Is there any programmable driver except for NANJG/7135 ? [With ATtiny or etc.]

Tivo’s East-092 Zilog MCU driven driver

Here is his Nanjg style with the Zilog MCU as well

These are actually very intriguing drivers, and well thought out too

The zilog programmer can get expensive though, and coding is different from the Amtel ATtiny MCU’s

Of course there is the multi-mode MELD driver by tterev3

heck check out the OSHPark projects page…plenty of different drivers, for the most part the ATtiny is the defacto MCU because it appears the easiest to deal with

2 x li-ion in series? Zener mod 7135 or mosfet drivers.

Yep, a couple people have done that.

There are a few mosfet drivers. Have you seen the oshpark project thread?

Wow, I never heard about those DIY projects, looks interesting.
Ill check whether there is anything that is good for me, THANKS!

All the components should be hand-soldered? Or is there an option to order those drivers with already soldered components?

Helios-, is there anything special about those drivers with the changed MCU?
Or I can take any driver and change the MCU?
Aren’t the pins of each MCU is different from one to another?

Just a few days ago I came across this 105C clone that does not seem to use a Attiny MCU (not that I know anything much about MCU's )

What LED configuration do you want to run? I ask because the drivers mentioned will not be suitable for driving a single XM-L with 2x Li-Ion.

You can add a tiny mcu with customs firmware to a 3T6 driver…
This drivers give constant current from 5-13V….
Check out the 3T6 review thread for more information about the driver.

A big point is how much space you have and what current do you need.

yep, a strange MCU :\ never seen it before

I am planning to use it with XM-L2 on copper base, I would like to drive it to 3.5A-4A.
Why the mentioned drivers are not suitable for XM-L?

Ok thanks, Ill check it, I have 20mm Dia. space.
I think this one is bigger, but ill check it deeply.

I think that i’ll stick to ATtiny, because all the other MCU’s programmers are quiet expansive :\

The most interesting drivers Iv’e seen till now, is:

  1. NANJG - mosfet moded
  2. NANJG - zenner mod
  3. Knucklehead V3.0 Buck Led driver by Mattaus - But the only problem I see now, is that all the components should be hand-soldered :frowning:

you can reflow oven (or hot air reflow station) the Knucklehead and pretty much all OSHPark drivers, wight has a buck driver as well (can’t find the link)

because all the 105c based drivers are linear, they dont step the voltage down, just pass it straight threw, to run on 2s cells it would output 6v (8.4v) so it could run 1 MT-G2 or 2s XM-L2’s, on 3s cells it would run 3s XM-L2’s but the only way to run a single 3v emitter (XM-L/XP-G) is to run on 4.2v only (1s)

12F683 is a PIC.

What kind of programmer and how hard to code for the modes? Is this a viable and easy to use equivalent of the AMTEL ATtiny MCU?



domo arigato

Cereal_killer’s explanation is close, but actually only describes the behavior of an FET driver. I’ll explain about each driver in your list:

  1. MOSFET modded Nanjg 105c work like Cereal_killer described. They do not have current or voltage limiting. They will always allow as much voltage (and current) through as they are able to. 2s will destroy your LED. Dimming is done via PWM and does not reduce instantaneous voltage or current.
  2. The 7135 based (commonly called 105c based) drivers will probably output the correct voltage for a single XM-L2 when run on a 2s Li-Ion battery. The linear driver will produce too much heat. The driver will die. A zener mod will protect the 7135's from getting too high a voltage on the VCC pin, but that mod will not protect them from too much wattage being turned into heat.
  3. The Knucklehead may be an option for you, but you should understand that it is still in testing. So far it's been determined that it *must* be heatsinked when run above approximately 2 amps. We do not yet know exactly how much heatsinking is required. Rufusbduck plans to do more testing on this. DBCstm does already have one working at nearly 4 amps; it is driving an MT-G2 which is a significantly higher Vout than your XM-L2. The lower Vout of an XM-L2 will put more strain on the driver. If you build one of these you will be in experimental territory.

WarHawk-AVG mentioned that I have a buck driver, but it is also experimental and not yet publicly available. Therefore it's not what you want right now. Here is the development thread.

There is good news! Since you have 20mm, you are actually not in bad shape. Can you tell us what host you are using?

Sorry for the late reply, I had a load week :frowning:

wight, Thanks for your great reply , Now I understood all the differences between the drivers.

I need something that I will be able to reprogram, because I’ll add it a hall effect switch (electronic switch), and I want it to work with 1-mode only.

I am using some diving flashlight hosts right now, one of them is TR-J2, another is maxtoch 2*26650, aletocn and non-branded Chinese.
All of them for diving, so heat is not my problem :slight_smile: it will be always cooled by the water.

About the LD-29, I have some , and it is nice, as it could be modded to 3.6A with a change of the current resistors, BUT, I couldn’t make it to work with 1-mode only. and I cannot reprogram it to work with an electronic switch.
Is that possible to change the MCU on LD-29? or in any other way, make it 1-mode and add an electronic switch?

And about your buck Driver,it looks amazing, Ill keep watching you posts, maybe your driver will be the one suit my purpose :slight_smile:


And another thing,

Is that possible to limit the current on the NANJG 105C- MOSFET MOD with current sense resistor?
Because PWM as I understood, will fry the LED.


Not possible. I tried to explain that in post #14. The “FET” drivers are entirely unregulated. They will always allow the maximum voltage possible.

From my reading buck drivers may potentially have problems when used magnetically controlled lights. Use caution there. TaskLED has recommendations, see ‘Magnetic Field Warning’ on this page -

I do intend to investigate the modding potential for the LD-29. It appears to use a MicroChip PIC controller. I would replace that with what I’m familiar with, an Atmel ATtiny13A. The pins on the PIC need to be scoped and maybe checked with a DMM to find where the (hopefully) PWM comes out and where the voltage divider for LVP is. I have no immediate plan to do this work, but it’s on my to-do list. Figuring that stuff out should make it easy to use a piggyback board or deadbug style wiring and one of the standard e-switch firmwares.

If you truly only wanted 1-mode and were willing to lose LVP (low voltage protection) it might be easy to do with the LD-29. Find the PWM pin and short it to the PIC’s VCC. That should leave the driver on high all the time with no low voltage protection or warning. It would still need a clicky switch though.

Thanks for your answer!
Yep I read taskleds’ warning, ill have to find a way connecting the hall effect and the magnets a little far from the driver :frowning:
Ill see if that could be done.

Yes, changing the driver into attiny would be awsome!

And about the 1-mode mod, how can I find the pins?
Pwm pin should be the one that leads to the mosfets’ gate?
And vcc is the one that goes to the battery + ?

PWM probably goes to the MOSFET’s gate. Possibly through a resistor or other component such as a gate driver it may be easier to follow the trace back from the gate pin towards the MCU.- VCC will not come from BAT+ because BAT+ can be higher than the maximum VCC (when used with 2 cells). VCC will come from a voltage regulator. The voltage regulator is probably one of the 5-pin packages you see on the board. You should own a digital multimeter (DMM), if not I recommend you purchase a cheap one. You can use your DMM to verify VCC.

IIRC the markings on the chip are legible - in that case you can just look at the datasheet to figure out where VCC is. Yeah, Microchip PIC 12F683.

On the PIC Vcc is called Vdd and it’s on Pin 1.

EDIT: I had a “blonde moment”. PWM will not go to the MOSFET’s gate. PWM will go to a “buck controller” chip. You still may be able to find the buck controller by following the gate trace back from the MOSFET.

I started work on the LD-29.