Page 1 of 2 12 LastLast
Results 1 to 10 of 13

Thread: some thermal conductivity improving experiments (continued)

  1. #1
    Join Date
    Oct 2016
    Posts
    209

    Default some thermal conductivity improving experiments (continued)

    The old thread wasn't nice to Bbe as the issue had nothing to do with his drivers and I couldn't modify the topic title. So continuing my thread from here. http://photonlexicon.com/forums/show...9-issues/page8

    We left off here:
    Trying to drive the multimode red, green and blue diodes without overdriving them (too much) inside chinese brass housings attached to chinese 10 mm optical plate with heatsinks attached below it with fans attached to those heatsinks and the diodes overheat.
    According to several other users and builds they shouldn't be overheating.
    We tried to find the issue via deduction. First we checked the drivers for issues, then the power cables (different AWGs), then the signal source (5V USB, ILDA from RIYA Lite), then compared collimator lenses, then with and without Lasorb connected, then with different 12V power sources (a cheap chinese LED PSU vs lab PSU vs 12V APU battery), then different LPMs and found no difference. Still overheated and produced far less power then they should. I sent the blue to Bbe and he said he got a stable output (no overheating) after polishing the housing and cleaning some junk in it and the diode.

    So if anyone is building their first high power one and is experiencing this issue you might find this topic useful.

    Bbe, I just got your package. It was sitting in the postal office and they were saying I have nothing all this time. Today they gave me 3 packages, one was for me and 2 packages I realized were for completely different people. You get the idea how well they do thir job.

    Anyway, time to compare your blue module with mine with the actual G2 lenses since you've applied better thermal paste and polished the surface and post my results.

    By the way, have you replaced the power cables that goes from the driver to the diode with thicker red/black colored one or was that me?

  2. #2
    Join Date
    Oct 2016
    Posts
    209

    Default

    Finished swapping the collimators with my chinese G2 lenses. The power output is now just slightly less than what DTR gets, like 3.8W vs 4W at the same current. Got a better thermal paste too and polished some edges of the housings with very fine sandpaper. I dunno if the last two were even necessary but I did it anyway, it took only an hour.

    I don't actually measure the temperature inside the housing but from touch they aren't hot anymore, barely warm. Only tested so far for 5 minutes each diode fully on (constant 5V signal). Don't experience power drop.
    I'm driving NUMB07 at 3.2A, P73 at 1.6A and NUGM01 at 2A.

    The chinese G2 seems to work fine with P73, gets close enough to the diode to focus it.

    Should I perform longer tests or different test to make sure the heat transfer is fine before adding the rest of the optics and putting it in my case?

  3. #3
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    1.6A for P73? this is insane! I suppose you tec the diode...at which temperature? and which is the power you get @1.6A from G2? (single diode)
    Thanks!

  4. #4
    Join Date
    Oct 2016
    Posts
    209

    Default

    I will be able to measure the actual temp closer to the diode in the housing later.
    But the housing itself is not even warm to the touch, unlike the green and blue.
    I get about 1.1W output before any mirrors and optics. This is with the cheap $9 "G2" lens from China, the Aixiz G2 lenses still haven't arrived, I assume it will be around 1.2W. My goal is to get 1W from each after beam correction that is why I'm aiming for 1.2W straight from the diode.

    The P73 diodes seem to be very fragile when it comes to ESD and no soft starting. I've killed two from not using a Lasorb and one from not using a driver with soft start feature. The Nichia greens and blues seem careless about ESD but I use Lasorb on them just because they are so much more expensive than red diodes.
    But as far as overdriving and heat goes, P73 performs great. I admit though that I've only tested them for few hours in CW mode total. I don't know how much this shortens their lifespan if at all. If the point of TEC cooling here is to just keep the temperature below the max operating temp (40), like somewhere in the 25 range, I *think* even passive cooling (no fan on) works but I'll need to actually measure it to be certain.
    I've had a chance to test the Oclaro 700mW earlier this year and even though its recommended output is more than P73 I couldn't overdrive it as much.

  5. #5
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    Oclaro 700mw have slightly worse divergence than P73. I'd like to stay with P73 if you need better divergence and mainly if you overdrive.
    If Aixiz G2 is really BB AR coated up to 700nm yes, you could get maybe about +10% gain on 638nm. It seems 'current' chinese G2 is not AR coated above 520nm, this is why you need more current to reach 1W. You could also get better performance with Dave's FL2mm collis which are specific 638nm AR coated.
    I'd like advice you on not deliver CW 5v lab PSU modulation @1.6A to the mits for few hours, just few minutes while setting the current...unless you get the diode really really cooled. But This way you are going to suffer condensation problems, so you should isolate the reds from outside in this scenario.
    So in my tests, I found a good compromise just cooling diodes about 22ºC (no condensation), to get stable power and a good efficiency. However I wouldn't like to go above 1,4A...better 1,2A. That said, if you want to reach its max...1,4A is my advice, and
    anyway you are not going to distinguish drop in laser power from 1,6A to 1,4A with simply your eyes.
    And yes, Nichia green/blue have on board ESD protection, not reds. As you did it is advisable an on-delay start (or soft start) feature on the red driver. If driver lacks this feature you could add a delay-on relay which delivers voltage to the driver after PSU is powered on.
    You could also switch ON-OFF such relay as a simple but effective E-STOP.
    good luck!!
    Jordi
    Last edited by jors; 05-25-2017 at 11:33.

  6. #6
    Join Date
    Aug 2010
    Location
    London or Spain depending on the weather
    Posts
    1,390

    Default


    @ Jors .....Excelent advice as usual.
    Cheers

  7. #7
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    Hey Lari! Thanks! and nice to hear from you
    Hope all is going well...summer is here

  8. #8
    Join Date
    Jun 2007
    Location
    Monroe, Mi USA
    Posts
    818

    Default

    Yes...Thank you Jodi....as usual....you share all your hard earned results....with great detail.....and...Thank you Finale for sharing.
    Anyway...Thanx
    Last edited by CDBEAM; 05-27-2017 at 05:10. Reason: Removed question
    Beam Axiom #1 ~The Quantum well is DEEP ! Photons for ALL !!
    .
    Beam Axiom #2 ~Yes...As a matter of fact...I DO wear tinfoil on my head !!
    .
    Beam Axiom #3 ~Whe'n dout...Po ah Donk awn et !!
    .
    Beam Axiom #4 ~A Chicken in every Pot, and a Laser Lumia in every Livingroom !!
    .
    Beam Axiom #5 ~"Abstract Photonic Expressionism"....is "Abstractonimical" !!
    .
    Beam Axiom #6 ~ "A Posse ad Essea" ~ From being possible to being actual ...is the beam target !

  9. #9
    Join Date
    Oct 2016
    Posts
    209

    Default

    Actually I was quick to judge the thermal conductivity of the other blue I had and my greens. They are running below their max recommended current but still do drop slowly in light output over time. I can file the back of the diodes some more as well as the backplate some more. I've already polished the bottom of the housing. Or is that not enough and I have to take them out carefully and also polish the flanges?

    Weird though that the red doesn't have this problem.

  10. #10
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    Of course you're going to improve thermal transfer if you file surfaces to be completely flat, apply thin layer thermal paste and tight hard. But diode always will drop a bit in power (and increase wavelength) while it gets hot, until it reaches working temp and gets stable. That point will be more or less depending on ambient temperature and your cooling system design. The Only way to not suffer from drop in power is active cooling with TEC and precise controller. That said, if you can remove heat efficiently from diode with just convection, (copper hosts etc), drop in power is not much, I doubt you could notice the drop just with your eyes.

    Quote Originally Posted by Finale View Post
    Actually I was quick to judge the thermal conductivity of the other blue I had and my greens. They are running below their max recommended current but still do drop slowly in light output over time. I can file the back of the diodes some more as well as the backplate some more. I've already polished the bottom of the housing. Or is that not enough and I have to take them out carefully and also polish the flanges?

    Weird though that the red doesn't have this problem.
    Jordi Luque


    AtenLaser.com
    Barcelona

    "Let there be light"

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •