Led Light Panel
- Thread starter CrazyElectroGuy
- Start date
If he is reducing the quantity of LEDs it may be practical to reduce the numbers per chain. The problem is always the same, tolerance of components, such as the power supply. When you get into the tight tolerance situations you have to have a 12V DC power supply that meets or exceeds 12V, else you are running into problems.
You aren't going to find too many pots that handle ½A too well. They exist, ususally as wire wound types, but it is better to use a low power pot controlling high power components, such as a transistor.
The circuit I drew compensates nicely for variance in power supplies, it will need some tweaking. I don't like the high parts count, and we're dealing with high currents relatively speaking. It actually works better with tight headroom situations, the parts run cooler among other things.
The problem with Darlingtons is they will drop 0.6V across the Collector Emitter as a minimum (saturation), so you've lost that voltage. Add up all the voltage drops and your out of room, completely.
If you're going to use transistors anyhow, why not configure them as a constant current source, which they do naturally? You calculated 10.8V drop, which leaves 1.2V to play with, which is not much to work with. Figure the 0.6V the Darlington will eat off the top and it is even less.
15VDC will make a difference, but that is not a popular voltage for power supplies. You can get ATX PSUs off the shelf.
You aren't going to find too many pots that handle ½A too well. They exist, ususally as wire wound types, but it is better to use a low power pot controlling high power components, such as a transistor.
The circuit I drew compensates nicely for variance in power supplies, it will need some tweaking. I don't like the high parts count, and we're dealing with high currents relatively speaking. It actually works better with tight headroom situations, the parts run cooler among other things.
The problem with Darlingtons is they will drop 0.6V across the Collector Emitter as a minimum (saturation), so you've lost that voltage. Add up all the voltage drops and your out of room, completely.
If you're going to use transistors anyhow, why not configure them as a constant current source, which they do naturally? You calculated 10.8V drop, which leaves 1.2V to play with, which is not much to work with. Figure the 0.6V the Darlington will eat off the top and it is even less.
15VDC will make a difference, but that is not a popular voltage for power supplies. You can get ATX PSUs off the shelf.
Last edited:
Actually 1 ma is not that bright. It is visable, but as a display (and not that great).
Here is something I drew out, it combines PWM with current regulators. You can get 3.6Ω ¼W off the shelf, put two in parallel and you have your current limiting resistors.
It goes to 350ma per chain, each resistor (R4-R6) has ¼W disappation, and the resistor/transistor drops around 0.8V for each leg, leaving 11.2V for the LEDs. Not sure what the PWM range is, 20% to 80%? It can be increased by reducing R2, say 10Ω or so. Do NOT use Darlington's for this.
************************************
Thinking about it, this could be a bit underdesigned. If the basic concept looks good I'll rework it, but the 1/3 of an amp through each transistor will get toasty.
This is for both posts above.
Honestly, I would rather not have to get into playing with IC chips really, especially 555 timers, I always end up blowing them up somehow.
Benard, yes I do need to have control of the individual strings, so I can "mix" the colors together.
so benard, you saying that I should use a 5.4Ω 2 watt resistor for green and blue, and a 19Ω 2.5 watt resistor for red?
I will be using 3 red, 3 green, and 3 blue 1-watt leds.
I would like to use the ATX power supply I have in my room right now, its 1200watt, I think I have a link for it posted on the first page somewhere!
I have multiple ways to cool them.
I have old pentium 1-4 heat sink/fan combos, as well as newer processor heat sink/fan combos, some sheets of aluminum in my shop, since it is easy to work with, drilling holes, mounting, shaping, etc. I've also got quite a few 12v computer fans, that range in size from 1" upto 120mm or larger.
I found 2 power supplies thats not from a computer,
1) is 12v 2000ma from an old massage chair, (picture in post after this one)
2) is from a old usb cd-rw drive, I believe it is +12v/3A +5v/3A, thats what it shows on the transformer, though I have a feeling thats the wrong output for the power supply.
I attached two pictures of the power supply, and one of the various cooling fans and computer power supplies that I managed to pickup that was within 3 feet of my desk.
Attachments
post for the other power supply
Attachments
That is likely not a regulated power supply, which means it's output voltage is likely to be around 16VDC with no load. How do I know this? It doesn't say "regulated 12V" anywhere on it. At 2 amps it will be within hailing distance of 12V.
It will work just fine, but you're going to have to have some kind of support circuitry if you try to use that one.
Maybe you can have best of two worlds[ or circuits ]. With a hand full of clip-leads & a junk box I but together a modified Bill's circuit for B or G.The Radio Shack 276-2020 worked better than expected. Max I set for 400 mA.Shorted 15Ω with only 50 mA increase. Substituted 530Ω for 330Ω with no noticeable differance. Worth a try??
Attachments
Obviously 144 leds all in parallel will work only if the power source is close to the Vf of the diodes otherwise it becomes a very nice heater. The idea to change brightness with a pot or any device is kind of up for grabs. These are not lamps but rather diodes following a exponential curve so be aware of that. for that many lamps 48+48+48 i would go directly to the line and install a proper diode and resistor ASUME 12O VAC then if the diodes are 2v Vf then 48x2=96 20ma for the string then 1.2k 2w resistor per string.
I had a similar thought, doing it this way.
R4,5, and 6 could be three 3.6Ω ¼W resistors in parallel, which would handle the .3W worst case, and they are off the shelf.
This circuit isn't nearly as efficient as PWM (generates more heat), but it is more efficient than what I've been seeing so far.
Attachments
first things first ok, let me get this straight, I AM NOT GOING TO BE USING 144 LEDS ANYMORE, I'm using the 1-watt leds from future electronics, I don't have the part numbers, but they are star hex, and their cheaper than allelectronics, and they have the same current output of 350mA, IIRC. I'm not on my laptop right now so I don't have the bookmarks.
when you say "support circuitry" you mean, heavy gauge wire (18-16 gauge), and higher voltage electrical components?
when i get paid, if I still have money left from bills, I'll order the leds, so I can start testing out the different circuits.
when you say "support circuitry" you mean, heavy gauge wire (18-16 gauge), and higher voltage electrical components?
when i get paid, if I still have money left from bills, I'll order the leds, so I can start testing out the different circuits.
"support circuitry " ' looked over all posts and found no mention. As for wire size ,long runs maybe from power supply, # 16 would be my choice, short runs #24 @ only.026Ω/ft.[ it's what I have most of].
Sorry if we ask questions that have already been answered, it happens quite often as by the time an answer is composed a lot of water has gone over the dam.I guess that is why quotes are often used so that answers are attached to the intended question.
Looking forward to specs. on your new LEDs.
Sorry if we ask questions that have already been answered, it happens quite often as by the time an answer is composed a lot of water has gone over the dam.I guess that is why quotes are often used so that answers are attached to the intended question.
Looking forward to specs. on your new LEDs.
this is the post for the "support circuitry!
I understand, it happens since its a large forum, and a pretty long post also.
Here are the links to the leds, the pages has the specs.
RED:
http://www.futureelectronics.com/en...-emitters/colour/Pages/3671054-LXHL-MDAC.aspx
BLUE:
http://www.futureelectronics.com/en...-emitters/colour/Pages/9703604-LXHL-MB1C.aspx
GREEN:
http://www.futureelectronics.com/en...-emitters/colour/Pages/5914064-LXHL-MM1C.aspx
Last edited:
They are still LEDs, and the base problems are still the same. We've adjusted the circuits to compensate for the higher currents, but you will still need resistors or current regulators, and those LEDs are going to be a lot more expensive to burn out if you get it wrong.
I'm not sure what you are refering to, so I apologize if this comes off weird. We respond to posts on a thread by thread basis, I won't remember your issues on another thread. This is your project, I'm just trying to help. You need to combine any problems in one thread IMO, otherwise it will get pretty confusing for everyone.
If the Emitter is providing .5A, and the transistor gain is 50, then you still have 10ma going into the base of each transistor amp.
I'm not sure what you are refering to, so I apologize if this comes off weird. We respond to posts on a thread by thread basis, I won't remember your issues on another thread. This is your project, I'm just trying to help. You need to combine any problems in one thread IMO, otherwise it will get pretty confusing for everyone.
Part of the problem is the gain of the transistors. If there are many legs it will need to be reworked. We'll have to find out how many LEDs the OP is planning on using total, and if there are more than one per leg.
If the Emitter is providing .5A, and the transistor gain is 50, then you still have 10ma going into the base of each transistor amp.
This is my only thread that I have on the forum, and the only one I have posted in.
if your talking about the "support circuitry" quote, its one the last post on page 3.
---
I know that they are still leds, I was just letting people know that I am NOT going to be using the 144 leds that I had first intended to use at first.
As for the resistors, and the transistors, thats what I'm "trying" to find out what I need, what size resistors, and which transistors I should use.
There are so many post saying use this size, or use that size, I'm getting very confused on what rating resistors I should use.
The rating depends on how hot it will get. Generally you are better off over rating resistors, but plan on dissipating some heat. You need to plan for that now. Part of the planning is how you plan on laying it out.
It can be confusing, but the math is simple. I respect you not wanting to use something like a 555, but that only means more heat, since PWM is out of the picture and it doesn't generally generate heat like analog does (which is the type of control you're talking about).
Until you know how you're going to lay the LEDs out, how many per chain, and the current you're planning on using with them, we can suggest, but we can't offer definitive answers. A third of an amp is a pretty substantial current, which brings us back to the heat problems we've been talking about.
I could be wrong, but you are still going to use multiple LEDs of each color, right?
Reducing the number of LEDs does reduce the other components needed, but you are making up for it by making the remaining components work harder, a lot harder.
Just curious, do you know what surface mount comonents are? The other name for them is SMT (Surface Mount Technology). These parts will not have leads, but are meant to solder direction onto something like a surface of a printed circuit board. Not hard to work around, matter of fact they will fit nicely on the bottom of your heatsinks and fans.
It can be confusing, but the math is simple. I respect you not wanting to use something like a 555, but that only means more heat, since PWM is out of the picture and it doesn't generally generate heat like analog does (which is the type of control you're talking about).
Until you know how you're going to lay the LEDs out, how many per chain, and the current you're planning on using with them, we can suggest, but we can't offer definitive answers. A third of an amp is a pretty substantial current, which brings us back to the heat problems we've been talking about.
I could be wrong, but you are still going to use multiple LEDs of each color, right?
Reducing the number of LEDs does reduce the other components needed, but you are making up for it by making the remaining components work harder, a lot harder.
Just curious, do you know what surface mount comonents are? The other name for them is SMT (Surface Mount Technology). These parts will not have leads, but are meant to solder direction onto something like a surface of a printed circuit board. Not hard to work around, matter of fact they will fit nicely on the bottom of your heatsinks and fans.
I'm working on the layout right now
Thank you. Dont get me wrong, ive had. Electronics training about 4 or 5 years ago. I just dont use it often, so im pretty rusty.
I know that im making it difficult for everyone and im sorry for that.
For the heat i know its going to be quite a bit, but i can build that desk to accomadate several 120mm fans for cooling. It make the desk more "techie"!
I will have the leds, laid out like this...
R G B R G B R G B (There will be about 2" of spacing between each, however, After the Blue, there will be about 4-5" of space for the next series of leds.
R 2" G 2" B 5" R 2" G 2" B 5" R 2" G 2" B (If you see what I'm trying to get at here.
EDIT:
I will have 3 leds per series, each series is it's own color, Red on a string, Green on its own string, and blue on its own string.
The leds are 350mA each, so thats 1050mA, or 1.05A per string, for all 3 strings that will be 3150mA or 3.15A.
Specs for the leds are as follows:
- Color: Red
- Radiation Pattern: Batwing
- Flux: 42 lm
- Radiometric Power:
- Forward Current: 350 mA
- Forward Voltage: 2.95 V
- Color: Blue
- Radiation Pattern: Batwing
- Flux: 16 lm
- Radiometric Power:
- Forward Current: 350 mA
- Forward Voltage: 3.42 V
- Color: Green
- Radiation Pattern: Batwing
- Flux: 53 lm
- Radiometric Power:
- Forward Current: 350 mA
- Forward Voltage: 3.42 V
Yes, 3 of each color, Red, green, and blue
I really can't reduce the amount of leds, since I need it to illuminate about 6 feet of desk area, I know that with the spacing that I'm using, I probably won't quite make it, but its not a big deal on illuminating the entire 6 foot area.
I do know what SMT is, I've used them in the training, though very little time using them. Most of the High-Powered leds seems to be smt contacts. I planned on mounting them to the heatsinks and fans, for better efficient cooling, so again heating won't be a big problem.
EDIT:
I've add an attachment that shows the spacing a little better!
Attachments
Last edited:
OK, so far so good. We seem to be on track as far as the controller goes. Unless you have a reason against, I'd still plan on putting the Red LEDs on one leg, Green on another, and Blue on another. If you go with transistor type current regulators (as opposed to simple resistors) you should be able to use that wall wart (12VDC@2A) you were looking at.
I wouldn't worry too much about some of the discussion between Bernard and I about the final layout of the schematics.
One advantage of having 3 LEDs per leg is it takes the heat off the transistors nicely, which is good.
Do you have a source for low ohmage resistors? Locally I can 1-10 ohms in the standard values at ¼W.
I wouldn't worry too much about some of the discussion between Bernard and I about the final layout of the schematics.
One advantage of having 3 LEDs per leg is it takes the heat off the transistors nicely, which is good.
Do you have a source for low ohmage resistors? Locally I can 1-10 ohms in the standard values at ¼W.
