home / lasers

Laser Diode Mounting

I once ended up with a large amount of bare 50W laser diode bars, and a need to mount them, which turned out to not be all that bad. After I did a little research I ended up at Robin's site, die4laser.com, which had some information on mounting bare laser diodes.

Practice Attempts

I decided to start out mounting some 1550nm DFB lasers that I was not very attached to (they were designed for telecom applications, and were not very useful to me), and got a basic procedure down. I started by smashing a TO-220 packaged MOSFET (dead one of course) to get the copper heatsink mounting tab, and dremeled/filed it down to look something like a laser diode mount. Next came a small piece of double sided copper clad, to give me a spot to attac the cathod contact. Then I bolted it down to my DIY Hotplate, and heated the whole mess up to 130ºC, just above the melting point of 93ºC indium based solder I was using. Finally I applied a drop of flux, followed by the laser diode, and another drop of solder on top, and then the cathode wire. After a few tried I finally managed to get a diode which was right side up, pointing the right direction, and reasonably well aligned with the front of the mount:

The Big Stuff

Next up was the diode bars, which are essentially 19 laser diodes that have not been cleaved apart yet. This is both good and bad, on the one hand they are much easier to handle than the individual bars, but on the other hand it means that the solder has to be perfect across the entire bar, if there are any voids the emitter above the void will overheat and short out the entire bar.

First, I put the copper mount on the heatsink, held it down using some socket head cap screws with a piece of 1/4" silicon tubing (to act as a crude spring) and let everything heat up. Then, I wiped the surface with a flux soaked q-tip, quickly dabbed some solder, and wiped it around with the q-tip to tin the surface. The q-tip also served to remove most of the solder, ideally only a very thin layer is required--too thick and the solder oozes out and can short out the bar. Then, the same procedure happens for the top of the bar, and the electrical contact (piece of copper sitting the right of the bar on the hotplate)

Then the moment of truth, I set a ceramic spacer on top of the mount (to keep the copper top contact level), added the copper foil, gently added the copper top contact, and finally gently screwed it down to make sure the bar was not floating on top of a pool of solder. Then, I used a q-tip to wipe any solder which may have oozed out the front (any the emitters are right at the bottom surface of the bar, so any extruding solder at the bar-copper interface would block the output) and make sure the bar is perfectly aligned with the edge of the mount. If everything went well, it looked something like this:

After everything cooled down, cleaned the front facet with isopropyl alcohol, and transfered the bar to a water cooled copper block with a large area thermal power meter directly in front of the bar to measure the output power. Then I tested them by applying 40A, full rated power, for a few minutes. If everything went well, the power would not drop measurably during the run, however if something did go wrong some of the emitters would explode and the power would be greatly reduced:

I found that the step of tightening down the top contact to squeeze out any excess solder greatly reduced the chances that there would be a failure like the one above (presumably because it helped reduced any voids in the solder) and I had about a 90% success rate with the method.

For the final finishing touch, I made simple cases out of ESD foam and some clear boxes I found:

For more details about the diode themselves, see my page on Diode Bars.

More Information

I can be contacted at contact@krazerlasers.com