Drivers: High power LED driver for laser diode?

[mpoulton]

There are several commercial solutions for driving high power LEDs with electrical specs very similar to these blue diodes. Has anyone besides me tried using these LED drivers for lasers? I was running a series pair of two blues on a Luxdrive BuckPuck 3021-D-E-1000 with no problems... until the driver overheated. My fault. Anyways, I'd be interested to know if anyone has really evaluated these types of LED drivers for laser use:
http://ledsupply.com/buckpuck.php

There's even a four-channel DMX512 controlled driver board available:
http://ledsupply.com/quadpuck.php

[Ragnarok]

I've wondered about this too. There are several of those "buck puck" drivers listed on the site, with mA ratings within the blue diode operating parameters. Why would they be overheating if the load is similar, and is there any reason they could not be heatsinked?

I also noticed there is a wiring harness for those pucks that has a pot for output adjustment. Would that help?

R.

[chad]

If you just want to run them in CW then , yes they will work. Be warned that they probably don't care too much about the safety of the diode. And if they are dimming like the dmx ones. They are going to be using a low frequency pwm and that will show up in scanning.

chad

[mpoulton]

I'm not exactly sure why mine overheated, since they are not supposed to require external heatsinking. However, I was running full current with a 12V input and a load voltage of about 9V. This is a fairly stressful operating point for a buck converter since it's on most of the time but not quite 100% of the time, so I'm not too surprised it had a problem. It was also about 95 degrees ambient! When I get my replacements I will heatsink them.

The dimmable Luxdrive modules do not dim with slow PWM. The dimming signal changes the current setpoint of the buck converter itself. They are rated for modulation up to 10kHz, and may be able to do blanking faster than that. The buck converter operates at hundreds of kHz and the setpoint should be variable from cycle to cycle. Someone might want to try these with a projector and see if it will blank fast enough.