box is 100r/220r/470r/1k/2.2k/4.7k/5.6k/8.2k/10k/22k/33k/47k

the other side goes from 56k right up to 1m..and 4.7m,.//very handy for subbing on a breadboard quickly...

That kind of illustrates my point. 1k through to 10k is useful for this purpose, but there's a "dead zone" between 10k and 22k that could be useful for dimming LEDs appropriately, that is hidden from us.  If it was 20 years ago, and everybody's parts bin had 300-600mcd reds and greenies, like Boss used to use, 220R - 4k7 would be useful.  But in an era when 10,000mcd is nothing special, and costs a nickel, one needs a bigger range of resistances.

[davent]

When i built mine, 12 steps from 500r to 8k2, there an addition series 10k resistor i can switch in to get through the dead zone.
dave

[Mark Hammer]

Smart.  And so simple.

[rockola]

LED light is blinding the heck out of me. Is there any way to tone down the brightness?

Nobody has yet mentioned tape. It's what at least three pieces of equipment in our household have over their blinding LEDs. Electrical or even gaffer tape will work (our printer has grey gaffer tape on top of the power LED and you can still see it just fine).

[antonis]

All above well said concerning brightness attenuation but I didn't notice somebody to mention the fact that it isn't wise to run an LED out of its "nominal" current rating.. 
(not only from power consumption viewpoint but from LED's lifetime also..)

[BJM]

it really does pay to test brights BEFORE soldering, because of the number of times the led in question turns out to not follow the lead identifing convention you most use - ie the die isn't on the anvil, or there is no body flat, or the long leg isn't anode.

You can test a LED quite simple with a multimeter (lowest resistance setting) without destroying it. You can't see how bright it is but you can identify cathode/anode.