Searching \ for '[pic] 16f88 PWM LED control' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/ios.htm?key=pwm
Search entire site for: '16f88 PWM LED control'.

Exact match. Not showing close matches.
PICList Thread
'[pic] 16f88 PWM LED control'
2005\04\03@044249 by Chris Williams

picon face
I'm plaing to use a pic16f88 with the internal osc to do pwm to control a
single RGB led, I'm not sure how to even begin on this code. if someone has
some examples I woulc use to get started, I would very much appriate it.

thanks.

2005\04\03@053217 by Russell McMahon

face
flavicon
face
> I'm plaing to use a pic16f88 with the internal osc to do pwm to
> control a
> single RGB led, I'm not sure how to even begin on this code. if
> someone has
> some examples I woulc use to get started, I would very much appriate
> it.

In you favourite language implement the following.

Do forever

   Adjust Vlimit here if desired (or almost anywhere else)

   For Vcount = 1 to Vcountmax ' Vcount steps once each loop
                                                       ' and resets
when max reached

       if count <= Vlimit then
           turn LED on
       else
           turn LED off
       endif

       Optional delay

   Next

Loop

LED will be lit for the ratio Vlimit/Vcountmax
Increasing Vlimit increases on time.
The "For" loop can be accessed once per timer tick or after a delay or
just run flat out. If running flat out there will be some "dead time"
while the code outside the For/Next loop is run.



       RM

2005\04\03@073847 by Jan-Erik Soderholm

face picon face
Chris Williams wrote :

> > I'm plaing to use a pic16f88 with the internal osc to do pwm to
> > control a single RGB led,...


And Russell McMahon wrote :

> In you favourite language implement the following.
> ...
>         if count <= Vlimit then
>             turn LED on
>         else
>             turn LED off
>         endif

Now, controling a RGB LED is a little more "involved" then
just turing a single LED on/off, isn't it ? :-)

To Chris :

Could you be more specific about how you'd like
to control the LED ? What about colors ?
How many "steps" do you need for each LED (R, G
and B) ? Maybe just a few distinct colors ? Are you using
a common cathod or common anode RGB LED ?

Jan-Erik.



2005\04\03@074620 by Chris Williams

picon face
Heres the LED I'm using lsdiodes.com/multi-color/5mmfullspectrum.htm
I would like to randomly cycles though all possible colors if possible.

{Original Message removed}

2005\04\03@083146 by Jan-Erik Soderholm

face picon face
Chris Williams wrote :

> Heres the LED I'm using
> http://lsdiodes.com/multi-color/5mmfullspectrum.htm

OK. Now, I don't like that page. Quote : "Keep in mind that while
green and blue require around thew same voltage, applying the
same amount [of voltage] to the red will fry it..."

Now, *NO* LED requires any specific voltage. They require
a specific *current*, normaly max 20 mA. Amateurs...

Common cathod meens that you can't (easily) drive the LED
from a higher voltage sourcethen the PIC is running from.
Common anode is easier to use in that regard.

> I would like to randomly cycles though all possible colors if
> possible.

There is no such thing as "all possible colors" !
There are an unlimitied, infinite number of colors.
You have to make a design decission here, how many
different colors do you want ? You absolutly **HAVE**
to decide that first.

Regards,
Jan-Erik.



2005\04\03@112342 by Bob Ammerman

picon face
Jan-Erik opined...

> Common [cathode] meens that you can't (easily) drive the LED
> from a higher voltage source [than] the PIC is running from.
> Common anode is easier to use in that regard.

Sure you can. You just have to think upside down. Let us assume you have un
unregulated 6V supply. Instead of using a positive regulator to create +5,
use a negative regulator to create -5, relative to the positive terminal of
the unregulated supply. The PIC connects between that -5 (as gnd) and the
positive side of the unregulated supply (as +5). The LED's common cathode
connects to the negative of the unregulated supply.

Bob Ammerman
RAm Systems



2005\04\03@122014 by Byron A Jeff

face picon face
On Sun, Apr 03, 2005 at 02:31:45PM +0200, Jan-Erik Soderholm wrote:
> Chris Williams wrote :
>
> > Heres the LED I'm using
> > lsdiodes.com/multi-color/5mmfullspectrum.htm
>
> OK. Now, I don't like that page. Quote : "Keep in mind that while
> green and blue require around thew same voltage, applying the
> same amount [of voltage] to the red will fry it..."
>
> Now, *NO* LED requires any specific voltage. They require
> a specific *current*, normaly max 20 mA. Amateurs...
>

True to a point. The LED does requirea minimum voltage to light.
The forward voltage is the mount of voltage the LED will drop
when you run the specified current through it.

But 1V will not light an LED that requires 2.1V forward voltage.

I've run LEDs at 120VAC. But I used a 10K resistor to drop the
max current to 12mA.

> > I would like to randomly cycles though all possible colors if
> > possible.
>
> There is no such thing as "all possible colors" !
> There are an unlimitied, infinite number of colors.
> You have to make a design decission here, how many
> different colors do you want ? You absolutly **HAVE**
> to decide that first.

It's a matter of precision because PWM will have a discrete number
of values. 10 bits of precision on each of the three channels will
give you 1Gig of potential colors. Many will be below the just
noticeable difference threshold. For simplicity I'd start with 8
bits per channel.

BAJ

2005\04\03@124219 by Lindy Mayfield

flavicon
face
I was curious about these LED's:  Can you mix colors?

-----Original Message-----
From: spam_OUTpiclist-bouncesTakeThisOuTspammit.edu [.....piclist-bouncesKILLspamspam@spam@mit.edu] On Behalf Of Bob Ammerman
Sent: Sunday, April 03, 2005 17:23
To: Microcontroller discussion list - Public.
Subject: Re: [pic] 16f88 PWM LED control

Jan-Erik opined...

> Common [cathode] meens that you can't (easily) drive the LED
> from a higher voltage source [than] the PIC is running from.
> Common anode is easier to use in that regard.

Sure you can. You just have to think upside down. Let us assume you have un
unregulated 6V supply. Instead of using a positive regulator to create +5,
use a negative regulator to create -5, relative to the positive terminal of
the unregulated supply. The PIC connects between that -5 (as gnd) and the
positive side of the unregulated supply (as +5). The LED's common cathode
connects to the negative of the unregulated supply.

Bob Ammerman
RAm Systems



2005\04\03@125352 by Jan-Erik Soderholm

face picon face
Lindy Mayfield wrote :

> I was curious about these LED's:  Can you mix colors?

Depends on what you mean with "mix colors".

The RGB LED can produce one single color at a time.
That color can be any any color that any combination
of the three LED:s (red, green and blue) can produce.
There might be some specific colors that ca not be made
with just red, green and blue, I'm not sure. What about
brown ?

If you put *two* RGB LED:s together you would be able to
produce two different colors that could me "mixed", in a way...

Jan-Erik.



2005\04\03@130051 by Lindy Mayfield

flavicon
face
I meant to add current to the Red lead and the Blue one, for example.  Would it produce a different color.  

{Original Message removed}

2005\04\03@131807 by Jan-Erik Soderholm

face picon face
Lindy Mayfield wrote :

> I meant to add current to the Red lead and the Blue one, for
> example.  Would it produce a different color.

Different from pure red and pure blue, yes.

red+blue will give you some violet color, I'd guess.
Any color from red over voilet to blue depending on the
amount of current (or the PWM pulse ratio) on the
red and blue led.

But note that the RGB led is made from three separate
led-chips in the same round housing. So a clear housing
will normaly produce one red and one blue spot (partly overlapped)
instead of one single violet spot. And the green led lights in another
slightly different directions. I like the "diffused" RGB leds better,
since they tend to give a more even "mixed" light.

Jan-Erik.




2005\04\03@172104 by Jinx

face picon face

> I'm plaing to use a pic16f88 with the internal osc to do pwm to control a
> single RGB led, I'm not sure how to even begin on this code. if someone
> has some examples I woulc use to get started, I would very much appriate
> it.

Check out the thread

[PIC:] RGB Color mixing or randomization

in the first week of February (4th onwards)

OP of that thread mentions getting 24-bit colours using an F88

2005\04\03@183550 by Hector Martin

flavicon
face
RGB LEDs can produce any color inside their spectrum, the same as your
computer monitor. It is basically limited by how pure the red, green,
blue are, and the maximum intensity of each.

I'm thinking "all possible colors" means any color. Thing is, you can
do several things:

-No PWM, plain Red,Green,Blue only, cycling all three.
-No PWM, combinations: Red,Green,Blue,Cyan(Blue+Green)'light
blue',Magenta(Blue+Red)'purple',Yellow(Red+Green). Plus white and/or
black if you like. Total 8 colors. This is just 2^3.
-PWM, middle intensities (half red, etc). Many more colors (3^3=27).
-PWM, more intensities. You can do 4 intensities each
(0%,33%,66%,100%) which gives you 4^3=64 colors.

etc all going up to what would be hicolor in PC-speak (32768 colors,
which is 5bits red, 5bit blue, 5bits red (32 colors each, 32768 colors
total)
Or truecolor, millions of colors (8bits each, 256 possible intensities
each, total of 16777216 colors).

The program for N-colors PWM (N colors for each channel, not total)
for RGB would be something like

each timer tick/overflow/interrupt/whatever
   increment PWMCnt
   if PWMCnt>=(Ncolors-1)
       PWMCnt=0
       Rvalue2=Rvalue
       Gvalue2=Gvalue
       Bvalue2=Bvalue
   endif

   if Rvalue2>PWNCnt
       Rpin=1
   else
       Rpin=0
   endif

   if Gvalue2>PWNCnt
       Gpin=1
   else
       Gpin=0
   endif

   if Bvalue2>PWNCnt
       Bpin=1
   else
       Bpin=0
   endif

return

And then, in the main loop, run something that changes Bvalue, Rvalue,
etc. Note that I cache the value to prevent changes during a PWM cycle.

Try with several values for the timing. For example, for a 256 values
per channel PWM, try something like each 1ms and go lower until the
flickering disappears. Usually 64 values per channel will suffice, and
will flicker much less and won't need the loop run so often.

To cycle through all colors (hues, no black or white in this example,
no brightness or saturation, just hues) you can run something in the
main loop that does, for example:

(start with red=100%,blue=0%,green=0%)
repeat forever:
   (total color is now red)
   ramp green from 0% to 100%
   (total color is now yellow)
   ramp red from 100% to 0%
   (total color is now green)
   ramp blue from 0% to 100%
   (total color is now cyan)
   ramp green from 100% to 0%
   (total color is now blue)
   ramp red from 0% to 100%
   (total color is now magenta)
   ramp blue from 100% to 0%
   (total color is red again)
end repeat

note that after a ramp the color remains at the end value, i.e. "ramp
red from 100% to 0%" does NOT mean touch the blue or green channels,
just leave them as they were after the last ramp. If you want to avoid
timers, and you do not intend to do anything else other than cycle the
LED, you can do something like the following, in which case you can
also remove the caching of the values in the PWM subroutine:

subroutine ramp(*channel,from,to):
   *channel=from
   repeat until *channel>=to:
       increment *channel
       repeat N times: (N depends on the speed on the effect, how
much each separate color stays before changing into the next tick, in
units of PWM cycles)
           call PWM subroutine
           delay (this is the PWM delay)
       end repeat
   end repeat
end subroutine

main program:
repeat forever:
   ramp(&Gvalue,0,max_color)
   ramp(&Rvalue,max_color,0)
   ramp(&Bvalue,0,max_color)
   ramp(&Gvalue,max_color,0)
   ramp(&Rvalue,0,max_color)
   ramp(&Bvalue,max_color,0)
end repeat

(where max_color equals Ncolors-1)

Note that I use */& the C way, since passing the value to the
subroutine won't do any good, you want to be able to CHANGE the value.
Or make 3 ramp subroutines, one for each color, or pass a value 0 for
red, 1 for green, 2 for blue, or whatever.

Of course the purity of the colors will be affected by the different
intensities. I recommend hooking up the outputs from the PIC to some
transistors and adding a pot to the base to be able to adjust
individual brightness.

HTH,

--
Hector Martin (hectorspamKILLspammarcansoft.com)
Public Key: http://www.marcansoft.com/hector.asc

2005\04\03@192733 by Russell McMahon

face
flavicon
face
>> > I'm plaing to use a pic16f88 with the internal osc to do pwm to
>> > control a single RGB led,...

> Now, controling a RGB LED is a little more "involved" then
> just turing a single LED on/off, isn't it ? :-)

Indeed. A little. You need to nest 3 at once for full sequence

   For Vred = 1 to max
       For Vblue = 1 to max
           For Vgreen = 1 to max
               If Vred < RedLimit then RedLED  = on else RedLED = off
               If Vgreen < GreenLimit then GreenLED  = on ......
               If Vblue < BlueLimit then BlueLED  = on ....
           Next
       Next
   Next

This would cycle through all colours BUT probably in an unsatisfying
way. Changing the red/gree/blue nesting would give some other
unstaisfying way :-)

What is the OBJECT of the exercise?

AFAIR there was a thread on exactly this topic s few months ago.

Yep ...

Michael Brown,  Friday, 4 February 2005 2:09 a.m.

Wrote:

Hello all,
I've been tinkering with a 16F88 and some RGB (actually RGBB) LEDs.  I
can successfully PWM the LED at 256 different linear duty cycles per
color resulting in 24-bit color possibilities.  This looks pretty good
when ramping the colors up and down.

Now the crux of the problem.  I would like to vary all three colors
continuously so that I end up with a good assortment of mixes.  The
problem is that no matter what kind of algorithms I come up with they
either take way to long to cycle thru or they repeat and miss allot of
colors I'd like to see.  I can change LED colors up to every 13mS (one
full period time) without causing ugly flashing.  To hit all 16
million
possible values would take about 60 hours, so this is not practical.

_______________

Semifinal post on this included

From: "Jan-Erik Soderholm" <.....jan-erik.soderholmKILLspamspam.....telia.com>
> michael brown wrote :

> > Thanks Bob, but I've decided on a scheme that I like.  It's
> > giving me a virtual rainbow of nicely saturated colors...
>
> Was that the overlapping read/gren/blue up/down ramps you
> described in an eariler post ?

Yes.  Did you find some LEDs you liked?  I need to get some more that
have diffused lenses.

       RM

2005\04\04@041630 by Russell McMahon

face
flavicon
face
>> Common [cathode] meens that you can't (easily) drive the LED
>> from a higher voltage source [than] the PIC is running from.
>> Common anode is easier to use in that regard.

> Sure you can. You just have to think upside down. Let us assume you
> have un
> unregulated 6V supply. Instead of using a positive regulator to
> create +5,
> use a negative regulator to create -5, relative to the positive
> terminal of
> the unregulated supply. The PIC connects between that -5 (as gnd)
> and the
> positive side of the unregulated supply (as +5). The LED's common
> cathode
> connects to the negative of the unregulated supply.

Both these comments reflect a dangerous (to the health of the PIC)
application area.
You CAN do what they are implying but you need to take care to design
right or (PIC) death or worse may occur.

If you run a PIC from V1 and drive the, say, cathode of a common anode
LED (or the anode of a common cathode LED using the second method)
from Vhi where Vhi > Vdd then.

- When the LED is on the PIC pin is at a fraction of a volt wrt to
Vss.

- When the LED is off the PIC notionally sees Vhi on it's output PIN.
This will be a very high impedance Vhi due to the LED resistance. Any
series resistor will further reduce what is seen somewhat. If Vhi is
very high (eg 120V behind 10k) then the LED will not turn off and LED
current will flow in the protection diode. If Vhi is say 6v (as
suggested) with Vdd = 5v the 1v difference will cause a very small
voltage to flow in the LED. Vpicpin will tend towards Vdd + 0.6v. For
a white LED the current will be very small. For some red LEDs the
current will be much higher and the LED may "glimmer" but be
acceptably dim. If Vhi is 7 volts the situation will be worse again.

At some points in this saga some PICs will start to misbehave randomly
sometimes. The PIC is being run "out of spec" and anything MIGHT
happen. If you do this, know what you are doing!!!



       Russell McMahon

2005\04\04@050302 by ThePicMan

flavicon
face

>> Now, *NO* LED requires any specific voltage. They require
>> a specific *current*, normaly max 20 mA. Amateurs...
>>
>
>True to a point. The LED does requirea minimum voltage to light.
>The forward voltage is the mount of voltage the LED will drop
>when you run the specified current through it.
>
>But 1V will not light an LED that requires 2.1V forward voltage.
>
>I've run LEDs at 120VAC. But I used a 10K resistor to drop the
>max current to 12mA.

What about max inverse voltage? If you put 120VAC (here the "through
10Kohm" doesn't matter) you sure trigger the diode reverse voltage
limits. The 12mA inverse current may well destroy it.

2005\04\04@051337 by Jan-Erik Soderholm

face picon face
ThePicMan wrote :

> >I've run LEDs at 120VAC. But I used a 10K resistor to drop the
> >max current to 12mA.
>
> What about max inverse voltage? If you put 120VAC (here the "through
> 10Kohm" doesn't matter) you sure trigger the diode reverse voltage
> limits. The 12mA inverse current may well destroy it.

Maybe put a standard diod in the other direction in parallel with the
LED to limit the inverse voltage over the LED. Or maybe
a second LED, so you'd "use" both periods of the AC.

Jan-Erik.



2005\04\04@051426 by Jan-Erik Soderholm

face picon face
Russell McMahon wrote :

> Sent: den 4 april 2005 10:16
> To: Microcontroller discussion list - Public.
> Subject: Re: [pic] 16f88 PWM LED control
>
>
> >> Common [cathode] meens that you can't (easily) drive the LED
> >> from a higher voltage source [than] the PIC is running from.
> >> Common anode is easier to use in that regard.
>
> > Sure you can. You just have to think upside down. Let us assume you
> > have un unregulated 6V supply. Instead of using a positive regulator
> > to create +5, use a negative regulator to create -5, relative to the
> > positive terminal of the unregulated supply. The PIC connects
> > between that -5 (as gnd) and the positive side of the unregulated
> > supply (as +5). The LED's common cathode
> > connects to the negative of the unregulated supply.
>
> Both these comments reflect a dangerous (to the health of the PIC)
> application area.

[snip...]

Yes, you are right, of course. The LED will feed the higher woltage
back into the PIC pin when off. But, I never said (at least never ment
:-) )
that you should drive the LED directly from the PIC pin. What I ment was
that (if you'd like to drive an RGB LED from e.g 12 V), it's easier
to drive a low-side switching transistor using a common-anode
RGB LED, then to drive a high-side switching transistor using the
(more common as it seems) common cathod RGB LEDs in which
case you might have to use special "high-side drivers".

The reason to drive the RGB LEDs from a higher voltage (then 5V),
is of course to even out process variations in forward voltage drop.
If builing multiple items, it easier to get an even light level between
them. And maybe your PIC is only run from (e.g.) 3 V, in which case
you have to run your LEDs from ta higher voltage anyway (or use one
of the gazillions of LED-drivers from Maxim, few of them available in
DIP packing, of course... :-) ).

Jan-Erik.



2005\04\04@055850 by Jinx

face picon face
> > What about max inverse voltage? If you put 120VAC

> Maybe put a standard diod in the other direction in parallel with
> the LED to limit the inverse voltage over the LED. Or maybe a
> second LED, so you'd "use" both periods of the AC

With opto-LEDs that makes a good, safe zero crossing detector
or mains time-base reference

http://home.clear.net.nz/pages/joecolquitt/zc-detect-opto.gif

2005\04\04@061219 by Jan-Erik Soderholm

face picon face
Jinx wrote :

> With opto-LEDs that makes a good, safe zero crossing detector
> or mains time-base reference
>
> http://home.clear.net.nz/pages/joecolquitt/zc-detect-opto.gif

Or use a "AC opto-coupler" for fewer components. :-)
They are made exactly for this purpose.

Jan-Erik.



2005\04\04@085920 by Byron A Jeff

face picon face
On Mon, Apr 04, 2005 at 09:05:23AM +0100, ThePicMan wrote:
>
> >> Now, *NO* LED requires any specific voltage. They require
> >> a specific *current*, normaly max 20 mA. Amateurs...
> >>
> >
> >True to a point. The LED does requirea minimum voltage to light.
> >The forward voltage is the mount of voltage the LED will drop
> >when you run the specified current through it.
> >
> >But 1V will not light an LED that requires 2.1V forward voltage.
> >
> >I've run LEDs at 120VAC. But I used a 10K resistor to drop the
> >max current to 12mA.
>
> What about max inverse voltage? If you put 120VAC (here the "through
> 10Kohm" doesn't matter) you sure trigger the diode reverse voltage
> limits. The 12mA inverse current may well destroy it.

Of course. That's why you have an antiparallel diode that conducts in
the opposite direction. Or you can rectify the AC into DC.

The point is that you can shove amost any voltage through an LED as
long as the current is properly limited. The forward voltage defines
a lower bound of the voltage required.

One last thing: even if the voltage is lower than the forward voltage you
still have to current limit it. I'm sure that 1V@1A will release the magic
smoke from the LED.

BAJ

2005\04\04@090708 by Byron A Jeff

face picon face
On Mon, Apr 04, 2005 at 09:58:43PM +1200, Jinx wrote:
> > > What about max inverse voltage? If you put 120VAC
>
> > Maybe put a standard diod in the other direction in parallel with
> > the LED to limit the inverse voltage over the LED. Or maybe a
> > second LED, so you'd "use" both periods of the AC
>
> With opto-LEDs that makes a good, safe zero crossing detector
> or mains time-base reference
>
> http://home.clear.net.nz/pages/joecolquitt/zc-detect-opto.gif

That's esentially what I was using it for. I was testing the
possibility of using cheap motion sensing lights in a homebrew
security setup. I know that it's possible to tear down to the PIR
head. But I liked the structure of the lights with the stems
removed.

Obviously the lights got 120 VAC when the sensor switched on.
So I antiparalleled a local visible LED with an opto switching
12VDC to the console. I used a 10K 2W resistor for current limiting.

Worked like a champ.

BAJ

2005\04\04@121844 by Russell McMahon

face
flavicon
face
> One last thing: even if the voltage is lower than the forward
> voltage you
> still have to current limit it. I'm sure that 1V@1A will release the
> magic
> smoke from the LED.

I think you probably didn't mean that.
If Vapplied is below Vforward then you will get minimal current flow.
(Almost) any LED with 1 amp forward current flowing will be biased in
the operating range :-).

       RM

2005\04\04@123732 by Danny Decell

picon face
>
> One last thing: even if the voltage is lower than the forward voltage you
> still have to current limit it. I'm sure that 1V@1A will release the magic
> smoke from the LED.
>

How could you do that BAJ ? Pump 1 amp into a LED if the forward voltage
requirement was higher than the 1 volt? Don't you need 1 ohm of resistance
and since the forward voltage has not been met your LED resistance is going
to be way above 1 ohm. Getting the LED to draw 1 Amp from 1 volt is
something I can't get my head around :-) Help help, help me learn new ways
to fry my LEDS :)

- Dan


2005\04\04@125919 by Mark Rages

face picon face
On Apr 4, 2005 11:37 AM, Danny Decell <EraseMEd.decellspam_OUTspamTakeThisOuTcomcast.net> wrote:
> >
> > One last thing: even if the voltage is lower than the forward voltage you
> > still have to current limit it. I'm sure that 1V@1A will release the magic
> > smoke from the LED.
> >
>
> How could you do that BAJ ? Pump 1 amp into a LED if the forward voltage
> requirement was higher than the 1 volt? Don't you need 1 ohm of resistance
> and since the forward voltage has not been met your LED resistance is going
> to be way above 1 ohm. Getting the LED to draw 1 Amp from 1 volt is
> something I can't get my head around :-) Help help, help me learn new ways
> to fry my LEDS :)
>
> - Dan
>

The LED forward votage falls with temperature.   Maybe if you *really*
cooked the LED, you could get it to draw current at 1V.  But the LED
would probably be pre-fried by the heat before that point.

I went looking for LED voltage vs. temp curve and found this:
www.naisweb.com/e/relaye/semi_eng/semi_eng_gu15/idaarax.html
(see fig. 7).  Also, Bob Pease has a bunch of curves in
"Troubleshooting Analog Circuits."  My copy's not handy so I won't
give the page number.

Back on topic, this temperature sensitivity is why you set an LED's
current rather than give it a fixed voltage.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2005\04\04@133028 by Byron A Jeff

face picon face
On Tue, Apr 05, 2005 at 04:05:08AM +1200, Russell McMahon wrote:
> >One last thing: even if the voltage is lower than the forward
> >voltage you
> >still have to current limit it. I'm sure that 1V@1A will release the
> >magic
> >smoke from the LED.
>
> I think you probably didn't mean that.
> If Vapplied is below Vforward then you will get minimal current flow.
> (Almost) any LED with 1 amp forward current flowing will be biased in
> the operating range :-).

It's such an odd condition that I guess I was speculating.

Sorry.

BAJ

2005\04\04@144001 by Danny Decell

picon face
NO apology needed, Mark sent a link of some graphs that show how forward
voltage drop can decrease with rise in ambient temperature. Although I
always current limit my led based components, I still learned something.
Thank you!

Dan

{Original Message removed}

2005\04\07@040953 by Alan B. Pearce

face picon face
>Mixing:
>  Opaque silicon sealants are very useful.

I have been thinking in terms of making some art deco type fittings using
frosted glass or Perspex, to give diffused mixed light. Figured this should
work equally well.

2005\04\08@121306 by Tony Smith

picon face
> I have been thinking in terms of making some art deco type fittings using
> frosted glass or Perspex, to give diffused mixed light. Figured
> this should
> work equally well.
>


Have a look on the Agilent website for a document on light pipes.  A rather
interesting read, althought drill a hole in a LED to insert a fibre optic
cable is probably a bit much.  Let me know if you can find it, I'll send you
a copy.

Tony

2005\04\08@171117 by Marcel Duchamp

picon face

>>I have been thinking in terms of making some art deco type fittings using
>>frosted glass or Perspex, to give diffused mixed light. Figured
>>this should
>>work equally well.

It does. Something I've tried that worked well for me are glass
insulators as were once common on telegraph poles.  Common types can be
had for less than $1US each.  I bought several and sand-blasted the
insides to diffuse the light.  A 3 color RGB led was used and driven
with 3 different frequency sine waves.  The slow pulsating colors,
randomly changing over time, produced the effect I was looking for.

Regarding the discussion earlier on pwm resolution, this topic will be
in the same category as CD quality vs. analog tape vs analog vinyl, ad
nauseum.  Each person will have to decide what they like.  For my
project, I'm generating 8 bit pwm and when a given color led gets down
to a few counts, the steps are very obvious to me.  I limit any given
led to a minimum of 1 count; going to zero is a sharp discontinuity, at
least up close.  Back a ways, you won't see it. No interrupts were
required; I'm getting it done with a PIC10F206 driving 3 leds.
MD

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...