Intergalactic_Captain
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 228
Registered: 4-9-2004
Location: somewhere where i don\'t know where i am
Member Is Offline
Mood: frabjous
|
|
The Gross LED - Need some PWM pointers
Been a while since I've posted here, and this a non-chem project so it belongs in Misc. A lot of you guys are techies so I figured I'd post the
project and see if I can get some help.
Long story short, I saw the commercial for Home Depot's christmas-light tradein the other day and decided to check it out... So, I ended up walking
out the door with two 50-count, 5-color strands at $4 and change each - The particular brand I found is simply raw 5mm leds (resistors on orange and
yellow?) wired in series like old-school christmas lights. Ten each of each color - red, green, blue, orange, yellow...
So I went to radioshack, grabbed the biggest hunk of protoboard they had, and got to work... Fair warning, these are not the highest quality - Light
output is on the "high" side, but not quite "high intensity"... Molding of the lens portion sucks, as there are obvious peak and valley cones.
However, for simple things, at this price I know for a fact that I could not source so many at such a price anywhere else in town - Ordering online?
Yes... Brick and mortar? No... Not by a longshot...
So that brings us to the title... The "Gross LED" - When it's finished, it will contain 144 leds in 12x12 form - 12 g, 12 r, 12 b, 12 g, etc...
Essentially making one bigass RGB led... ![:D](./images/smilies/biggrin.gif)
...So now for the fun part... Each "row" is being built in groups of 3 - 3 led's in series with one resistor per group, each group tied in parallel
and each color row tied in parallel to form one "super group"... The math is done for this part, all said and done there will be 3 "super groups"
consuming about 12W total and ~8W at the led's...
Each "super group" will get it's own PWM signal to control brightness - Here's where it gets tricky... I plan on building a modular 3-channel drive
circuit based around TIP3055's for power handling, with generic PNP's driving them off of a 555/556 circuit...
Plan is to use two 556's to control frequency and duty cycle - half of one 556 in astable to control frequency, this triggers other half and other
556, each with cheap potentiometers to control duty cycle...
Soo... Aside from apparently taking up way more space than I expected to describe this... My main questions apply to the PWM portion - What should
be my target frequency? Most of what I can find seems to deal with high-power drivers in the KHz range - For this project I'd rather go low -
100-200Hz or so... Ideal? I don't know - Would it look cooler if it's moving? Yes - You'd be able to see the blips... So, is this doable with
556's (or 555's if I have to)? I toyed around with the idea a while ago for a motor controller, but I believe with led's I should be able to get away
with simplistic designs...
Power module should be simple - I've built more than a few... Probably go old-school rather than switchmode, 10:1 full wave off the wall to whatever
regulator I can dig up... Necessity is about 1A @ 12V out, will probably build it to handle 5A or so - Everything I do gets repurposed....
Power handling should be simple enough - As I said, TIP3055 triggered by RadioShack PNP's from the 555/6 outputs - this should invert the negative
pulse to trigger the 3055 and light my led's... This section will also be built modular - Designed for simplicity and driving any simple 12V load,
PWM in particular...
...So, if anyone has read this far, I'd like some input regarding my PWM scheme - Will it work the way I want it to? This is a shits and giggles
project - One way or another I'm going to end up with each individual module - What I'm concerned with is how to design the 555/556 section - Ideally,
100Hz and adjustable 5-95% duty cycles - Is this something I can do or am I barking up the wrong tree?
If you see me running, try to keep up.
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
In order to eliminate flicker, you want length the PWM period to be significantly shorter than the characteristic period for persistence of vision to
take hold. 200 Hz should be fine. 100 Hz stabilizes images, but since you're making a light field, the flicker "signal" may be more prominent, so I'd
using something higher. Search for "555 PWM" to find adequate PWM circuits, all you need is a single 555, a potentiometer control, and a few
miscellaneous components.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Under motion, you'll want an even faster switching rate, like 1kHz. 100kHz is simply out of the question for TIP3055, but it's such a terrible
transistor I'd go for something else on principle alone. RadioShank carries IRF540s, which will do a fine job, and can be driven direct from a 555
(use a series 47 ohm resistor though).
Once upon a time, I made this circuit:
http://www.youtube.com/watch?v=ztBzA7_hjHk
http://webpages.charter.net/dawill/tmoranwms/Elec_LEDs.html
I prefer comparators to 555s because they're more versatile and, given the limitations of the 555's capability for modulation and its output, I'd end
up with all the extra transistors either way.
You could take the same circuit, from the low voltage FWB (all the control circuitry should run from 12V just fine), and instead of the optocoupler
LED and resistor, use a gate drive circuit more like the junk on this circuit:
http://webpages.charter.net/dawill/tmoranwms/Circuits_2010/1...
See the TL494 chip with its internal transistors, imagine that's the last transistor in the modulator circuit. The 1k to ground, 2N4403 (or 2N2907 or
2N3906, anything will do) and 1N914 are additional. Then the MOSFET, source to GND, drain to LED bank, gate as shown.
Tim
|
|
Intergalactic_Captain
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 228
Registered: 4-9-2004
Location: somewhere where i don\'t know where i am
Member Is Offline
Mood: frabjous
|
|
Thanks AX and Watson - But I think I may have misrepresented flicker here. I want to run this at a relatively low frequency - If you can follow my
logic, as a stationary device it will simply be a giant RGB LED - However, IF it was moving, and keep in mind that it is being built in modules, it
will still be a RGB LED - with the upside that you'll be able to see the driving frequency to varying degrees depending on relative speed... I am NOT
trying to build a POV device here...
So, keeping that in mind, I want to do this with 556's - Mostly because they're in my random parts boxes...
The idea is that I set up one half of a 556 to output (for example) a 100Hz pulse in astable mode... This is used as a signal generator for its
remaining half and the other 556 - Each channel of which is set up via potentiometers to provide a variable duration pulse in monostable mode... Each
of these drives one transistor array which in turn drives each color-string of leds...
So... There's the conundrum... There's a shitload of 555/556 circuits all over the net - But most of what I'm searching for gives me a single
variable schematic - ONE 555... However, when done in this manner, both the frequency AND duty cycle are altered - What I would like to know is if
using one 555 as a signal generator and another as a variable pulse modulator is a viable option for "simple" PWM...
...As far as progress, all 144 led's are now mounted - Next step is finishing the common rails and resistors, then onto everything else... Hands are
too twitchy to keep going so I'm calling it a night... Pics will be available if anyone's interested...
If you see me running, try to keep up.
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
Run one 555 as an oscillator and the other as a one-shot pulse generator. Trigger the pulse with an edge transition from the
oscillator. Don't know if you'll find a single schematic already out there that does this.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
^
http://web.engr.oregonstate.edu/~moon/ece323/LM555.pdf
Plug Fig.4 into Fig.8. Easy with a single 556, obviously the pinout changes. ![:)](./images/smilies/smile.gif)
You only need one astable, so you can use subsequent 555/556s for more PWM. Two 556s gets you three channels, which is awfully convenient!
Tim
|
|
|