Go Down

Topic: Driver board? (Read 878 times) previous topic - next topic


Hey I've been working with power LEDs recently and am becoming increasingly more annoyed with all of the little things you have to deal with. I would love a board that just takes care of everything and has a 3 watt RGB led onboard along with all the other components to correctly drive it. I'm thinking of something with just power in, ground, and 3 signal wires (one for each LED using pwm input from arduino or led driver). I'm not great with hardware so if anyone knows about somethings similar and can point me in the right direction that would be great. Possibly using a little 8 pin PIC for handling the signal well.

I would probably not want to pay more than about 12 dollars per board or it would get pretty expensive quickly. If somebody has schematics for this sort of thing or is willing to draw some up I would love to have them manufactured so that I wouldn't have to deal with hunting down pesky hardware problems all the time.


If you have already a LED driver board...what's the problem?

LEDS must not be fed too little current or they appear too dim...and they must not be fed too much current or they will die! So, an accurate control of the LED drive current must have some sophistication and complexity to manage things and that is not always on the cheap. A 3-W LED can handle small PWM duty cycles at a much higher current than if the current is applied steadily to make light, so current must be limited and adjusted over the PWM duty cycle range. Current can be limited by use of a single resistor, but  it can get quite hot and waste a lot or most of the power used, so this is probably not an optimal, yet it is a cheap and SIMPLE solution. Maxim and Linear Technology and Texas Instruments have chips that are designed exactly for this purpose, but they require just a few components to make them work.

The simplest driver would be to use a Low Voltage threshold power MOSFET, it's gate  connected to the PWM output, the Source pin connected to ground (minus of the power supply), the Drain of the MOSFET connects to the cathode of the LED (and to an inductor and a paralleled diode), positive of the LED to the LED power supply (3-12Volts at 1 to 3-Amp). A small inductor in series with the LED with a diode across serves to time limit the rate of rise of current so that the PWM would have enough some reasonable range of pulse width before the current exponentially reaches the maximum allowable current. This whole driver circuit would have only 4 parts, the toroidial power inductor (>300uH at 4-amps min), a MOSFET rated at >6-Amp 12V D-Source rating that has a <3V gate (a low V threshold gate power MOSFET), a 3-amp diode in parallel with the power inductor with it's cathode towards the positive supply, and the power LED to be lit.  The PWM signal must have a high frequency rep rate, from a period of a few microseconds to a max <.2mSec. Be aware if the PWM duty cycle exceeds a certain limit, the whole thing goes up in smoke and people will run screaming into the night. 


To verbally repeat the circuit wiring: The whole circuit is a series circuit, first the power LED anode to the positive LED power supply, the LED cathode connects to both the cathode of a >3A power diode and also to one end of the toroidial inductor (The inductor is in parallel with the power diode), then the anode of the power diode and the other end of the inductor connect together to the Drain of the power N-Channel MOSFET, it's Source directly to ground, its Gate connected directly to the PWM output of an Arduino pin.


Jun 03, 2011, 05:59 am Last Edit: Jun 03, 2011, 06:02 am by Kctess5 Reason: 1
Thanks for the replies. I have it working but its not at full brightness (mostly because of transistors instead of mosfets), also I have a lot of problems with connections failing and tracking them down is a pain. I don't have the soul to take it apart and replace the transistors but I might at some point. Also apparently my circuit has some other issues which were discussed on one of the LED blinder threads I started. I know that these problems are mostly my fault due to the fact that this was really my first big high power project. I just think it would be awesome to have a little board with like a terminal block on each side (one side power & ground in, other side power & ground out and on an adjacent side the 3 pwm inputs) so that they could be sort of daisy chained and all of the necessary components onboard. I might draw up a circuit diagram based on your descriptions and have you check it over.

Be aware if the PWM duty cycle exceeds a certain limit, the whole thing goes up in smoke and people will run screaming into the night.

What happens if they are at 100% duty cycle (aka fully on)? Would they burn out?

Plus the way I have it set up has for some reason burnt out 2 of my beloved arduinos :( I ordered a new atmega328 from ebay and I'm hoping it will fix the problem so I don't need to buy a whole new board


Did you check here to see if such a thing exists?

If not, what you want does not sound that difficult to wire up on a protoshield board.
Or to make a PCB that you could a bunch of.
Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.


Like I warned you, if the duty cycle exceeds a certain limit, the current will rise to destroy you power LED's. You must slowly ramp up the duty cycle from 0 and find at what point the LED's start to get too hot and adj your software not to exceed this limit.

BTW, the cost per channel of the parts I mentioned would be less than a few dollars, particularly if you have access to an electronic surplus store nearby.  I built something like this using free power n-channel MOSFETS that I extracted from defective PC motherboards. They are there waiting to be salvaged. The toroidial inductors are another matter and cost a few bucks each new, but can be salvaged from used PC power supplies for  free as well. The larger the toroidial core, the more power it can handle. Almost every ATX power supply has about a 3-in Diam. toroidial transformer on the output side and already wound with a bifiliar winding approximating what would be optimal for your LED control circuit.

The closest thing I saw on the web cost about $30 plus tax and shipping and only handled <1 Amp. Not enough current for a high-powered LED.


How many LEDs? It looks like you said just one? If so you might check out the ShiftBrite (http://macetech.com/store/index.php?main_page=product_info&cPath=1&products_id=1). They are easy to use and don't require other circuitry.


That shiftbrite thing is exactly what I'm thinking of except with a 3 watt led. That is pretty cool, I might have to use those at some point

Go Up