Alright, hope everyone had a good break over the holidays!
Still no beamshots I’m afraid, we have half a meter of snow outside so those will have to wait a bit.
-
I’ve got some new super exciting runtime graphs though! Yay!
I had a go at maximizing the heatsink stuff inside the driver cavity and then ran the same setup as the last test to see if I could see any difference in the output behaviour.
All I really did was wedge some bits of thin aluminium sheet in between the driver assembly and the cavity wall, press some TEM cubes into the middle of the 7135 slaves (just to try and homogonize the temperature within all those 7135s) and apply a bit of thermal paste where needed.
I also touched up a solder joint on one chip that looked suspect and could potentially have caused a flicker on one particular emitter. That was something slightly annoying I noticed each time the light was warming up. I probably didn’t solder one 7135 ground tab perfectly with the copper sink rings causing it to heat up faster than all the others.
Not sure yet if that has been fixed but I definitely hope so.
It’s a mess in there and I wasn’t too hopeful that it would actually do all that much but the test results are pretty significant!
( Faded graph is old comparison test done before the heatsink improvements, solid is the new test. )
Everything was set up the exact same way as the first test except that the starting temperature of the light was around 4 degrees C cooler.
So I’m very pleased with these improvements. It’s really obvious that the light is now able to maintain considerably higher output for longer, the steeper part of the output curve now starts at 4.5mins rather than at 2. Temperature on the outside of the light also gets considerably higher and continues on towards 70 degrees, at around the 5.5 minute mark I saw a larger flicker in the output and decided to call it a day at that point! Haha, made me jump a bit tbh!
I suspect that was caused by one entire driver puck getting too hot and calling it quits, didn’t investigate which one because I wasn’t too happy running the light at that kind of temperature.
Next I’ll do a test with the battery pack cable again and see what real world performance looks like now.
Seeing how much better things are running now, I may well go back in and pot the entire driver assembly.
Cheers
Linus