Any interest in high-power LED driver shields?

I'm working on a buck converter constant current source capable of 350mA/700mA with a wide range of input voltages and output voltages for flexibility in driving any sort of LED load. I'm developing it in shield form, with three independent drivers capable of PWM dimming. There will be an on-board attiny which will generate the PWM signal, controllable easily by I2C so that the host Arduino barely has to do any work. Stacking these shields allows you to independently dim as many RGB LED's as wanted.

Sparkfun sells a similar 700mA triple driver, but it's very pricey at $50 and doesn't let you dim as easily as mine.

Is there any interest in this shield? So I know if I should make a bunch of boards.

A preview:

I suppose I would say I might be interested in one (as an alternative to my master for Blinkm which the regulator blew up on) but it depends on the price. Any ideas on how much it might be?

Is it 700mA per channel then? (or 350 or what?)

I might suggest you double the spacing on the LED connection pins to allow 0.2" spacing screw terminals (common size).

Not convinced by the spacing of the bucks in relation to the LED connections... closer is normally better :wink:

Mowcius

I'm planning on selling both a 350mA and 700mA version, since those are the two common max currents I've seen. Going from 350mA to 700mA is just a matter of soldering on another current sense resistor in parallel. Easy to customize the output current yourself too, since I've left room to put three current sense resistors in parallel.

I believe that if the LED's are going to be connected through some wires then a few inches of PCB trace won't matter too much. I might move them closer but I still need to layout the on-board MCU.

Might have to get a little creative to fit in six 0.2" spacing terminal blocks. Also I'm looking for a way to be able to stack these boards while being able to screw/unscrew the terminal blocks.

Pricing: probably around $25 + $5 shipping for a fully fitted board.

Might have to get a little creative to fit in six 0.2" spacing terminal blocks.

Six?

Six connections... 3 terminal blocks...

Also I'm looking for a way to be able to stack these boards while being able to screw/unscrew the terminal blocks.

Hmm well you shouldn't really be screwing/unscrewing LEDs when it's powered up so I don't see why you need to be able to do this.

Shields are pretty much only used for prototyping anyway so it's not likely that it'll be so restrictive in a case that you can't take the shields apart...

Mowcius

Just a matter of convenience :slight_smile:

The driver handles hot swapping of loads anyway. It's a nice feature to have.

Oof, slightly tighter fit than I'd thought:

Waiting for these boards to be manufactured. More details on my blog: http://electronoblog.allanw.org/2010/10/triple-high-brightness-led-driver-arduino-shield/

Allan

If you have the space, it would be nice to be able to program the attiny directly from the host Arduino, ala http://arduino.cc/en/Tutorial/ArduinoISP. I'm working on a dual high current motor driver board that I'm planning to set up that way.

That's a pretty good idea if it can be programmed just by stacking them. But then you'd have to put jumpers in so that it's not always connected that way.

Can't an Arduino program the ATTiny by the ISP port anyway though?

Naturally, it's possible to just jumper to the ISP port. You would indeed need programming jumper to make sure the pins weren't connected all the time, but they don't take up much space. If all else fails, you could use solder jumpers on the back of the shield.

I got my version of high-Power LED shield working. It acts like high power blinkM (got an idea from here)

for detail, check http://neuroelec.wordpress.com/

video High power LED arduino shield - YouTube


Nice job. I had to put my design on hold after measuring some huge EMI spikes at switching instants.

Since the Linear driver is fixed frequency, is there a limited voltage input/output range determined by your inductor choice?

Frequency is adjustable with AD5254 (3 channel for current adjustment, 1 for frequency adjustment). At 2.1Mhz, it seem to working OK with 5-24V that I tested. Mostly follow Linear tech reference design which support up to 45V. Love to test EMI but I don't have ocilloscope :wink:

Might have to get a little creative to fit in six 0.2" spacing terminal blocks. Also I'm looking for a way to be able to stack these boards while being able to screw/unscrew the terminal blocks.

have you thought about molex connectors such as:

http://www.mouser.com/ProductDetail/Molex/39500-0007/?qs=sGAEpiMZZMsD

just a thought,

Carl

Hi, I am very interested in acquiring some of these LED shields, can you please tell me how we could proceed. Best regards

Paul

Nice job. I had to put my design on hold after measuring some huge EMI spikes at switching instants.

Since the Linear driver is fixed frequency, is there a limited voltage input/output range determined by your inductor choice?

Turn out you are right. I need to adjust frequency depending on output voltage. glad to connect frequency adjust pin to ad5254.

Hi, I am very interested in acquiring some of these LED shields, can you please tell me how we could proceed. Best regards

I have couple of mistake on my first design. Not a major things though. I manually fixed them. . If you are interest, I will sell couple of extra boards (about $35). Please contact me through my blog.

I would love one if it could power very high power LED's.

Like these: http://www.luminus.com/stuff/contentmgr/files/0/cf0621e078a2f77a303c6e0ddd72bc09/misc/pds_001230_rev_03_cbt_90_w_product_datasheet_illumination.pdf

Yes, they really are this bright. :sunglasses:

For those that TLDR, here's the short version:
2200 Lumens - And these do exist, I've seen these in person.