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

Does anyone know the replacement parts required for 12V operation from 1S? I see loneoceans tested but I can’t find any info which components to replace and what values to replace them with.

I like to know too but I dont yet use 12V led. Maybe you can ask him by email? loneoceans seem to be busy from blf recently, but he reply my email when I was building gxb172. Are you making your own gxb172 also?

Ok, I’ll just email him later on. I’m not building one yet so no hurry.

I’d be in for 1 of these. Wish I had the expertise and time to do that. Impressive. Not to many boost drivers out there.

Very impressive. I’m in for 2 if they could be sold.

12v operation has been updated in Loneoceans BOM REV.A2.
It should perform very similar to the H2-C driver.
Jensen567 has a nice thread already setup with his test data. Buck and Boost Drivers, Testing, Modding, and Discussion (Pic Heavy)

RevA2 says it was updated January 2019.

There is also a matching board update dated Feb 3, 2019. OSH Park ~

“GXB172 - 17mm Boost LED Driver - Rev A2 - Adjusted silkscreen, soldermask expansion - Adjusted components for better clearance - Added bypass R15 (~560R works fine) if required - Optional: R15, C12, R3, C15 Otherwise identical to GXB172 Rev A.”

Looks like Gao has found some time to put some finishing touches on this driver.

There are parts on the BOM that refer to E-switches. Is E-switch support new?

If anyone does make these for the community I would like one please

That was what I was thinking too. Maybe his working on implementing the e-switch. Good to see him back on this project.
I made the changes on the driver for 12v operation with a XHP35.2 and got 3 amps at the led but I haven’t tested how well it might handle this.
It might not make that current with the voltage sag from a li-ion, I was using a power supply. It will be voltage dependent in a flashlight.

Here is someone offering to build a few: Any intrest in loneoceans's 50w boost 17mm driver?

Eswitch should be easy to implement in code, hes using an ATiny and has pins that are unused which can drive indicator leds or an eswitch. This can easily be implemented in code quick and dirty without any changes to the design however Ideally, design should be edited to include pull up and pull down resistors for switch support. I don’t plan on doing this for my first run however I can mess with the code and add eswitch support, just solder your switch to v+ and a pin on the controller (micro will blow if above whatever the spec sheet says).

Internal pull up in ATtiny is good enough. I’ve never had issues with that. Hardware supports E-switch as parts are listed in BOM. Don’t know about firmware though.

When looking at BOM, it look like loneoceans already have on board resistor and cap for switch debounce in hardware already! But from his webpage it look like eswitch can be support but no firmware written for it yet.

It might be possible to do an e-switch right now by momentarily opening the power to the +vcc pin and making the attiny reboot. Maybe this is what he is doing and the resistor and capacitor he’s added just debounce the switch.

As far as I know the code has not been released. Only the hex has been published and there is no way to modify it.

The good news is that he appears to be back working on this driver. His intention all along was to release the code when he was happy with it.

It’s probably best to sit tight for a couple weeks and give him time to finish what he is working on.

Thanks for the heads up, I saw that post after I put up mine and responded there too :smiley:

I just finished shoving the GXB172 modified for 12v operation in a Convoy S2+ with a XHP35.2 4000k from Arrow.
I got 2600 lumens on start up and drawing 17 amps or so my meter says. I like the beam much better than the xhp50.2 in the convoy.
Has a nice brightly defined hotspot. Much more throw for sure. Shinning both lights together it looks like the xhp35.2 just drowns the 50 out. But I know its only because the hotspot is so much more intense. I like it. :wink:

How accurate do you think your 2600 lumen is? From what I’ve heard people say about the 35.2, it should not be much or any brighter than the older 35. The older 35, even with a very powerful driver, would not get to 2600 lumen OTF. This driver is not super powerful, like 3A+, it says it should be 2.5A. So overall I would not expect 2600 lumen, but maybe the 35.2 has a real improvement in output?

The HD bins go higher than HI so its possible i’d bet

My lumen tube is calibrated to the set of maukka lights. So it should be close. I have looked over the test done on the xhp35HI several times, if you look at TA’s test he hit 2710 lumen from a cold start, that’s the little ramp near the end of the graph pushed at 3 amps. I don’t think anyone has done a death test on the xhp.35.2 which I’m using in a highest bin in the datasheet E2. 590 lumens at 350 ma. TA hit 2710 with a xhp35HI and the HD would be even higher. I’m not absolutely sure of the results I got but I’m pretty confident it’s close in the reading. The xhp50.2 in the same setup (6v) read I think the highest I saw was 3200 lumens at about 15 amps on the input side, still using the stock switch with two blue springs. Should do better with the LO fet switch. The xhp35.2 switch is stock with 20awg bypass.

On the driver 12v configured GXB172, using a .020 feedback resistor. I saw 3.2 amps from a power supply. I was worried the led may experience the blue tint shift at that current but I have not seen any blue yet. I was using a 30Q battery which will sag unlike the power supply.

If that’s an older test from TA then it was probably when his Lumen Tube was reading too high. That’s also the raw led with no reflector or lens.

Are you saying you saw 2600 lumen at “turn on” with the driver modded to output 3.2A and it was powered from your bench psu and not a battery?

All those things do make a big difference. I typically ignore the first second or two and let the output settle for 15 to 30 seconds so it’s more in line with the ANSI-FL1 specs.

I would be curious to see the complete light measured using a battery and after the output starts to settle. (Some lights settle after 10 to 15 seconds, some might need 30 seconds)

It’s awesome the led didn’t turn blue at 3.2A.

BTW, are you saying the psu drew 3.2A on the input side? That sounds way too low. A psu supplying power to the driver might need closer to 13-15A at 4.2v to output 3.2A at 16 or so volts, right? How did you measure the 3.2A? I’m a little confused.