Results 1 to 24 of 24
  1. #1
    more carbon=more awesome
    Reputation: The Understater's Avatar
    Join Date
    May 2007
    Posts
    1,315

    Question about b3Flex.

    Hi all.

    Trail testing my new light last night, I found that when I switched it up to the full 3 amps, it would occasionally blink.... as in the light would flash off for a split second before coming back on. Now the driver is brand new, and the only settings I have played with are the brightness settings. Everything else should be set to the default modes.
    I've had a read up, and the only thing I can think of is that it could be the medium voltage warning tripping, but as I understand it, this is set to off as a default. I'm using a 4s1p pack of Panasonic 3400mah NCR18650B cells.

    Could the two MTG-2s on full three amps draw enough power to pull the voltage down to a point where it gives me that warning?

    Other than that one slight niggle, the light was super awesome. (and bare in mind that I don't use many superlatives...)
    Posting on the basis that ignorance shared is ignorance doubled.

  2. #2
    mtbr member
    Reputation:
    Join Date
    Sep 2010
    Posts
    150
    Your light looks super nice - if it lights up the trails equally well, then it's super awesome (a word I never use).

    I think you are right about the medium voltage warning. You can programm the voltage setting (8 clicks), default is 14.3V. The vf of the mt-g2 at 3A is 6.2V, for two in series 12.4V, minimum input for regulation with the driver (+1.1V) is 13.5V.
    These tests show that it doesn't take too long (20 minutes?), until the cells drop bellow (14.3V/4) 3.6V, when discharged with 3A.

    To get rid of the medium voltage warning, just set it to 3V and you can still use the low warning if needed.

    I'm still learning from you here - so please correct me, if I'm wrong!

    PS: World's first MT-G2 bicycle flashlight - not really!
    Last edited by mrradlos; 05-16-2013 at 06:29 AM.

  3. #3
    mtbr member
    Reputation: mattthemuppet's Avatar
    Join Date
    Jul 2004
    Posts
    2,287
    I agree with mrradlos, looks like a voltage sag problem tripping the mid-voltage warning. As the driver doesn't remember that it's tripped already, when you go down a power level, battery voltage bounces back up above the warning and then trips it again when you go back to high.

    You're also pushing the limits of a buck driver with that set up - ideally you want no. of cells to = no. LEDs + 1, but that's with 3V LEDs. Given that you're running 2 6V LEDs in series, you'd need a 5 series (or 5 cell) pack to keep the driver in regulation for the whole battery discharge curve. As it is, your light will start dimming once you get below LED Vf + 1.1V, i.e. about 1/3 to 1/2 way discharged, not counting voltage sag.

  4. #4
    more carbon=more awesome
    Reputation: The Understater's Avatar
    Join Date
    May 2007
    Posts
    1,315
    Would running two packs in parallel help at all?

    b.
    Posting on the basis that ignorance shared is ignorance doubled.

  5. #5
    mtbr member
    Reputation:
    Join Date
    Sep 2010
    Posts
    150
    Quote Originally Posted by mattthemuppet View Post
    As it is, your light will start dimming once you get below LED Vf + 1.1V, i.e. about 1/3 to 1/2 way discharged, not counting voltage sag.
    The Vf of the mt-g2 at 3A is 6.2V, for two in series 12.4V, minimum input for regulation with the driver (+1.1V) is 13.5V. 13.5V diveded by the for cells is just under 3.4V. This test show that the cells are drained 1/2 ore more reaching this point (remember your med-warning was allready at 14.3V). The light will afterwarts start to get dimmer on the highest level. For me, that is the way I would like it to be! If you don't like that, just programm a lower level at 1A and you can drain the pack with a constant level of about 1'500 lumens (still bright).

    A fifth cell in the pack would change that as mattthemuppet explained. Two packs in parallel would just double the time until the light starts get dimmer.

    Why don't you set the med warning to 13.5V and test if you can notice the dimming after the warning, perhaps the low warning set to 12.8V - I bet there is still plenty of light.

  6. #6
    mtbr member
    Reputation: mattthemuppet's Avatar
    Join Date
    Jul 2004
    Posts
    2,287
    it would halve the current draw per cell, so they would sag less under load, but it still wouldn't allow the light to stay in regulation throughout the whole battery discharge because of the Vf of the LEDs and overhead of the driver. As mrradlos pointed out, it's not necessarily a bad thing - some people like it as a battery discharge warning.

    Given that adding a 5th cell would mean tearing down your pack and buying a new PCB, I'd just leave it as it is. Just bear it in mind for the next build

  7. #7
    mtbr member
    Reputation:
    Join Date
    Feb 2009
    Posts
    300
    Quote Originally Posted by The Understater View Post
    Hi all.

    Trail testing my new light last night, I found that when I switched it up to the full 3 amps, it would occasionally blink.... as in the light would flash off for a split second before coming back on. Now the driver is brand new, and the only settings I have played with are the brightness settings. Everything else should be set to the default modes.
    I've had a read up, and the only thing I can think of is that it could be the medium voltage warning tripping, but as I understand it, this is set to off as a default. I'm using a 4s1p pack of Panasonic 3400mah NCR18650B cells.
    Hey did you solve the problem?

    I got a similar problem with my 3xXML build!
    I use a 4s 5000mAh polymer pack where I monitored the voltage under load - and it dropped nowhere close to the warning points!
    But every time I switched to high (3A) it started to blink!
    According this behaviour I also had a conversation with George from Taskled - but he couldn't help me either! (he said that probably some solder flux caused the problem due to some mA running to resistors or so....)

    The vf of 3 XML's in series at 3A shouldn't be the problem, or I'm missing something?
    ..just ride...

  8. #8
    more carbon=more awesome
    Reputation: The Understater's Avatar
    Join Date
    May 2007
    Posts
    1,315
    I got rid of the annoying flashing by lowering the warning thresholds but the last ride out I definitely noticed it fading slowly over the course of the ride. Particularly as I loaned the battery for my bar light to a mate and ran my helmet light at levels four and five. After a discussion with the guy from Night Lightning here in NZ I'm going to run a second pack in parallel. He reckoned I could run them both with one protection circuit but I think I'll use two circuits and one plug.
    Posting on the basis that ignorance shared is ignorance doubled.

  9. #9
    mtbr member
    Reputation:
    Join Date
    May 2005
    Posts
    2,235
    Quote Originally Posted by Whitedog1 View Post
    I use a 4s 5000mAh polymer pack where I monitored the voltage under load - and it dropped nowhere close to the warning points!
    Were you measuring the voltage at the battery or at the driver?

  10. #10
    mtbr member
    Reputation:
    Join Date
    Feb 2009
    Posts
    300
    I used a Y-cable: one end to the driver and on the other end I measured the battery!
    ..just ride...

  11. #11
    mtbr member
    Reputation:
    Join Date
    May 2005
    Posts
    2,235
    Quote Originally Posted by Whitedog1 View Post
    I used a Y-cable: one end to the driver and on the other end I measured the battery!
    Really should measure voltage at the driver. You could be seeing voltage drop through the cable and connectors.

  12. #12
    mtbr member
    Reputation: ktronik's Avatar
    Join Date
    Dec 2006
    Posts
    424
    There is a small issue With the bflex, that when turned on slowly to high and when set at 3000ma mode, it can sometime go into program mode or flash when switched up, for this issue please wire a 25v 100uf cap across the input pins, this will stop the voltage spike re-setting the system.
    K

  13. #13
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    995
    Quote Originally Posted by ktronik View Post
    There is a small issue With the bflex, that when turned on slowly to high and when set at 3000ma mode, it can sometime go into program mode or flash when switched up, for this issue please wire a 25v 100uf cap across the input pins, this will stop the voltage spike re-setting the system.
    K
    I believe I was one of the first to encounter this issue shortly after the b3flex came out and immediately brought it to George's attention. I was experiencing it even with 2000 and 2500ma settings with a 14.8V li-po driving a 3 XMLs in series load. In my case it was resolved with a 47uf capacitor across the input but I suspect it will vary based on the length of the power leads used.

    Karl

  14. #14
    mtbr member
    Reputation: ktronik's Avatar
    Join Date
    Dec 2006
    Posts
    424
    Thx Mate, yes 10-100uf should work depending on input lead length....

  15. #15
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    53
    Is this issue still present in currently shipping b3flex?

  16. #16
    mtbr member
    Reputation:
    Join Date
    Feb 2009
    Posts
    300
    Quote Originally Posted by ktronik View Post
    There is a small issue With the bflex, that when turned on slowly to high and when set at 3000ma mode, it can sometime go into program mode or flash when switched up, for this issue please wire a 25v 100uf cap across the input pins, this will stop the voltage spike re-setting the system.
    K
    Quote Originally Posted by kwarwick View Post
    I believe I was one of the first to encounter this issue shortly after the b3flex came out and immediately brought it to George's attention. I was experiencing it even with 2000 and 2500ma settings with a 14.8V li-po driving a 3 XMLs in series load. In my case it was resolved with a 47uf capacitor across the input but I suspect it will vary based on the length of the power leads used. Karl
    Hell yehaaa, that is definitely the problem I tried to discribe some Posts above!

    Is George resolving this issue? Is a new b3flex Version coming?
    ..just ride...

  17. #17
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    995
    Quote Originally Posted by Whitedog1 View Post
    Hell yehaaa, that is definitely the problem I tried to discribe some Posts above!

    Is George resolving this issue? Is a new b3flex Version coming?
    The problem you described in your previous post doesn't sound like the same issue. The specific issue I encountered *only* happens when first turning the light on, not when changing brightness levels. Apparently the problem is due to a voltage spike causing the B3flex's processor to reset, and when it comes back up it sees the push button being held down and goes into programming mode.

    What you've described really does sound like the voltage at the B3flex's inputs is sagging enough on high to trigger one of the low voltage warnings. I encountered problems like this with 18650 cells, but pretty much eliminated them when I moved to Li-Po batteries. If your wires are very long or thin you could still have enough sag at the inputs to trip a low voltage warning. Are you measuring the voltage right at the B3flex's input or at the battery end?

  18. #18
    mtbr member
    Reputation:
    Join Date
    Feb 2009
    Posts
    300
    Yeaa you may be right!
    My Problem is definitely the blinking when switched to high mode and not going into the programming mode!
    But you also mentioned a flashing when switched up?

    I measured the voltage at the cable next to the battery - not at the Driver!
    But i used 0,75mm˛ wires which are pretty thick for a bike lamp......
    Lamp is currently at a friend for testing, so i can not remeasure it now!
    ..just ride...

  19. #19
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    In terms of resolving 'this problem' - it really is inherent with longer cables and jumping straight to high or a high level. The instant that happens there's quite a large current draw and the rather small capacitor on the driver isn't enough to make up for the drop across the wire length and connectors that are often used. It is a very short brownout that you won't see with a meter - only a scope would catch it.

    The only solution for existing boards is to add an extra capacitor at the driver end of the wiring. Something in the 100uF will pretty well guarantee to resolve the brownout issue.

    In future versions (not this year), I'll look at ramping the current up slower when changing levels or on initial jump from off to a high level to reduce the instantaneous current pulse demand. The problem is obviously worse at higher current settings since the ramp up is to a higher level.

    It is hard to second guess how long some folks' cables are and whether they have connectors AND the series resistance of the cells in use. Obviously a lot of these conditions are out of my hands in terms of the driver itself, since it works just fine if supplied from shorter cables without connectors and with batteries that have low series resistance. Protected cells are typically worse since they introduce circuitry in series with the battery output. Often protected cells will trip on a short high current pulse.

    The 'simple' solution for problem situations is to add that external capacitor and size it as required based on your 'system' (i.e. battery resistance, cable lengths, connectors etc).

    cheers,
    george.

  20. #20
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    995
    Quote Originally Posted by georges80 View Post
    In future versions (not this year), I'll look at ramping the current up slower when changing levels or on initial jump from off to a high level to reduce the instantaneous current pulse demand. The problem is obviously worse at higher current settings since the ramp up is to a higher level.
    This does seem like an elegant solution in that it would not only solve this issue but also provide for a pleasant user experience with a gentle transition between brightness levels. One thing I'd really like to see in future versions would be the ability to do field upgrades to the firmware. Obviously this would require some sort of interface (USB, Bluetooth) to achieve, but it would be nice to be able to upgrade the firmware to address minor bugs or feature enhancement when the changes don't require hardware revisions. This same interface could also be use for programming of the various modes that we now currently do via counting clicks on the momentary switch. If you used something like Bluetooth this could also open up opportunities for remote control of the driver.

  21. #21
    mtbr member
    Reputation: ktronik's Avatar
    Join Date
    Dec 2006
    Posts
    424
    Or even using the stat led as an visual feedback to aid in the programing...

    In all the years I have been using these drivers, longer than most, this is the first issue I have had, they have been rock solid, and very dependable... Keep up the great work George!!

  22. #22
    mtbr member
    Reputation:
    Join Date
    Jan 2010
    Posts
    287
    The killer with BT or USB is the real estate to provide that on the boards. Folk already complain about the size and a lot of folk don't need/require the interface.

    More likely I could provide some pads to enable soldering an optional BT unit on there. USB is out, since it then adds all the problems with providing considerable power to the USB chip and then a connector and that means a hole in the case etc...

    The other issue is that I've pretty well maxed out the current uC flash memory, they are 8k in size and the firmware is up in the 7.5 - 7.8k size...

    I've looked at a new uC series in the same family that comes in 16k and 32k and is physically the same size as currently used on the h6flex/b3flex/hbflex and that would open up a larger range of future code options. The new uC also has built in SPI etc so I could use that as the BT hardware interface.

    It is something that has been percolating in my small brain as the next step in the flex architecture.

    The new uC also has a boot code space, so has the necessary support to allow for field firmware updates. The boot code would interface to the BT interface to do the updates or configuration (versus the menu system).

    Thanks for everyone's continued support of my LED drivers.

    cheers,
    george.

  23. #23
    Light freak
    Reputation: scar's Avatar
    Join Date
    Jun 2005
    Posts
    2,409
    Thanks for all your support George. You rock!



    ****

  24. #24
    mtbr member
    Reputation: znomit's Avatar
    Join Date
    Dec 2007
    Posts
    1,660
    Quote Originally Posted by georges80 View Post
    Thanks for everyone's continued support of my LED drivers.
    We only buy them because they are the best!

    Soft starts and ramped level changes would be cool (love my nitecore flashlight for this).
    DIY LED Bike Lights:
    A few Dynamo builds and some Small battery lights

Similar Threads

  1. Did I cook my b3Flex?
    By roxtar in forum Lights DIY - Do It Yourself
    Replies: 18
    Last Post: 03-06-2013, 01:21 PM
  2. Bottom Braket Question, yet another pf30 vs bb30 question.
    By ntm1973 in forum Drivetrain - shifters, derailleurs, cranks
    Replies: 4
    Last Post: 01-21-2013, 08:30 PM
  3. Replies: 4
    Last Post: 11-06-2012, 03:02 PM
  4. b3flex
    By georges80 in forum Lights DIY - Do It Yourself
    Replies: 107
    Last Post: 05-06-2012, 05:23 PM
  5. What chance a B3flex?
    By JezV in forum Lights DIY - Do It Yourself
    Replies: 6
    Last Post: 02-02-2011, 03:03 PM

Posting Permissions

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