<< Lume X1: 40W Single-Cell Boost Driver with Anduril2 and UDR >>

OMG Yes!

FW1A X1 withFC40 would be a spot on.

Even better IMO would be the B35AM Nichia in the FW1A and light pebbled, narrow TIR or a 21700 KR1 (would thinner straight battery tube create enough girth for 21mm cell?)

And honestly, I’d rather run this in a triple FW3A in series rather than a parallel with the Lume1.

PSOC 4000 goes down to 1.45x1.56mm BGA. :wink:
I also think that the effort to port FSM would be high compared to the benefits though.

This looks a great driver, as far as it requiring an E-switch, does this mean it could be installed into a convoy host with a tailcap mod in much the same way as the GXB driver does?
The RGB LED support looks pretty cool too!

But this would be still linear driver, where difference between Vf and BattV would be fired on the chip?

No, a buck driver is different than linear. The difference between cell voltage and the LED vF is not just burned off, it’s “bucked down”. This should help.

Boost voltage to a 3S MCPCB for triples could be fun.

I don’t know if this is in response to my earlier comment or not, but I can’t agree more. If there is a release of this X1 for the FW lights I’m in!

Thanks! The Lume X1 is an E-switch topology. I chose that specifically over clicky topologies because the clicky switch adds significant input resistance, and is non-ideal for this high power driver (40W nominal). This is the reason why getting the GXB172 to work well in a clicky design is challenging and requires very careful assembly and was why I designed the FET-based-tailswitch to reduce the switch resistance - the tailcap mod is not the same as an e-switch, if that's what you're asking. In an e-switch, the battery is always hard connected to the driver and the switch is just a signal net connected to the MCU; in a clicky or a clicky with my tailcap FET switch, the power flows through this switch when it is On. The Lume X1 also uses ToyKeeper's Anduril, which is designed for E-switch use only.

D'AVerk, the Lume1 driver in the Fireflies flashlight is a 6A Constant Current Buck + FET + Charging driver, not a linear driver. It has a fair bit higher efficiency than linear drivers. The buck regulator converts the battery voltage to a lower voltage to match the forward voltage of the LED at a desired forward current. In an ideal world with ideal switches and inductors etc, there would be 0% loss (as opposed to linear drivers where an ideal linear regulator would still have the same loss as a simple resistor), but in the real world, the efficiency at most outputs is around 90 to 95% with some losses in DCR, gate drive, and switching losses in the buck regulator elements.

Thanks! It does take a non-trivial amount of time to layout a board for each flashlight so I'm trying to pick something that has potential for lots of people to enjoy. I noticed that Kaidomain sells some individually-wired 3-XP MCPCBs (looks easy to connect in series with two solder jumpers), but the FW3 series is a little challenging for the Lume X1 to be run at its full potential due to its very shallow driver cavity (limiting inductor size), and generally small-sized PCB with E-switch ring (this takes up a lot of layout estate). I'm curious about the Amutorch E3 with its unibody construction, and I ordered one to see how good it is. It could be a good candidate for me to build my 'Zebralight SC700d' compete, though I suppose ideally it'll be nice to work with a manufacturer to develop a flashlight that would fit this driver better. I'm also looking at making a revision to use the AVR 1-series too..

:person_facepalming: forgot about the limited Z height of the FW lights. Bummer. I’ve gotten so spoiled by the tail e-switch layout (YMMV), but it complicates design and either adds girth or necessitates proprietary cells,the latter BLF hates with a fiery passion. The Lume1 will have to be good enough for the FW3a then and we shall find or create a new host moving forward. :slight_smile:

Well I don't think this limitation rules out the FW lights; all I'm saying is that it limits the maximum power due to component choice. There are methods to get around it, such as milling out a slot for the inductor like I did with the KR1, using a board cut-out to sink the inductor into at the cost of layout estate and PCB assembly complexity, or simply just reducing the power output. I actually purchased a FW1A with the idea of trying this out and driving a XHP70, but am currently weighing different options to see what I should try next.

However, I agree with you that while tail e-switch can be nice, generally a side e-switch makes more sense to me. It's simpler, more robust, more compact (especially if integrated on the driver PCB itself), and saves in mechanical and EE BOM cost, and could reduce assembly time. And like you mentioned, it can reduce the girth as well, so a 21700 flashlight can be made a similar size as say a Noctigon KR1. If I'm not mistaken, the 21700 Zebralight Sc700 is in fact, smaller than the 18650 KR1 by a noticeable margin.

I also really like lights with e-switch on board. Utorch S1 Mini, DQG Tiny 4th 18650 are some of my favourtes. Simple, compact and cheap. Though not always applicable, f.e. it wouldn’t work well on a zoomie (unless that light had a really short zoom travel). And sometimes prone to failures (DQG Tiny 18650 has problems with it).

I got to agree. I really like e-switch lights and I see their benefits, but it would be cool to have some state-of-the-art clicky driver options for the most popular hosts. I don’t even mean boost or buck drivers (would be very nice tho)… I mean linear drivers, 17 mm, with stable timings (OTSM) without tantalum caps. I’m sure Mike C already has his own, if not several.


That’s it… I’ve held off on the E3 until today. Just ordered my first Amutorch! It looks like your Lume Aux RGB boards should probably fit. That’ll save me some time, no need to make the aux board from scratch.

One challenge I see with the Amutorch E3 is that the charger bay is very shallow, like the FW lights and the Noctigons... In this particular design, it's very difficult to mill out a slot from the top (since it's unibody), so I'm hoping that I can either do something creative if it doesn't fit:

  • Using a spacer under the screw, which would actually improve DC contact with the body and the screw (since I can crank the screw down tighter without flexing the PCB), though this means the driver isn't well sinked to the body.. also need to check switch placement too
  • Use parallel, flatter, inductors with option of going both sides of the PCB
  • Use a board cutout and manually solder the inductor (so it sits 'inside' the PCB)
  • Mill out a slot from the button-hole side

If this works it could be a first real step towards a 'BLF Zebralight-styled' flashlight with Anduril. Yes the Lume1-FW3A Aux RGB boards should fit. The MCPCBs (single LEDs) need to be in series, so I'm planning to connect the pads using copper tape + soldering to keep them as flat as possible for the aux board to sit on top. However if I do manage to add some LEDs under the switch dome itself (likely need to find a replacement clear rubber switch boot), then I'll likely not install the aux LEDs on the front.

You can also scratch the traces to transform the 1S3P MCPCB to 3S1P.

Yes, though I'd much prefer not to do it because the only thing separating the traces from a short is the thin dielectric between the copper trace and the copper core; a small disaster waiting to happen. I'll also need to add some protective covering for reliability anyway, which would be almost the same resultant thickness, and may not look quite as clean.. It will work but it's definitely not something I'd be comfortable in a product that lots of people could end up using/replicating, especially so in a high power lithium battery system.

Depends how thick the dielectric layer is, on a Noctigon MCPCB I had no problem cuting the traces while keeping enough thickness for isolation

When did noctigon triple mcpcbs stop being configurable?

A long time ago. I tried to buy some from Hank incase he had some laying around in October 2019 with no luck :frowning: