Interesting RGB LED project and issues.
Archiver
Posts: 46,084
I'm working on an interesting project that is going to be my first
time using a microcontroller. The need I have is to be able to control
50 points of light which must, primarily, be extremely responsive (200
updates/second), and secondarily, be bright (yes, vague) and have an
accurate color rendition (yes, subjective). For these reasons, I have
chosen LEDs as the light source.
My first question deals with the LED selection. I am tempted to go for
an RGB LED, but I understand that at least some of the older LEDs
leave a white that isn't well mixed. Is this still an issue today? Or
should I go for an RGB LED plus a white LED? Or individual
red,green,blue LEDs and then give them my own frosted cover and hope
they mix well? Or individuals plus a white? AAIIIEEEEE!
Happily, I believe I have found what appears to be the most
appropriate chip to control them. Earlier in this forum, someone
recommended the MAX7219 for RGB control. I have to say that this is a
terrible idea. You may be able to control the intensity for a group of
8 LEDs, but not each individual RGB level of each individual LED.
What I did find was the MAX7313. It can individually control 16 LEDs
with 16 intensity levels for all LEDs, and then 16 intensity levels (4
intensity bits) for each individually addressable LED (using PWM
functionality provided by the chip itself). It uses the 2-wire
I2C/SMBus interface, and can be configured for a maximum of 64 of
these devices on a single bus. (With 64 of the chips, you have a
maximum of 1024 brightness controls.) Groups of 16 individual LED
levels can be set with a single write operation of 8 bytes, which
makes for very quick updates.
Earlier, I mentioned it needed to do 200 updates per second. The
updates aren't calculated real time but based on a predetermined
pattern (or patterns). I'm likely to chose something like a 24AA512
for the long term storage, and to pull in groups of 8 (or more) bytes
into RAM on the microcontroller and feed that to the MAX7313. The
24AA512 has a hefty amount of storage and multiple devices can sit on
a single bus, but I was wondering if anyone had any experience or
opinions on these, too.
I have chosen the Parallax BS2P24 with the BS2p40 Demo Board as a base
for the project. (Specifically for the easy and supported I2C
functionality as well as general level of support and the decent
specs.)
Any comments out there? Any pitfalls I'm heading into I should avoid?
(Like putting too many devices on an I2C bus, or the RGB LED issues.)
Thanks.
time using a microcontroller. The need I have is to be able to control
50 points of light which must, primarily, be extremely responsive (200
updates/second), and secondarily, be bright (yes, vague) and have an
accurate color rendition (yes, subjective). For these reasons, I have
chosen LEDs as the light source.
My first question deals with the LED selection. I am tempted to go for
an RGB LED, but I understand that at least some of the older LEDs
leave a white that isn't well mixed. Is this still an issue today? Or
should I go for an RGB LED plus a white LED? Or individual
red,green,blue LEDs and then give them my own frosted cover and hope
they mix well? Or individuals plus a white? AAIIIEEEEE!
Happily, I believe I have found what appears to be the most
appropriate chip to control them. Earlier in this forum, someone
recommended the MAX7219 for RGB control. I have to say that this is a
terrible idea. You may be able to control the intensity for a group of
8 LEDs, but not each individual RGB level of each individual LED.
What I did find was the MAX7313. It can individually control 16 LEDs
with 16 intensity levels for all LEDs, and then 16 intensity levels (4
intensity bits) for each individually addressable LED (using PWM
functionality provided by the chip itself). It uses the 2-wire
I2C/SMBus interface, and can be configured for a maximum of 64 of
these devices on a single bus. (With 64 of the chips, you have a
maximum of 1024 brightness controls.) Groups of 16 individual LED
levels can be set with a single write operation of 8 bytes, which
makes for very quick updates.
Earlier, I mentioned it needed to do 200 updates per second. The
updates aren't calculated real time but based on a predetermined
pattern (or patterns). I'm likely to chose something like a 24AA512
for the long term storage, and to pull in groups of 8 (or more) bytes
into RAM on the microcontroller and feed that to the MAX7313. The
24AA512 has a hefty amount of storage and multiple devices can sit on
a single bus, but I was wondering if anyone had any experience or
opinions on these, too.
I have chosen the Parallax BS2P24 with the BS2p40 Demo Board as a base
for the project. (Specifically for the easy and supported I2C
functionality as well as general level of support and the decent
specs.)
Any comments out there? Any pitfalls I'm heading into I should avoid?
(Like putting too many devices on an I2C bus, or the RGB LED issues.)
Thanks.
Comments
>I'm working on an interesting project that is going to be my first
>time using a microcontroller. The need I have is to be able to control
>50 points of light which must, primarily, be extremely responsive (200
>updates/second), and secondarily, be bright (yes, vague) and have an
>accurate color rendition (yes, subjective). For these reasons, I have
>chosen LEDs as the light source.
>
>My first question deals with the LED selection. I am tempted to go for
>an RGB LED, but I understand that at least some of the older LEDs
>leave a white that isn't well mixed. Is this still an issue today? Or
>should I go for an RGB LED plus a white LED? Or individual
>red,green,blue LEDs and then give them my own frosted cover and hope
>they mix well? Or individuals plus a white? AAIIIEEEEE!
Here is a web site that should answer all of your LED-specific questions:
http://ledmuseum.home.att.net/
Regards,
Bruce Bates
<snip>
>My first question deals with the LED selection. I am tempted to go for
>an RGB LED, but I understand that at least some of the older LEDs
>leave a white that isn't well mixed. Is this still an issue today? Or
>should I go for an RGB LED plus a white LED? Or individual
>red,green,blue LEDs and then give them my own frosted cover and hope
>they mix well? Or individuals plus a white? AAIIIEEEEE!
<snip>
For info about LED's I would start with Don Klipsteins site.
Here's a page about mixing RGB to get white:
http://members.misty.com/don/ledrgb2w.html
And more general LED info:
http://members.misty.com/don/ledx.html
Steve D.
[noparse][[/noparse]Non-text portions of this message have been removed]
> For info about LED's I would start with Don Klipsteins site.
> Here's a page about mixing RGB to get white:
> http://members.misty.com/don/ledrgb2w.html
Thanks for the page. That *was* one of the detials I was going to have
to work out (the correct levels of each color in order to get white).
But was I was wondering was more of a different question.... inside of
an RGB LED itself, when you have red, green, blue on (and at the
appropriate levels), how well do the colors mix together to make
white? Or does it look a splotch of red, a splotch of green, and a
splotch of blue?
that even in that small surface mount package i get blotches of green, red,
and blue. but you may want to check it out for yourself.
www.bivar.com
Regards,
Gary D
Qualcomm Inc.
At 09:00 AM 3/4/2004, you wrote:
>--- In basicstamps@yahoogroups.com, laurasdog@w... wrote:
>
> > For info about LED's I would start with Don Klipsteins site.
> > Here's a page about mixing RGB to get white:
> > http://members.misty.com/don/ledrgb2w.html
>
>Thanks for the page. That *was* one of the detials I was going to have
>to work out (the correct levels of each color in order to get white).
>But was I was wondering was more of a different question.... inside of
>an RGB LED itself, when you have red, green, blue on (and at the
>appropriate levels), how well do the colors mix together to make
>white? Or does it look a splotch of red, a splotch of green, and a
>splotch of blue?
>
>
>
>To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
>from the same email address that you subscribed. Text in the Subject and
>Body of the message will be ignored.
>
>Yahoo! Groups Links
>
>
>
>