Results 1 to 19 of 19
  1. #1
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673

    Did I cook my b3Flex?

    Got my dual XM-L done, set the driver to 3A duo mode, and set it to high to see what the run time would be.
    Within about 10 minutes it kicked to low.
    I figured the thermal protection kicked in but it was barely warm.
    Reset the protection to 90*C (highest level) and same thing happened; again, barely warm.
    Turned the protection off, opened the rear housing to better dissipate any heat, and tried again.
    This time, after about 20 minutes it shut down completely. The housing was very hot.
    Even after cooling it will not power back on.
    Am I ****ed?
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  2. #2
    mtbr member
    Reputation:
    Join Date
    Jan 2007
    Posts
    754
    I hate to say, but it sounds likely. But, you should start by making sure power is getting to the driver. Then carefully check the rest of the wiring. Having the thermal protection go off w/o the case being warm may indicate the thermal path to the outer case isn't good enough. You should also check the LEDs themselves. It's certainly possible one or both of them gave up the ghost.

  3. #3
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    Thermal path should be good. The LED and driver support is bolted directly to the case.





    There was just a HUGE difference in outer housing heat between even the highest level of protection and none.
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  4. #4
    mtbr member
    Reputation:
    Join Date
    Jan 2007
    Posts
    754
    Quote Originally Posted by roxtar View Post
    Thermal path should be good. The LED and driver support is bolted directly to the case.
    I'm not so sure, thermal conduction is based on area. And while you have a big contact area between the LEDs and driver to your L-bracket, and from your L-Bracket to the case, all the heat has to pass through the profile of the L-Bracket which actually isn't a very big area. it's not like I've done the math or anything, just something I'm pointing out that may need further investigation.

    Then again, if the light had a lot of air moving over it, everything may be fine. I would never recommend turning the thermal protection all the way off.

  5. #5
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    719
    Quote Originally Posted by MtbMacgyver View Post
    I'm not so sure, thermal conduction is based on area. And while you have a big contact area between the LEDs and driver to your L-bracket, and from your L-Bracket to the case, all the heat has to pass through the profile of the L-Bracket which actually isn't a very big area. it's not like I've done the math or anything, just something I'm pointing out that may need further investigation.

    Then again, if the light had a lot of air moving over it, everything may be fine. I would never recommend turning the thermal protection all the way off.

    I am guessing the problem is a thermal one, I think you just have the low voltage cut out set wrong causing it to cut out. I hope for your sake i am correct because this is a very easy fix, solved by a bit flex programming.

  6. #6
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    13
    It sounds cooked to me. I agree with mtbmacgyver that the thermal path to the housing was not adequate.

  7. #7
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    Yea, ordered a new one from George. Also sent in the toasted one to see if it could be salvaged.
    Oh well, another diploma paid for.
    When I cooked it, the outer housing was extremely hot; way too hot to touch so the thermal path might be OK. Possibly the reason it felt only luke warm on the initial tests was that I didn't notice and get to it till it had already cooled off.

    Anyway, perhaps this housing just isn't enough surface area to dissipate the heat from two XM-Ls driven to 3A.

    Does anyone have any ideas for a housing this size:

    http://www.hammondmfg.com/pdf/1455K1201.pdf

    with better heatsinking capabilities?
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  8. #8
    Rep Power Infinity +2
    Reputation: brad72's Avatar
    Join Date
    Jun 2009
    Posts
    1,566
    The hammond case should be ok when you are moving and the driver may kick to low when stationary. As a rule you will only need the full 3A to the leds when you are going fast so perhaps a remote bar switch would be good so you can drop down power levels when going slowly.

    Regarding the L bracket to mount your leds to as long as you have used some thermal epoxy or some CPU heat sink paste and a couple of screws (which you have) to secure it to the housing everything should be ok. It is also a good idea to lightly lap the mating surfaces to reduce and surface imperfections that may hinder the thermal junction soundness.

    I have found also that even the air temp inside an enclosure is enough to trip the thermal overloads of the driver. Perhaps you could run a bracket from the rear cover to mount and heatsink the driver and orientate the driver so it is near the L bracket holding the LED's. You will still get thermal cutout as the air temp and radiant heat inside the housing will still trip the temp sencing IC.

  9. #9
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    Yeah, overheating was likely in this case. Consider 3A to the LEDs, typical Vf of 3.3V and 2 of them. That's 3A x 3.3V x 2 = 20W just from the LEDs. That's a LOT of heat to move down through the bracket. With relatively thin aluminium in your bracket, you've created a 'bottleneck' for the heat to travel through.

    Wouldn't hurt when you get it going again to put a thermocouple or your pink pinky on the heatsink to see what is actually happening temperature-wise...

    cheers,
    george.

  10. #10
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    Another thing I've thought of is to vent the enclosure. I do limited night riding and only ride in rain if I get caught unexpectedly so the chance of night riding in rain is slim.
    I'm thinking two 1/8" holes on the bottom of the case, near the front and two more on the rear endcap near the top. This should allow cold air in from the bottom front and hot air out the top rear.
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  11. #11
    mtbr member
    Reputation: mattthemuppet's Avatar
    Join Date
    Jul 2004
    Posts
    2,300
    I really wouldn't go for that approach if I were you. It's not just rain you need to be worried about, but condensation, bugs, dirt etc. I'd focus more on getting a housing with more surface area and making a bigger heat path from LEDs to housing.

  12. #12
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    OK, second try.
    Thermal glued some more surface area to the heatsink:





    Also, the plastic trim on the housing, while creating a nice clean look, also served to decouple the front and rear housing caps. getting rid of them changed the spacing for the optics so I had to machine some material from the front cap:



    Got the new b3flex and put it all back together:







    And:



    However, it still cut out very quickly (10 minutes @ inside temp) and the housing was just moderately warm.
    WTF?!
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  13. #13
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    Took the back cover off and the heatsink was also just reasonably warm. No where's near the "hot to the touch" that the 60*C thermal protection setting is supposed to cut out at.
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  14. #14
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    You say "cut out"... I assume you mean it steps down to the L3 level when the temperature monitoring trips ???

    Do note that the temperature the driver will trip at is the temperature of the driver since it is measuring its own temperature. So, that temperature WILL be hotter than the housing since the driver temperature will be the ambient temp (the heatsink it is attached to) + it's own heat losses.

    What thermal epoxy are you using? The driver you sent seemed to have something that was somewhat 'spongy'. Arctic Alumina 2 part epoxy when cured is quite 'brittle'.

    cheers,
    george.

  15. #15
    mtbr member
    Reputation: roxtar's Avatar
    Join Date
    Sep 2009
    Posts
    673
    Yes, I meant steps down.
    Heatsink paste is Fujik BAB.900
    FUJIK Silicone Thermal Glue (50ml Grease-Like) - Free Shipping - DealExtreme

    BTW George, thanks a ton for the quick service getting the new driver out.

    I'll try the whole thing outside in the cold.
    Democracy is two wolves and a sheep voting on what's for dinner.
    Liberty is a well armed sheep, contesting the vote.

  16. #16
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    Never used that product, so I have no idea of its real world performance versus some reviews that just say "it works really well'....

    I prefer to use a known product, like Arctic Alumina 2 part epoxy. With such products you need to apply a very thin even layer and then slide things around to push out any air pockets etc.

    Anyhow, with your current light, you have 20W and not that much surface area on the outside and that is what counts to dissipate heat. I'm sure with a little airflow as you pedal along the whole unit will be a lot cooler.

    I'd still verify that the b3flex is well mounted to the heatsink.

    I use a thermocouple to verify actual temperatures when I'm testing drivers, with their small tip I can touch individual components. I also use an IR gun to make more gross measurements.

    cheers,
    george.

  17. #17
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    719
    Ever since I ordered and mounted two new H6 flex's a couple months back I have no interest in 2 part epoxy. The thermal tape included was so easy clean and fast. I just used an alcohol prep pad to clean surfaces first and stuck it down. I will be using tape to stick down LEDs from here out as well. I would say put that stuff with all the boards. It was so easy.

    Also I never get my moneys worth from arctic 2 part products as there always seems to be left over and it always goes bad before I use it up. So another plus for tape.



    Quote Originally Posted by georges80 View Post
    Never used that product, so I have no idea of its real world performance versus some reviews that just say "it works really well'....

    I prefer to use a known product, like Arctic Alumina 2 part epoxy. With such products you need to apply a very thin even layer and then slide things around to push out any air pockets etc.

    Anyhow, with your current light, you have 20W and not that much surface area on the outside and that is what counts to dissipate heat. I'm sure with a little airflow as you pedal along the whole unit will be a lot cooler.

    I'd still verify that the b3flex is well mounted to the heatsink.

    I use a thermocouple to verify actual temperatures when I'm testing drivers, with their small tip I can touch individual components. I also use an IR gun to make more gross measurements.

    cheers,
    george.

  18. #18
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    The thermal tape I supply is good stuff... BUT, not as good as a well applied thin layer of arctic alumina.

    For the drivers that have components only on the top surface (lflex, h6cc, h6flex, hbflex etc etc), the thermal tape works great since the area for the thermal path is larger AND I have tested and characterized those drivers over full power range and the tape is more than adequate.

    The b3flex and the maxflex6 have very small thermal paths due to their two sided construction. As such the thermal path requires the best possible interface material and for those, the arctic alumina 2 part epoxy when applied CORRECTLY provides the needed interface. It's thermal conductivity is much better than the thermal tape and IS REQUIRED for the b3flex/maxflex6.

    So, yes, I like using the thermal tape as well, but unfortunately it's not appropriate in the case of the b3flex/maxflex6 when those two drivers are run at higher output power and therefore higher thermal losses.

    cheers,
    george.

  19. #19
    mtbr member
    Reputation: adleysh's Avatar
    Join Date
    Jun 2012
    Posts
    68
    [/QUOTE]However, it still cut out very quickly (10 minutes @ inside temp) and the housing was just moderately warm.
    WTF?![/QUOTE]

    Don't worry about your light cutting out after 10 mins in a static test. I have recently built a light were I knew the surface area was not great enough to dissipate the heat created.

    When I first tested it under static conditions it would dim at the 60 degree cut off after only 5-6 mins but once in use on trail even on constant full power it does not trip the thermal threshold.

    I found that the L bracket arrangment in an old build just didn't have the contact area required to dissipate heat enough and had to move on to a C shaped plate which made contact on the roof and floor of the case using artic alumina to glue it in place.

    Good luck.

Posting Permissions

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