Results 1 to 15 of 15
  1. #1
    mtbr member
    Reputation:
    Join Date
    May 2011
    Posts
    237

    AMC7135 Handling 6.6V?

    I'm looking at the popular KD 8x7135 driver that puts out 3040ma. http://www.kaidomain.com/product/details.S020073

    I'm wondering if it can handle 2 XM-L in series, drawing 6.6V of current. The datasheet for the AMC7135 lists maximum voltage at 6V, so I assume it's not recommended, but it may work reasonably well regardless. Obviously input voltage would have to be greater than output since it's a buck driver.

    My current light is 1 XM-L in series with a set of XP-G in parallel. I am using this: Regulated CV/CC LED Driver Circuit Board for Cree MC-E/SSC P7 Emitters (8.4V Max Input) - Free Shipping - DealExtreme
    It has worked very well with a 6AA power source. I am looking for something that puts out a full actual 2.8A at least though. This driver puts out about 2.3A.

    Any recommendations? Thanks.
    Bamboo is real.
    Boola Boola.

  2. #2
    Sun Worshiper
    Reputation: oreophilus's Avatar
    Join Date
    Mar 2005
    Posts
    7,368
    Quote Originally Posted by pinkston32 View Post
    I'm looking at the popular KD 8x7135 driver that puts out 3040ma. http://www.kaidomain.com/product/details.S020073

    I'm wondering if it can handle 2 XM-L in series, drawing 6.6V of current. The datasheet for the AMC7135 lists maximum voltage at 6V, so I assume it's not recommended, but it may work reasonably well regardless. Obviously input voltage would have to be greater than output since it's a buck driver.

    My current light is 1 XM-L in series with a set of XP-G in parallel. I am using this: Regulated CV/CC LED Driver Circuit Board for Cree MC-E/SSC P7 Emitters (8.4V Max Input) - Free Shipping - DealExtreme
    It has worked very well with a 6AA power source. I am looking for something that puts out a full actual 2.8A at least though. This driver puts out about 2.3A.

    Any recommendations? Thanks.
    the KD driver does not accept more than 4.5v, it will damage if more voltage is applied, immediately

    bring the +ve from the battery to the switch and from the switch to the +led of the first XML led, the -ve of the first xml led goes to the center of the driver +ve to drive the second XML, normal connections from the driver to the second xml led, so +led on the driver to +led on the xml led, and -led on the driver to -led on the xml led, the -ve bat connects to the -ve of the driver

    the driver controls the current for the whole circuit (3A), the first xml drops the voltage so the driver only gets half of the 2 cell Li-Ion pack or 6 cell AA voltage of 7.2 to 8.4v (the driver drops the voltage .7v and burns off the excess voltage into heat).

    when using this configuration, it's advisable to heat sink the driver
    Last edited by oreophilus; 02-03-2013 at 12:25 AM.
    .

  3. #3
    mtbr member
    Reputation:
    Join Date
    Apr 2010
    Posts
    488
    @pinkston32:That KD driver is not buck but it's linear driver based on AMC7135 chips in parallel. Overhead of the voltage turns into heat. Maximum voltage for this chips is 6V (absolute max. 7V), so 6.6V would be pushing the limit. Keep in mind that XM-L demands about 3.3V@3A current (6.6V for 2S setup), but the power supply would be probably 2S LiIon setup which is 8.4V fresh off the charger. Since 7135 is rated up to 6V only, 8.4V would kill the driver. But, you could wire it like oreophilus suggested. If you don't understand the wiring oreophilus suggested, Google "Poormans 7135 driver setup".

  4. #4
    mtbr member
    Reputation: HEY HEY ITS HENDO's Avatar
    Join Date
    Feb 2008
    Posts
    712
    Any recommendations? Thanks.
    yes,.... going from 2.3A to 2.8/3A is just not worth it
    the difference in brightness will be hardly noticeable
    and your run time will obviously be shortened!
    my XM-L bar light has a www.taskled.com multi level driver
    plugged into a watt meter i found a sweet spot at 2A drive level
    brightness Vs runtime Vs heat management
    ...Scun.thorpe, UK

  5. #5
    banned
    Reputation:
    Join Date
    Aug 2012
    Posts
    2,049
    Quote Originally Posted by HEY HEY ITS HENDO View Post
    yes,.... going from 2.3A to 2.8/3A is just not worth it
    the difference in brightness will be hardly noticeable
    and your run time will obviously be shortened!
    my XM-L bar light has a www.taskled.com multi level driver
    plugged into a watt meter i found a sweet spot at 2A drive level
    brightness Vs runtime Vs heat management
    Agreed ... After testing on a bench, and reading other peoples results 2A is the sweet spot.

    FYI:
    Not on their website, but in my store, at least ... Costco 3 pack, 500 lumen, CREE XML flashlights $22, using 3 C-cell arrangement, Hi/Lo/Strobe/SOS, Brand = FEIT Electric ... Verified (by me) driver is hitting the LED with 2A using a 4.5VDC supply ... That's 9w ... You can't buy the LED's for this price.

  6. #6
    Sun Worshiper
    Reputation: oreophilus's Avatar
    Join Date
    Mar 2005
    Posts
    7,368
    Quote Originally Posted by bikeabuser View Post
    Agreed ... After testing on a bench, and reading other peoples results 2A is the sweet spot.

    FYI:
    Not on their website, but in my store, at least ... Costco 3 pack, 500 lumen, CREE XML flashlights $22, using 3 C-cell arrangement, Hi/Lo/Strobe/SOS, Brand = FEIT Electric ... Verified (by me) driver is hitting the LED with 2A using a 4.5VDC supply ... That's 9w ... You can't buy the LED's for this price.
    for XML U2
    there is no sweet spot, you get a maximum of 165.2 lumens per watt at .15A and it declines thereafter

    if you wanted the most efficient XML U2 current, it would be .15A, but that selection doesn't provide that many lumens

    a plot of the lumens per amp gives a steady curve (from Cree data on XML U2)
    Name:  xml-chart.jpg
Views: 442
Size:  15.3 KB


    selecting the current to run leds does not depend on a sweet spot (which doesn't exist), it depends on appropriate heat sinking and available battery capacity

    design the lamp to dissipate the heat for the current used and for the run time required for the battery selected
    .

  7. #7
    banned
    Reputation:
    Join Date
    Aug 2012
    Posts
    2,049
    Quote Originally Posted by oreophilus View Post
    for XML U2
    there is no sweet spot, you get a maximum of 165.2 lumens per watt at .15A and it declines thereafter

    if you wanted the most efficient XML U2 current, it would be .15A, but that selection doesn't provide that many lumens

    a plot of the lumens per amp gives a steady curve (from Cree data on XML U2)

    selecting the current to run leds does not depend on a sweet spot (which doesn't exist), it depends on appropriate heat sinking and available battery capacity

    design the lamp to dissipate the heat for the current used and for the run time required for the battery selected
    The sweet spot I think HEY HEY ITS HENDO and I are referencing, is the balance of lumen output versus runtime, and current flow is a bad indicator of consumption.

    That chart is worthless without a stated voltage !!!

    Convert it to watts, and you'll see that the small addition doesn't do much as it relates to lumen output.

    9w versus 10w ... Check it out

  8. #8
    Sun Worshiper
    Reputation: oreophilus's Avatar
    Join Date
    Mar 2005
    Posts
    7,368
    Quote Originally Posted by bikeabuser View Post
    The sweet spot I think HEY HEY ITS HENDO and I are referencing, is the balance of lumen output versus runtime, and current flow is a bad indicator of consumption.

    That chart is worthless without a stated voltage !!!

    Convert it to watts, and you'll see that the small addition doesn't do much as it relates to lumen output.

    9w versus 10w ... Check it out
    besides lumens to power efficiency there are other considerations like driver selection, reflectors/lens choice, intended use, helmet or handlebar, single or in conjunction with other lights, geographic location, weight, number of leds, cost of components, complexity, build time, modes or on/off, aesthetics

    for 1 and 2 led lights, I build them at 3A, get the most lumens out of them, and the batteries are selected for the run time required

    helmet 3 and up are different to handlebar 3 and up, so get more efficiency consideration
    .

  9. #9
    mtbr member
    Reputation: HEY HEY ITS HENDO's Avatar
    Join Date
    Feb 2008
    Posts
    712
    a plot of the lumens per amp gives a steady curve
    I agree 100%, BUT....
    this steady curve is nothing but numbers, nothing!!
    in the real world going from 750 > 1000 is not a 25% increase in perceived brightness
    ...Scun.thorpe, UK

  10. #10
    Sun Worshiper
    Reputation: oreophilus's Avatar
    Join Date
    Mar 2005
    Posts
    7,368
    Quote Originally Posted by HEY HEY ITS HENDO View Post
    I agree 100%, BUT....
    this steady curve is nothing but numbers, nothing!!
    in the real world going from 750 > 1000 is not a 25% increase in perceived brightness
    When I parallel a sense resistor with resistance wire, solder one side and just touch the other side on/off to apply the resistance or not, I see the difference. I cut the resistance wire until I reach the amps I'm going for, while seeing the increase in illumination.

    when Cree comes out with a new XML bin, everyone is always amazed at the extra 65 lumens (from T2 to U2, 65 lumens jump for each new bin at 3A)

    well, for U2, between 2A and 3A, there's an extra 230 lumens
    Last edited by oreophilus; 02-03-2013 at 03:59 PM.
    .

  11. #11
    banned
    Reputation:
    Join Date
    Aug 2012
    Posts
    2,049
    Quote Originally Posted by oreophilus View Post
    When I parallel a sense resistor with resistance wire, solder one side and just touch the other side on/off to apply the resistance or not, I see the difference. I cut the resistance wire until I reach the amps I'm going for, while seeing the increase in illumination.

    when Cree comes out with a new XML bin, everyone is always amazed at the extra 65 lumens (from T2 to U2, 65 lumens jump for each new bin at 3A)

    well, for U2, between 2A and 3A, there's an extra 230 lumens
    I'm only impressed that the design is marginally more efficient ... But 65 lumens is only a stop the presses moment for a few insiders within the lighting industry.

    I mean, really ... What percentage of increase is it ?
    Almost nothing ... It's measurable on the bench, but in daily use ... It's nothing one can perceive with their eye.

    Also,
    As I said before ... 2A means nothing without knowing what voltage the 2A is driving.

    That Costco flashlight I mentioned is pushing 4.5V @ 2A ... That's 9W !!!
    It barely get's warm, and it runs for about 4.5 hrs. before the low voltage kicks in, and that's using 3 C-cells (Duracell) ... Imagine what it might do with a good rechargeable battery of greater capacity.

    Look at Relative Luminous Flux versus Current chart for the XML, and also the Maximum Current versus Ambient Temperature chart ... IMO the difference between 2A and 3A is trivial, except on a test bench, and cooler running, longer running, plus longer life of all components is a big plus, IMO ... YMMV

  12. #12
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    286
    Quote Originally Posted by bikeabuser View Post
    Agreed ... After testing on a bench, and reading other peoples results 2A is the sweet spot.

    FYI:
    Not on their website, but in my store, at least ... Costco 3 pack, 500 lumen, CREE XML flashlights $22, using 3 C-cell arrangement, Hi/Lo/Strobe/SOS, Brand = FEIT Electric ... Verified (by me) driver is hitting the LED with 2A using a 4.5VDC supply ... That's 9w ... You can't buy the LED's for this price.
    So, did you measure 4.5V at the LED or 4.5V is the battery voltage and you used a meter in series with the tailswitch/battery pack to measure the 2A ?? If the later, then that's 9W (approx) INPUT power and the LED is seeing 2A X its Vf in watts (assuming a linear regulator based driver, which is highly likely).

    cheers,
    george.

  13. #13
    mtbr member
    Reputation:
    Join Date
    Jul 2007
    Posts
    1,119
    while everyone's expert opinion is interesting to read - I think you guys might be getting into arguements that really don't help the OP at all.

    there's some major issues with the OP's original ask - namely the misunderstanding of current vs. voltage.

    pinkston: your 2 LED's in series present approximately 6.6v of what is called Forward voltage (vf) - it is not current. Current is measured in amperes, and LED's are driven by current. Your driver is considered a constant-current device, which is supplying current at whatever rate it is rated at. In this case - it is 3.04A. At 3.04A, a XM-L will present roughly 3.35vf. The 2 of them in series will double that vf to 6.7vf total. The drive current remains the same in a series LED setup.

    Unfortunately, the driver you mention has a maximum input voltage (voltage from your battery) of 4.5V. while the AMC7135 may be rated at a 6V maximum, other components on the driver may not be rated for more than the 4.5V, hence the lower input requirement.

    2 XML in series will not work on a linear driver with a powersource of 4.5V or less. (At least not supplying the advertised current).

    2 XML in parallel though would work, but each emitter would split the current provided, so you'd be driving each emitter at roughly 1.5A for about 800lm output each. (on paper)
    Last edited by adrenalnjunky; 02-04-2013 at 09:02 AM.

  14. #14
    mtbr member
    Reputation:
    Join Date
    May 2011
    Posts
    237
    Thanks for all of your help everyone. I'm pretty sure I researched and learned all of this stuff at one point, otherwise my current light wouldn't be working. It's coming back to me. Interesting reads here.

    Adrenalnjunkie hit the nail on the head. I seemed to have forgotten these are current regulating drivers. The driver I am currently using is can only supply the same or lower voltage, depending on what the Vf of the LEDs is. Meanwhile the 7135 won't work with the higher voltage power supply and therefore can't supply the voltage required to the LEDs. Thanks for that!

    As for my current question and 2.3A vs 2.8A or 3.0A, I'm still not sure. I still think I'd be happy with a higher current driver, especially consider my reasonably short night rides in cooler air. Mounted to copper pipe caps, at 2.3A they only get slightly warm. My battery supply currently is 6xAA (getting about 2-3 hours I think) but I will be switching to a pair of 18650 and checking out the run times.

    I am temporarily running my light as a fish tank light right now connected to a 7.5V AC/DC converter for about 10 hours straight daily without any issue. For my fish tank lights, I would definitely stick to less than 2.5A though.

    My next build is going to a be a 2 XML build. I will probably continue to use my current light as a bar light and use this as a helmet light. I'm getting my XML LEDs and parts from some Chinese sites. The LEDs about $5 each. The Costco lights are a good deal but I don't have a membership since I live pretty far from one.
    Bamboo is real.
    Boola Boola.

  15. #15
    mtbr member
    Reputation:
    Join Date
    Apr 2008
    Posts
    153
    With just two added resistors, you can make an AMC7135 work with 6x AA (or 2x li-ion) to power two LEDs in series (~6.6V output). The trick is to power the Vdd pin with less than the full battery voltage, using a simple voltage divider. I did this in a modded 2D LED light which is now using 6xAA (using two of KD's 3xAA to D holders) to power a ~6.6V 700mA LED (DX's 5W LED). Photo here.

    You may be able to use your existing DX 20330 driver to drive the two XMLs in series. It should be able to power two XMLs in series without any modding. If you want to increase the current, you can mod the sense resistor to let more current flow, by piggybacking a few SMD resistors on top of the existing sense. The driver appears to use the SM5241 chip, which uses a 0.2V sense voltage, so you want to drop the total sense resistance down to 0.2V/3A = 67milliohms.

    There's an interesting discussion on this driver over at CPF here. Links to the datasheet and a schematic, as well as some modding for modes are discussed.

Posting Permissions

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