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

Dang it. Too late :frowning: Bought already, but who knows if it’s shipped. Will have to get another pogo adapter because I’m not rewiring things every time.

Actually I had sent Neal (who sent Lumintop) the updated board files to have the swapped miso and mosi lines a long time ago, so we're not sure why Lumintop went with the older board files for fabrication. In any case, I don't have confirmation yet that future batches will swap the MI and MO programming pads, but I have certainly raised it up to Neal again. Apologies for the trouble but I did my best to convey the message to use the latest files for production, and I can confirm that Neal has received updated files.

contactcr is right - the easiest way is to simply get a bunch of female-female 0.1" headers and make your own cable, if you're using the Emisar pogo-pin flashing kit. These headers are cheap and plentiful. You can also use female-male headers if you unplug only one of the sides. This set on Amazon is fairly cheap and has all 3 variants: https://www.amazon.com/Elegoo-EL-CP-004-Multicolored-Breadboard-arduino/dp/B01EV70C78/.

If you're planning to modify the firmware, please get the latest version specifically for the LUME1-FW3X from either TK's repository, or the github page with the latest Anduril firmware.

There are some modifications that were made to Anduril to help take advantage of the additional hardware in the Lume1 driver, so it's a little more advanced than other Anduril-based flashlights. Some of the base files (non hwdef or cfg files) are modified specifically for Lume1. One example is the external temperature sensor handler. The comments in the code should describe these in detail.

For soldering the Aux LED boards, Scallywag has the right idea. Use some flux and just a touch of solder. I highly recommend using tweezers to hold the wires and pre-cut and tin them to length (details in the datasheet PDF here: https://github.com/loneoceans/lume1-fw3x-anduril/). Solid core AWG30 is great, but any other thin wire of your choice will work well too. There are also detailed instructions about prepping the flashlight for driver replacement.

Just a suggestion: this 8-stranded multi-colour wire-wrapping wire works really well, but the insulation is prone to melting at low temperature, so it needs a little finesse when soldering to prevent the tweezers squeezing the (melting) insulation: https://www.amazon.com/URBEST-Solder-Coated-Wire-Wrapping-Celsius/dp/B01LWI20M0/. This wire is also great for PCB rework since it's solid core.

For the FW1A, I actually do not have one myself so I cannot test nor verify that it will work, but there should be no reason why it shouldn't work electrically, as long as it fits mechanically. Just be careful with the turbo-mode; direct drive FET may be too much for a single XPL LED, so you may have to limit it by perhaps using some 'low-drain' cells like the Panasonic 18650B, and make sure that the heat-sinking of the LED is done really well. However, there should be no problems with FET drive if a 3V XHP50 is used.

I ended up getting this:

30 awg ptfe stranded

It should arrive in the next few days. If it’s any good I could probably sell a few feet of each color if someone needs.

For 30-gauge, I’d recently ordered this after dealing with overly fat switch wires in another light recently: multiple colors, 60-ish feet each, 30 gauge silicone, $12.95 prime as of posting

The dupont wires I ordered were this item for my flashing kit recently (with SOIC8 clip, I don’t have a progkey yet), which looks to be an identical item to the one LoneOceans posted and $0.50 cheaper at the time of this post. Either will work if prices fluctuate.

I prefer to wait for the final version, but can you confirm that the discount code will work for future batches?

Unfortunately I have no idea how long the discount code will work for, nor have any control or knowledge of how many batches there will be (or if there will be future batches for that matter); this would be a question for Neal and Lumintop. They have a business to run so I assume that if it's selling well they will make more of what will sell; likewise with what they want to sell and which flashlights they want to put it in. I've just been able to convey my preferences to them but I don't know how much influence I have.

I'll be happy to help answer technical questions instead, thanks!

It seems that some folks are interested to change the firmware with their own modifications - what sort of changes are people thinking of making? Just curious.

A couple of firmware changes I might make:

Disable the FET for FW1A
Switch to single pin aux led control to use Lexel’s boards

I ordered a few with the mixed up programming pins

Just a note, unfortunately the Lume1 boards are unlikely to be compatible with Lexel's boards. I don't know what the exact schematic of his aux boards are, but the control is likely based on pulling one Attiny pin high or low to turn the LEDs on or off (or through pull up resistors in MCU).

However, the VCC voltage I'm running the Attiny at is different from the battery voltage, so the system may not work as designed. Likewise, the MCU used is different (1634 vs 85), as well as having different pins. Instead, the lume1 tri-LED AUX boards will allow you to recreate all functionality and colour combinations provided by Lexel's boards, as well as additional features provided by Anduril since Lume1 has individual RGB LED control exactly the same as the Emisar D4V2. This includes voltage display, a variety of flashing modes, colour combinations, and bright/dim settings.

The RGB Aux boards are included with the Lume1 driver for now as well, but tri-LED boards will require DIY assembly.

For FET disabling, yes the easiest way to do it is with the config files, just a one line change. The Lume1 also has 4 additional solder jumper pads that need some firmware to be taken advantage of. Perhaps toggling FET on/off could be one of them..

Anyway, I'd love to garner more feedback from the Lume1-Fw3x driver; perhaps I could roll them into improvements for future drivers. One flashlight I'm eyeing is the Noctigon KR4 and similar series but I haven't had a chance to purchase one yet! It would be nice to also including USB charging I think.

Yeah the brightness will be wrong with the Lexel boards, but I’m hoping that is just a resistor change. The ones I have light up (very dimly) at 2.5V. They look better IMO than the tri led boards.

I ordered few of them. Hopefully, they will be delivered smoothly without problems.

Speaking of the problems, I am OK with the programming pad for MISO and MOSI. It is something I can deal with. If I understand correctly, loneoceans has not tried out the version for the actual production yet. Also, hopefully, the driver will work fine without other problems.

I would like to thank everybody who has made this driver possible.

Is the FET in the Lume1 lower resistance than the factory driver? A stock FW1A uses an XP-L HI or SST-20 with a FET.

It should be higher resistance, I think he is just being safe.

Awesome! I guess this means I need to get off my butt and get the code merged.

I’ve mostly been trying to finish up Anduril 2 lately, but after that I’m planning to merge the Lume1 branch and the AVR 1-series branch.

It’s unfortunate about the swapped flashing pads. I should probably get myself another flashing adapter so I won’t have to constantly rewire things. Perhaps it’s time to put in an order with Hank…

My mystery ebay 30 AWG stranded PTFE arrived. OD is ~0.77mm, not the skinniest ever but smaller than silicone and assuming it’s legit should withstand high temp. Will investigate more as my drivers get closer.

In the detailed information, the following is noted:

“Max Turbo is done with a DD-FET, but the current path does go through the sense resistor, adding 20mR in the path. This reduces slightly the absolute max brightness. It’s possible to ‘fix’ this with an additional FET to bypass the resistor, but I’m not sure if it’s really worth the effort. This however can be beneficial for some kinds of lower V_fwd LEDs (marginally).”

So in a sense, that makes this driver even better for 219B’s, and I would think it should be fine for a single XP-L.

A related note, the FET is not PWM’d. If I understood right, the other components interact with it, and the resulting rise time of the FET is somewhere in the ballpark of a full millisecond. So the firmware works normally up to max regulated current, and then the FET 100% on / full turbo only.

I’m a little curious what the effect of trying to PWM it at high frequency would be - might it smooth the output at an intermediate value with ripple at the PWM frequency, or would it risk burning something out?

i have some 30 AWG solid wire-wrap silver-plated wire with Kynar insulation, measures 0.5mm OD, wire diam is 0.25mm. 50 in a pack, 5” long with 1” on each end already stripped off. i doubt this wire is being used to carry any significant current, it could be used as a fuse in a high current circuit.

I’m sure it’s a non issue but I’m pretty sure mcpcb can be higher than 100c but the wire (insulation) will likely never have good enough contact to cause issues. I only had down to 28awg silicone so just decided to get PTFE

Could this be used with very low vF LEDs, like red for example?

My drivers are in the country with an eta of Monday Aussie time. :smiley:
Now to choose the lights and emitters to go with them!

Mine were clearing on Wed, apparently - good old Australia Post being ded. Still in customs last I looked.