Lume1-FW3X: Constant Current Buck-Boost & FET Driver with Anduril1/2 + RGB Aux

Yep… :+1: I have to agree… Lexel sent me a boat load of different micro resistors for these or his boards, I might mess around with the last driver, lowering the green and maybe the red a tiny bit more? Winter is coming… :smiley:

Hmmm……… I sure would like to get rid of this useless :rage: “MUGGLE MODE” :rage:

Another option with the provided resistors would be to use the 330Ω on the green, the 100Ω on the red and just solder bridge the blue. Would that fry the blue?

Hello everyone,

I'm glad some people have received their boards! Unfortunately I personally haven't yet received mine yet, though they are in the mail and I should receive them sometime today hopefully.. I see that there has been a good amount of feedback so far so here are my comments.


Yes the system should work with red LEDs. However, you should make sure that FET is disabled; this can be easily done by changing the config files (e.g. change the ramping tables) and reflashing the firmware so that the FET is never used. The system will run in buck-mode for the entire range if red LEDs are used.


Yes there is, please see my github for the latest versions. https://github.com/loneoceans/lume1-fw3x-anduril/


Thanks for the detailed feedback gokalpm. I'm glad you were able to get it working well in the end.
- I did not design the PCB with cutout because newer versions of the FW3x flashlights no longer have the notch, so I was told by Neal to make a round PCB. Filing would work to fit the older ones.
- The RGB aux board was designed for the Lumintop MCPCB not the Noctigon one

- The aux LED board should sit flush on top of the MCPCB. The Lume1 datasheet has an assembly section on how I put things together, and things to take note of as well.


That is a very clean build, thanks for sharing!



That's unfortunate to hear.. :( If you scroll back in this thread a while back when I got the first proto samples from Neal (the green PCBs), I found the resistors chosen to be non-ideal and made the Aux LEDs far too bright, and specifically requested them to be changed. I think Lumintop probably wanted really bright AUX lights, but that was not what I think they should be because I prefer the AUX LEDs to be very dim.

Personally, my recommended resistor values are:
1.8k for Red and Blue
5.6k for Green

These values allow for a good change in brightness when using the 'dim' AUX settings in Anduril as well (especially the breathing rainbow mode which I use most often), and they work really well in the nighttime, but will not be very visible in the daytime. These values are listed in the datasheet on Github too.


LED intensity is controlled by the resistors, so swapping wires from driver to aux board will not change the brightness, but will change the ordering instead.


No please do not use a solder bridge! I'll need to ask Neal why Lumintop didn't change the resistors to what I specified.


When I designed this board, it was immediately after I designed the Lume1 board, which I assemble and solder by hand. The actual driver board has smaller components so the LED pads seem easy in comparison, but you are right that normally this would be a challenging job because they're tightly spaced. [Update: it's done! https://oshpark.com/shared_projects/mtHUb4gp]

I quite like the bright aux that can be seen during the day. For night time, he dim setting is dim enough. I wish it was balanced though… I believe even with the swapping around of the resistors, the green is still brighter and the blue dimmer than the red. I think I may try 75Ω for the blue (since no resistor at all is a no go), 160Ω for the red and 330Ω for the green. I have ordered some 0402 resistors to try.

As someone who has never soldered anything before, it sounds like the aux lights are going to be out of the question, which is fine. Is just the driver swap fairly easy? What’s the process? Unsolder the wires from the mcpcb, remove the driver, unsolder the wires from the driver, solder the wires to the new driver, install the driver, solder the wires to the mcpcb?

Yes. As some others mentioned if you have a first batch FW3A with the dog eared driver slant you will need to re-create that “slant” in the new driver with sand paper or a file.

If you have a new batch FW3A with MCPCB screw you just need to check for any long burs coming from where that hole is tapped into driver cavity (honestly I forgot to even check mine).

Soldering to the driver itself is a piece of cake, the pads are large.

Soldering to the MCPCB is harder but as long as you have a clean, blunt tip and either A) some fresh leaded solder or B) flux paste you can probably do that too without much trouble.

I wouldn’t ordinarily re-use wires but in this case you could get away with it. They are good quality. Lumintop thermal paste is questionable but adequate. Those would be two “extra” things I always do because I have all the stuff for it already.

Wow, that is a good looking board :slight_smile: Nice work loneoceans!

I have a lot of soldering experience, and those RGB pads look great. My advice is to tin the pads and wires first while using plenty of flux. The connections should be a piece of cake at that point.

I have just casual experience during past 2 years and did it so it’s doable for sure.

It still requires working with 6 wires, 4 of which are very small, steady hands, good eyes or magnification, etc.

I think you take it for granted some if you are experienced with tools and probably young. Doesn’t mean it shouldn’t be made easier for DIY if its possible (and it was it seems cause the new revision seems better)

It seems that the examples of the values I see here so far for the AUX board resistors are lower than the ones above. I also like the very dim AUX LED’s. I hope the recommended values above will bring very dim like tritium at the low setting and brighter at the high setting as close as to the moon mode.

I was checking around this thread for more about with the AUX board. I just happened to notice something else. The higher value for the sense resistor was used in the prototype. I am guessing the sense resistor is the big one with the “R 020” label. What did the higher value do? Did it simply limit the current even further for the TURBO mode?

Yes. It is time to move on from 219B. Indeed, the prototype has been built with them (and several others). There have been no problems. So, my questions now are more for the curiosity. I will go ahead and use 219B.

Thanks.

@contactcr, Don’t work on them one at a time like through-hole components, work them all at once until the last moment.

1) With a dab of solder flux on your finger, scrub it onto all 4 pads at once.
2) With a small drop of solder on the solder iron tip, brush back-and-forth across all 4 pads at once a few times. Individual older domes will form over each pad. If any of them connect, repeat with less solder on the iron. Wiping perpendicular can fix it too. Eventually they will all have individual domes.
3) Tin each wire by rolling the wire over a solder drop on the iron’s tip.
4) Trim the wires if you stripped and tinned too much.
5) Hold a tinned wire on top of the correct solder dome. Tap with solder iron corner to add heat. The tinned wire should sink into the dome. Repeat for all wires/domes.
6) Done.

I have already done precisely this but I will elaborate since the combined information will be helpful to others.

In addition to the above:

- The right wire length helps tremendously

- When you feed your wires through consider how they will be attached to the boards so you aren’t crossing them over each other above the PCB (it probably wont fit otherwise)

- If your wire will retain any shape at all consider bending it a little into position so you aren’t trying to juggle tweezers, a clean iron tip, steady hand, etc.

  • If you have one use a smaller tip for only the final 4 RBG wires. A big chisel tip or similar will work better for everything else.

:+1: :beer:

Got 1 set in today - dunno when I'll get a chance to install it, hoping soon! I got 2 FW3A's

I did my FW3C today, and did a few improvements, took some images.

For the resistors, I ended up using r=330 b=330 and g=680 or 1k. It definitely tooks much better, especially with 1k.

I only had 0804 at hand, so I had to improvise the fit. The green fits sideways, but the blue needs to stay flat to clear the optics. Note that the image below was before I figured this out.

I also twised the ends of the enameled wires differently this time, instead of marking with sharpie. Worked nicely.

Installed sliced dogfarts as I previously mentioned. Absolutely fabulous leds, massively underrated.

I installed my Lume1 driver earlier today, I love the idea of a buck/boost for the FW3A, but I have encountered a bit of a problem. I am experiencing visible PWM until level ~20, I am not particularly sensitive to it, but on the very low levels it has an almost candle like flicker.

I am assuming that I received a dud, but I am curious if anyone else is experiencing something similar.

Will buy two once the mi/mo issue is fixed. Great work.

I’ve recieved my driver 2 days ago and after some playing with it, I want to write something about it.

First issue I encountered was when I tried to solder rgb wires to driver. I ripped one of the solder pads from the board :frowning:
They are so incredibly delicate, I have some expirience with soldering and never encountered pads so sensitive. Since i never really liked aux led boards, it isn’t that big of a deal for me.

Second issue is extremely strong flickering in lower end of the ramp. It flickers like it’s in candle mode up until I hit ramp level 35-40, which is too bright for a ramp floor. So after puting in this driver, I definetly won’t use my FW3A in “middle of the night toilet run”, unless i want flicker.

Other than that driver works well, especialy the thermal stepdown from top of ramp, which is much better than in factory driver. It doesn’t overshoot or step down too fast like it did with factory driver. There is also obvious improvement in battery life. Overall runtime is about the same, but it is fully stabilized now until deplating battery when cooled. I also feel like pwm of fet channel would be better than just on/off. Fet being just on/off causes turbo step down before flashlight body gets appropriately warm.

Overall it’s good driver but needs improvements, especially in low brightness flicker.
Sorry for all grammar mistakes that I unknowingly made.

Could it be that flickering is related to some bad conntact? Kawiboy would have notice it already...