"fade on" at startup

GooeyGus · Jun 29, 2008

Hi guys... I was curious about how one could make a diode 'fade' on to full power over the course of about 1 - 2 seconds. I'm guessing one could do it with some sort of capacitor... but I tried wiring one up a couple different ways and it didn't' seem to work. I think it would be a cool effect, and may also prevent severely over-driven diodes from failing due to the harshness of start-up.

Any ideas?

Maelstrom · Jun 29, 2008

Ask Dr.Lava how he did it on his circuit

styropyro · Jun 29, 2008

I think you could use a pretty big capacitor in parallel with the batteries (before the driver) , but it would suck down the batteries and the laser would take a while to dim out after you shut it off

woop · Jun 29, 2008

I would say that Drlava's flexdrive has that soft start feature built into the chip.
I don't think it is important to have a soft start of 1s... laser diodes are designed to be pulsed anyway. turning them on in a fraction of a second isn't going to hurt them

I find it annoying when a laser doesn't turn on full power straight away. but thats just me

rkcstr · Jun 29, 2008

Quick google search found this article on soft-start circuits for LDO linear regulators... didn't really look too closely, but there may be something useful in it:

http://www.powersystems.eetchina.com/PDF/2007DEC/PSCOL_2007DEC21_DROP_TA_100.PDF?SOURCES=DOWNLOAD

GooeyGus · Jun 29, 2008

Yes Dr. Lava's circuit has soft start built into the IC itself. BUT what his does isn't really what I'm talking about. His is a fast soft-start. you cant really see it fade on, it just takes a second before it turns on. I want to be able to actually see it fade on. I'm not doing it for any sort of diode preservation, I just think it will look cool.

woop · Jun 29, 2008

well you could put a huge capacitor across the LD. at first it would take all the current from the LD, then as it charges, more current would go to the LD until its charged.
that would require a pretty big cap though.

otherwise i can't think of a simple way without interfering with the actual current regulation
you could PWM the enable pin of your chip, if it has one

unfortunately that link is for voltage regulators, not current regulators

GooeyGus · Jun 30, 2008

How huge are we talking here? ;D

think a 16v/1000uf might give me what I'm looking for? Size isn't much of an issue, as its not going to be in a pointer.

woop · Jun 30, 2008

GooeyGus said:
How huge are we talking here? ;D

think a 16v/1000uf might give me what I'm looking for? Size isn't much of an issue, as its not going to be in a pointer.

depending on what currents we are talking about, i think you would need a much larger cap for any visible fading effect.
try it out and see.

likewhat · Jun 30, 2008

if you want it to fade on you should choose a capacitor such that R*C is a few seconds. So take the amount of time you want it to fade on for, say 3 seconds, and divide 3 by R to get what C you need to make it happen.

rkcstr · Jun 30, 2008

I have some 50000uf 16V caps, they're just slight skinnier and taller than a can of coke, though ;D

I'm sure there is a small circuit that would do it for you, though. Just have to find/design it

you'd probably need something to slowly raise voltage on the input side to avoid interfering with current regulation, or would produce a variable setting resistance on startup.

GooeyGus · Jul 1, 2008

rkcstr said:
I have some 50000uf 16V caps, they're just slight skinnier and taller than a can of coke, though ;D

I'm sure there is a small circuit that would do it for you, though. Just have to find/design it you'd probably need something to slowly raise voltage on the input side to avoid interfering with current regulation, or would produce a variable setting resistance on startup.

This is what I was thinking. Slowly raise voltage until the driver has enough to regulate. This wouldn't interfere with current regulation, and would provide the desired effect. Now I just gotta figure out how to do it ;D

pseudolobster · Jul 1, 2008

they're doing neat things with aerogel capacitors these days.. you can get multi-farad capacitors smaller than an aixiz module...
for instance, in every xbox there is a 2v, 1f aerogel cap that's something like 5mm diameter x 12mm... mind you a 2v capacitor won't help you in this application, but still...

other than the analog approach, you could always do something with a microcontroller or a 555 and a couple transistors to make something that pulses faster and faster till it reaches CW...

woop · Jul 1, 2008

you could use an RC circuit connected to the base of a transistor to slowly turn the voltage up. attach a linear driver to that and it would work
it would only require a cap, resistor and a power transistor.

GooeyGus · Jul 1, 2008

woop said:
you could use an RC circuit connected to the base of a transistor to slowly turn the voltage up. attach a linear driver to that and it would work
it would only require a cap, resistor and a power transistor.

details? To be brutally honest I dont know a damn thing about transistors, other than their basic function.

woop · Jul 1, 2008

something like this should work.
using a n-channel FET.
basically the 500K resistor charges up the cap, as it charges the voltage gets larger. a FET is a voltage controlled transistor so as the voltage on its input grows it starts to turn on, allowing current to flow. generally when the input reaches 5V the fet is completely on and basically act like a piece of wire.

replace the led and resistor with a real driver

"fade on" at startup

GooeyGus

0

Maelstrom

0

styropyro

0

woop

0

rkcstr

0

GooeyGus

0

woop

0

GooeyGus

0

woop

0

likewhat

Guest

rkcstr

0

GooeyGus

0

pseudolobster

0

woop

0

GooeyGus

0

woop

0

Attachments

Online statistics

Forum statistics

Share this page