[ GXB172 - 50W Single Cell 17mm Boost Driver! ]

Personally, I’m waiting on Lexel to release his boost drivers which use NarsilM.

I suspect that Richard at Mountain Electronics may eventually install RampingIOS on his boost drivers, but it’s a ways off still.

I don’t think Loneoceans is about the money or even care about selling them… he kindly designs and shares his ideas for free with us on here whenever he gets time for this hobby…. Loneoceans never released the GXB172 on oshpark, so if anyone wants to build this driver they’ll have to draw out 1st & from what I remember few members had difficulties routing the PCB on a 2 layer… 4 layers would be a lot more $$$.

in my opinion the GXB172 was a motivator for other members to push ahead with their own 17mm boost drivers… Lexel is getting pretty close to completion I think and RMM driver is already available.

How many Watts is Thrunite’s boost driver? 35-40W I guess? Might just try to buy a used one and rip that out. :smiley:

Edit: Nevermind it’s lower.

It’s designed badly though. The diode is being overdriven. There might be more problems, who knows.
Also the U.I. is bad (EC50 II) and there are software bugs when it gets hot.

I noticed the PCB is up on oshpark as of Sep 13 (also added to original post). Can we expect the BOM and firmware any time soon? I’m a little unclear on the status of the project since the information was supposed ot be released in March and there was still some tinkering happening in May, but it’s been radio silence since then.


I don’t understand one thing…
GXB172/202 is quite a busy design, if I count correctly 36 components take up pretty much whole board space.
Shocki’s boost is similar.
But MTN BST2 has lots of free space and I can see just 23 components on it. Wires obscure the view, so I am likely missing some but nevertheless the difference is large.

What is the reason for this?

Another question.

I’ve just recently found out that there are boost driver designs that do not increase the amperage draw as battery voltage drops. I thought all boost drivers drew higher amps.

I was testing out the Sofirn SP33 with xhp50.2. It seems to draw 5.80A at 4.2 volts all the way down to 3.7v, then you loose Turbo function. Turbo output at 3.75 volt is about 200 lumen less.

This design is new to me. If anyone is familiar with this style of boost driver I would like to find out more.


That would mean that the output is not constant. You can do that, but why would you?

JasonWW indicated that output didn’t change much.

Output power = input voltage * input current * efficiency.
If you fix input current and output power - as voltage drops efficiency must go up. To make that possible at high voltage it must be seriously low.

Is there no error in the measurements?

BTW there is one case where I’d like a boost driver to keep constant input current and vary output power. Haikelite SC26 with a shorty tube. The cell is the limit and to make it really powerful at full battery it would have to be throttled as voltage drops. Actually I have a hunch it might be overtasking the cell as it is.

From what I understand, with them setting a amperage limit of 5.8A it protects the driver circuitry and components from burning up. Maybe this is why they can sell the whole light for only $23. It’s not a particularly powerful driver at a rated 2500 lumen/2200 lumen measured, but it allows for a lot of turbo run time.

Some high-powered boost drivers may draw 10 to 12 or more amps from the battery which really stresses the driver components as well as the pcb traces.

Like I said, I really don’t know much about this style of boost driver.

Constant current output means roughly constant power output (+emitter temperature coefficient involved) and power input (+driver efficiency involved). That which you describe means suckass regulation, in some way of course.

Just gonna throw out a wild guess here… The GXB drivers had a sales pitch for true constant current. I don’t see this same sale pitch in the MTN driver description. I’m guessing that if MTN went through the trouble to make it true constant current, the same sales pitch would be used. I don’t know the components involved in making the true constant current part, I base this guess solely on that constant current is not mentioned at all in the MTN driver description.

Where are we with boost drivers? I’ve been away for a while, thought something would be ready by the time I got back. Have I missed something? Reading the description of the GXB172 it was almost ready. What happened?

Nothing happened, at least publicly. It is almost ready. Same with Shocki’s driver. BST2 is ready, but it supports clicky lights only. It doesn’t allow to drive 12V LEDs as high as we could hope for, but some expect this is only because that configuration has received fewer tests.

Lexel has had to deal with real life stuff, moving to new home, dealing with leaking pipes in walls, hand surgery, modding lights for customers, building drivers for Fireflies company, (maybe a new job, I’m not sure on that one) etc… So his boost drivers got moved down in priority.

On the plus side, he got some help from ToyKeeper to get NarsilM working on them. I hope he will start working on them soon once he gets the other stuff taken care of.

I hope Lexel has boost driver ready soon! :sunglasses: loneoceans, do you have boost driver ready?

This driver is not being sold as a finished driver.

I see MRsDNF.

to Loneoceans, is there a way I can build this myself?

hello bob_Mcbob! I found that loneoceans has actually all the detail on his webpage!!

The PCB, part list and code is all up for download on website with very long detail. Have anyone here on BLF try to make their own driver?

I may give it a go after the holidays. I went through the component list and started making a order of parts at Arrow electronics and Digi-key.
Adding up oshpark, coilcraft and the two component suppliers I’m at around $80.00 enough for two drivers and a few spare components just in case a let the magic smoke out. I’m thinking this might be a challenging little project for this tinkerer. Soldering 0402 stuff is not exactly fun for me but how sweet the reward.
If someone gives this a go, I would be interested in hearing how it went. :wink:

This sounds pretty cool, but I have a question about the firmware. According to the project’s page and some comments in this thread, the firmware is based on BLF-A6, Bistro, Biscotti, Anduril, and/or NarsilM. All of those are distributed under the GPLv3 license, which requires that derivative works also have their code distributed under GPLv3.

But the GXB172 page specifies a Creative Commons BY-NC license and doesn’t seem to include a link to any of the source code. It just has the compiled .hex file.

Could I get a copy of the code to include in the repository?