I'm not connecting it to the mains so please don't close this thread. :S As for the UV, it's techically in the near UV range which makes it strongly absorbable, but not really dangerous for the skin. It's the short end of the UV range that reaches ionizing radition energy levels. Having said that, my GF would only be using if for about 15 min. at a time. The design is based on commercially available products; although, that doesn't necessarily mean much these days.

 

Didn't say I was going to, that was for the poster giving bad advice. If you check my albums you will find a lot of examples of large LED arrays, usually for grow lights. For example...

LED Drivers

I have been a fairly prolific draftsman during my stay here.

Wall warts and old computer power supplies for laptops are pretty cheap nowadays, there is no reason to use mains as it is very dangerous.

Bill's Blog

 

@Billy
Thanks for that info. Are you suggesting that I use two transistors on the series chains and a pot instead of a 555 based pwm?

 

If I'm building a PWM based on an online schematic that may have a lower input voltage than that of my project, can I simply choose an appropriate MOSFET and keep the same design, or do other components need to be adjusted to accommodate for the power difference?

 

You should link to the schematic you've found. Not all PWM schematics found on the web are good, and may not suit your situation. At the very least, for a 555-type PWM circuit, you will need a voltage regulator to drop from 19v down to under 15v so that the 555 timer isn't fried.

 

I'll post a couple which I thought seemed like they should work well.

I assume the PWM goes between the power source and the LEDs. If I drop the voltage that the PWM receives, does that not mean that the voltage being supplied to the LEDs also drops?

 

Would something like this work well for my application?

http://www.reuk.co.uk/LED-Dimmer-Circuit.htm

 

Of course I would substitute an appropriate transistor.

 

I'll post a couple which I thought seemed like they should work well.

I assume the PWM goes between the power source and the LEDs. If I drop the voltage that the PWM receives, does that not mean that the voltage being supplied to the LEDs also drops?

Good question. Why couldn't a regulator run the 555 at 12 volts and the 555 output run transistors to common? That would allow longer series LED strings.

 

Here's a rough draft at throwing the PWM circuit together with the LEDs and current limiting resistors. I'm calling it a rough draft, as it has a mix of surface mount and thru-hole components; and a CMOS 555 was used because even though the pinout is the same as the BJT 555, this particular "canned" library part had the pins in more convenient locations for the schematic.

For some absurd reason, the .sch file is too large to upload; apparently Cadsoft Eagle 6.4 is a space hog.

 

Thanks for taking the time to put this schematic together, SgtWookie. I'm starting to see how it's all supposed to fit together. But to come back to my earlier question, if the voltage going into the PWM is 12V, doesn't that meant that the LED strings will be running on 12V too? I'm probably missing something; just trying to figure out what it is.

 

No, not in this case.

Take another look at the schematic I posted. +V is where your +19v gets connected. On the schematic, +V is connected to the input side of the 12v regulator, and also to the high side of every LED strings' current limiting resistor. 12v is used to power the 555 timer, which controls the gate of a MOSFET via a resistor. In the schematic, a IRF540 is shown, which would work OK for your purposes.

 

At first glance Wookie's diagram does look as if the voltage regulator feeds right into the LED array, but no, there are two different conductors involved.

However, is there ever likely to be a problem driving lots of current into the LEDs and then no current, with fast alternation? Maybe some power supplies would tolerate that better than others. A good hefty filter capacitor would help deal with the fluctuations.

 

Point well taken, John P. I actually meant to throw a rather large one (or several smaller ones) in there, but I was in a bit of a hurry.

 

Ok, I see. I'm ok with reading circuit diagrams, but sometimes I have trouble interpreting some parts.

So the V+ connects to the full psu power.
The 12V is the positive end of a 12V input. Can I step down the voltage with a voltage divider, or will I need something more complicated?

I have a few quick questions in regard to reading the schismatic:
What is "x" referring to in the diagram?
What is IC1 for, and how does it fit into the rest of the design.

Also, what kind of capacitor would I add, and in what location, in order to help deal with fluctuations.

Sorry for all the newbie questions; I am trying to learn as much as possible.
I appreciate all the help.

 

Ok, I see. I'm ok with reading circuit diagrams, but sometimes I have trouble interpreting some parts.

So the V+ connects to the full psu power.

Yes, and V+ is connected to all other points in the diagram labeled V+.

The 12V is the positive end of a 12V input.

All points labeled 12V are connected to each other, and it is the output of IC1, a 78L12 12v linear regulator that comes in a TO-92 package. Your current requirements for the 555 timer circuit will be small.

Can I step down the voltage with a voltage divider, or will I need something more complicated?

IC1 is a 78L12 fixed 3-terminal 12v regulator; the terminals are IN, OUT and GND. It needs to have an 0.33uF capacitor on its' input, and an 0.1uF capacitor on its' output, located as close as possible. These are shown on the schematic.

I have a few quick questions in regard to reading the schismatic:
What is "x" referring to in the diagram?

Those are thru-hole connection points where you can solder wires to. The "+V" and "12V" are merely schematic labels, and don't show up on a PCB. You need places to connect to a board.

What is IC1 for, and how does it fit into the rest of the design.

That is an active voltage regulator integrated circuit. It's much better than using something like a resistive divider.

Also, what kind of capacitor would I add, and in what location, in order to help deal with fluctuations.

Capacitors are shown on the schematic, close to where they need to go on a board. Those smaller than 1uF can be metalized poly, ceramic or mono caps. For 1uF or larger use aluminum electrolytic. For across +V and GND, a 25v or higher 1,000uF to 2,200uF should be adequate.

Sorry for all the newbie questions; I am trying to learn as much as possible.
I appreciate all the help.

Nothing to be sorry for; not a single one of us was born knowing this stuff.

 

Ah, IC1 is the regulator. That makes sense.

A few more questions:
1. So the "x" are simply "mounting points"? For example, I would/could take +12Vx from VO on IC1 and connect it to +12V near R1, and also connect +12V from VO on IC1 to +12V from V+ on IC2? The same with +V and +Vx on the high side of the LED line connecting to V+ on the input side of IC1, and the other going to the PSU V+? I hope I got that right?

2. What are M1D, K and K2 for?

And 3. You said I need a 25V cap between V+ and GND; does that not mean that C1 should be to that spec.?

Once again, I really appreciate all the help, SgtWookie, and am having a lot of fun learning this stuff. I have a triple channel 30V, 10A Lab power supply on its way and am making my through "Teach yourself electricity and electronics". As someone who loves physics, it's great to use it in such a practical and hands-on way.

 

Ah, IC1 is the regulator. That makes sense.

A few more questions:
1. So the "x" are simply "mounting points"? For example, I would/could take +12Vx from VO on IC1 and connect it to +12V near R1, and also connect +12V from VO on IC1 to +12V from V+ on IC2? The same with +V and +Vx on the high side of the LED line connecting to V+ on the input side of IC1, and the other going to the PSU V+? I hope I got that right?

Were a PCB generated from the Cadsoft Eagle schematic, the "X"es would be pads that have through-the-board plated holes in the center, so that one could stick wires through them and apply solder, making a good electrical and physical connection.

The basic idea is that 10 copies of a single board could be made; all of the copies would have all 96 LEDs and the 120 Ohm current limiting resistor, however only one of the boards would have the 12v regulator, the 555 timer, the MOSFET and associated components; this single board would then power the LEDs on the other boards.

I think that you really should start off by ordering a relatively simple kit or two and put them together; that will give you more of a feel for things. It's important to have successes early on; starting with a relatively simple kit gives a high probability of success, and one gains experience in a variety of skills required for more complex projects. Having a large project "go south" can really beat the dickens out of one's confidence level.

2. What are M1D, K and K2 for?

M1 is the MOSFET, D is the drain of the MOSFET.
K and K2 are connected to the lowest cathodes of the LEDs. An initial thought was that one could daisy-chain wire from the "master" board to the "slave" boards, however it would be better to have an individual pad for each slave board.

And 3. You said I need a 25V cap between V+ and GND; does that not mean that C1 should be to that spec.?

The caps after the 12v regulator could be rated 16v or higher, the remainder should be rated for at least 25v.

Once again, I really appreciate all the help, SgtWookie, and am having a lot of fun learning this stuff. I have a triple channel 30V, 10A Lab power supply on its way and am making my through "Teach yourself electricity and electronics". As someone who loves physics, it's great to use it in such a practical and hands-on way.

It IS fun!
Your power supply is somewhat overkill for a beginner, but you won't likely outgrow it.

 

Hija,

Just got back to this thread.

I noticed you already bought the 1000 leds like you said in the beginning of this thread.
The leds you bought was 630nm (just red).

Then later you said this:

The research I did set suggests that the effective wavelength is 320-360nm for this application.

Wait a second ... . I'm confused. You bought 630nm (plain red) leds and you are aiming for a 320-360nm (UV) setup?
Or I am missing something, or you'll just have a whole bunch of red light instead of UV?

Fuzz.

 

@Fuzz
Just ignore that. I meant to write 620 - 660nm. It's not a UV machine; I was thinking of something else. It's red light for red light therapy.

@SgtWookie
I get it now. The whole time I was imagining daisy chaining all the LED boards, but what you are saying is to have them all connect at the pads you marked.

I still don't see the purpose of K and K2. The MOSFET drain connects to the load correct? So all my LED boards would connect to X +12V and X M1D right? Or are K and K2 = to M1D and simply extra connection points?

After this project I will definitely choose projects that simpler and progressively increase in difficulty. This project wasn't actually that hard; it only became a little more complicated once I realised that I need a PWM. The rest of the project is fairly straight forward as far as I can tell. Anyway, I need to finish it as it's a gift so whatever I don't know I'm going to have to learn and learn quickly. I've successfully completed 2 prior projects so hopefully those victories keep me going. lol.