I have been looking at specialty bike light applications for about a year now. Specifically, a daytime and night time high brightness flash application for a red tail light.
I wanted to apply a constant light along with flash modes. I have not found this in commercially available drivers. This modification allows for a 350ma constant output where various modes can "ride" on top of this constant mode.
The base driver is the NANJG 101-AK driver, in this case, the 17-mode version. The reason for choosing this driver was the special modes including 1Hz, 1/3Hz, 10 second beacon, and a special police mode. Still applicable are the constant modes which add on top of the 350ma minimum.
The trace to the right AMC7135 gate was cut (small circle). The gate was then re-routed to the protection diode output (red rectangle). This provides full drive to the AMC7135 when power is applied.
Variations are easy to envision because these are 1400ma drivers. Here, the driver is configured for 1050ma and the constant modes would be as follows:
low - 350ma +5% of 700ma = ~385ma
med - 350ma +30% of 700ma = ~560ma
High - 350ma +100% of 700ma = ~1050ma
Adding additional AMC7135 will obviously change these numbers but you get the idea.
Since my recent tests with runtime and output at 1 amps I decided to take this mod to the next level.
I received some more drivers and few 16mm XM-L T6 emitters. I decided that my be-seen zoomies on my recumbent trike needed a consistent output and the constant mode that this mod provides. This exercise is a result of some knowledgeable people with regard to bike lighting where it is shown that constant lights are easier to "locate" on the road. But I cannot give up my strobes... so now I can have both modes. See the new Jetbeam ST Cycler specialty modes.
The zoomies currently run at ~1 amp into an XR-E Q5 with a single 18650. I find that this is plenty of light for their purpose and also provides sufficient runtime in strobe mode.
Today I cut the gate trace on one of the 7135 and rewired it to the protection diode to provide the constant 350ma output. I cut the gate trace to a second 7135 to disable it. Now I have a constant 350ma (140 lumen) XM-L with any one of 9 modes on top of it. Max output at this level would be around 280 lumens (700ma) throughout the single 18650 runtime. Originally, the Q5 would output 240 lumens and rapidly drop to 160 lumens and slowly wonder down to 120 lumens before the battery was depleted.
I fully expect to continue getting a 4+ hour runtime. The test is in progress now but I used 350ma x 4 + 50% duty cycle for strobe to equal 2100ma total as a baseline. As long as I use quality cells with a full charge, I should have no problem achieving the desired results. Runtime will improved with any of the other modes like the 1hz or police mode. Again, the real beauty of this is that the XM-L will self regulate all the way down to 3V at the battery at this reduced current. Heat too is proving to be very manageable. Running over 2 hours now and it is easy to hold. On a bicycle, heat will not be an issue.
If you look carefully, you can see the cut trace right next to the "R3" silkscreen and the trace under the added wire.
Last edited by Be-Seen Triker; 09-18-2011 at 02:28 AM.
The driver is a 17-mode NANJG 101-AK 1050ma driver.
With the disabled 7135, it now has a max output of 700ma in high mode.
We know these drivers have a Low mode at 5% and Medium mode at 30%.
So these are the modes after the mod.
Low - 368ma | 160 lumens
Med - 455ma | 182 lumens
High - 700ma | 280 lumens
Fast strobe + 140 lumens
Police strobe + 140 lumens
1hz strobe + 140 lumens
1/3hz strobe + 140 lumens
10 second beacon +140 lumes
SOS + 140 lumens
All strobes trigger the second 7135 for a full 700ma delivered to the emitter.
It is a bit harder to see the group switch trigger in low mode. If you know what to look for, you will notice it but it is only 20 lumens brighter than the constant. I really like this about this driver. You can hide the massive modes if the 5-mode will suffice, or even the solid output 3-mode group.
As a 3-mode driver, you now have some closer range outputs...
Low - 52%
Med - 65%
I find these modes much more usable than the stock 5% and 30% for say a P60 drop-in configuration. And you could use these new figures with any configuration you like, be it 1050ma, 1400ma, or hybrid 1750ma, 2100ma, 2450ma, or 2800ma. Another level would be to hardwire 2x 7135! I'll let you do the math.
FWIW: A 1050ma driver will also self regulate the XM-L on a single 18650 until around the last 10 minutes of battery life where the output drops quickly. More info on this HERE.
And if you missed the other features... these drivers do have mode memory and a low voltage alarm (slow low flash). The alarm will not disable the bypassed 7135! Only the triggered output of the MP.
Last edited by Be-Seen Triker; 09-18-2011 at 03:05 AM.
My math wasn't off too far.
On a 3100mah Callies Kustoms 18650, I got 05:40 to 3V.
At 5 hours it was 3.3V still drawing 700ma on high;
At 05:30 it was 3.17V drawing 700mah.
At 05:40 the driver's protection circuit kicked in and the removed cell was 3V exactly with only a few seconds rest.
350ma x 5.66hr = 1981 x 1.5 (fast strobe 50% duty cycle) = 2971.5ma
Runtime using the 1hz strobe with the previous test cell; Panasonic CGR18650D (2350mah) - 04:30
Last edited by Be-Seen Triker; 09-19-2011 at 12:55 AM.
I have looked for this kind of driver for a long time. (Makes for a nice bike tail light driver) I have 101-AK drivers also, but they don't look the same as yours. They lack the R3 component marked "472" and the tracing is a bit different. I cut the tracing closest to the 7135 like you did (despite the different tracing) and soldered a wire to the protection diode. Resulted in a single mode driver, steady on on/off only :-(
Have you figured out a way to mod this driver to supply a higher strobe rate than 10 hz? I was messing around with this driver and somehow I managed to get it to strobe double rate - but I have no idea how that could be repeated. I would love to have an LED driver similar to this one that strobes at 20 hz-100 hz