[Lab] christmas light pcb

Richard Guy Briggs rgb at tricolour.net
Mon Dec 24 11:57:31 EST 2012


On Mon, Dec 24, 2012 at 11:41:01AM -0500, Peter Sjoberg wrote:
> On Mon, 2012-12-24 at 06:46 -0800, Richard Wiens wrote:
> > Peter,
> >  
> > If you go to the Adafruit website and look at their material on
> > addressable LED strips you should get some clues.  You basically use
> > PWM to control each LED...the kind of chip will dictate whether you
> > need to keep sending the signal or just send it once and the chip will
> > keep PWMing until you change something.
> 
> The first chip has _only_ power and out from that is 3 wires, gnd,data,
> +5. Don't know if it can be parasiting to much from the power since it
> goes directly from the 7805 to all the cards. 
> They do not blink randomly, all/sections come on blue/green/white,
> change color in sequence, chasing light and so on, so it's some
> intelligence there, and with only one wire it must be one way only or
> some single wire protocol. Come to think of it I don't think it ever
> changes backwards, I think changes always goes from the power side to
> the end. Don't know if it need to be PWM since it looks more like on/off
> then dimming so it may only need 3 outputs but once it get the signal of
> what to turn on it can probably keep that on until new orders come.
> 
> The adfruit ones (306(HL1606) & 322(WS2801)) have 4 wires so the
> protocol can't be the same. 
> Now thinking about it, since this lights starts at power on and it's no
> separate controller box (was expecting the small box in the beginning to
> have some smarts but it was just a rectifier, 7805 and a cap) they must
> be custom chip or programmable (PIC).
> What I was hoping was to be able to control the lights individually to
> make some light painting or so. After christmas I will take a scope and
> see what the data line looks like and also try to hook it up to read the
> bits (what's the safest way to read 5V with raspberry pi's 3.3V
> unbuffered input short of buying some chip?) but don't expect much
> there.

A voltage divider will easily do that.

------+
      \
      / 10k
      \
      /
5V    +----------
      \
      / 18k
      \        3.3v
      /
------+----------

Use proportionately higher values if there is too much loading on your data
line.

> > Richard
> > 
> > 
> > 
> > 
> > From: Peter Sjoberg <peters-modlab at techwiz.ca>
> > To: lab at artengine.ca 
> > Sent: Monday, December 24, 2012 3:17:55 AM
> > Subject: [Lab] christmas light pcb
> > 
> > 
> > 
> > 
> > I picked up a Christmas light at walmart and trying to find info about
> > it. What interested me is that for $8 I got 15 RGB LEDs, each with
> > it's own little pcb and it seems like they can be controlled
> > individually! Of course it's not as simple as reading the specs
> > because I haven't found any (did google the numbers found) and the IC
> > is a black blob so no help there.
> > I opened up the first and last one to take some pictures. The first
> > one has 5V feed on the upper row marked "DI", the bottom row marked
> > "DO" has 3 wires going to next light where it comes in on the "DI"
> > row, out on a "DO" and my guess is that it goes on like that to the
> > end.
> > I took some pictures of it that can be seen at
> > http://flickr.com/gp/henahadu/73436S/ 
> > My question is - anyone seen anything like it ? Any clue on where/how
> > I can find more info? Somewhere else to ask ?
> > 
> > /ps
> > 
> > 
> > 
> > 
> > 
> > 
> > _______________________________________________
> > Lab mailing list
> > Lab at artengine.ca
> > http://artengine.ca/mailman/listinfo/lab
> > 
> > 
> > 
> 
> 

> _______________________________________________
> Lab mailing list
> Lab at artengine.ca
> http://artengine.ca/mailman/listinfo/lab


	slainte mhath, RGB

--
Richard Guy Briggs               --  ~\    -- ~\            <hpv.tricolour.net>
<www.TriColour.net>                --  \___   o \@       @       Ride yer bike!
Ottawa, ON, CANADA                  --  Lo_>__M__\\/\%__\\/\%
Vote! -- <greenparty.ca>_____GTVS6#790__(*)__(*)________(*)(*)_________________



More information about the Lab mailing list