Review: NANJG 105A 2800mA driver

NANJG 105A 2800mA 4-Group 6 mode Driver

Reviewer's Overall Rating: ★★★★


Size: 17mm
Voltage: 2.6-4.5V
Linear Regulator
Current: 2800mA
Modes: 6 in 4 groups
Price Paid: $4.70

Kaidomain 1)

1) This one was ordered early dec.2010 and seems to be a successor to the NANJG 105.


  • Well regulated and efficient driver
  • Has memory
  • Has a 3-level group with no nonsense and good spacing
  • Has a high PWM frequency giving a steady light at low level


  • Change of group only by soldering
  • Memory kicks in 2 secs after switching on (would prefer after switch off)
  • Can only be used with 1 Li-Ion cell

Features / Value: ★★★★

Mode Groups

The groups are as follows:

default: High(100%) - Medium(35%) -Low(4%) - Fast Strobe(14Hz) - SOS(decent speed)

1. High(100%) - Medium2(20%)

2. High(100%) - Medium(35%) -Low(4%)

3. High(100%) - Medium2(20%) - Fast Strobe(14Hz)

The PWM frequency is high at 4.5kHz.

Mode switching

The NANJG 105A has stars on the bottom side to select group by connecting one of the stars to the outer ring. Select group 1, 2 or 3 by connecting star 1, 2 or 3 (counting from the right) to the outer ring or to star 4 which is connected to the ring.

There are no further groups to choose by combining stars as star 1 takes precedence over 2 and 3 and star 2 take precedence over star 3.

Mode Memory

The mode memory snaps in abour 2 seconds after a mode switch (no indication). I normally prefers memory after switch-off so I do'nt have to tap twice to shift level, but this approach normally needs three more components.

Build Quality: ★★★★★

NANJG has a fine build quality and can withstand normal soldering with a fine tipped soldering iron.

This one runs with a MCU PIC12F629 instead of the NANJG 105's ATtiny13. This has hardly any influence on quality but NANJG has managed to make it run like it has PWM although the PIC MCU does not include PWM (or am I wrong here?).

Battery Life: ★☆☆

Linear regulators with 7135 devices normally are well regulated and efficient. At lower battery voltages the drive becomes direct with a very little loss. The high current is hard to the battery and demands a quality type battery (low internal resistance).

Light Output: ★★★★☆

The output current 2.8A is fine to drive XM-L, P7, MC-E LED's. The high current could heat up the light too much but then it is fairly easy to remove one or two of the 7135 on the bottom side (giving 0,35A each) with a heat gun for removing old paint (use a narrow nozzle cone accessory).

The driver has no boost capability so it must be expected that the output falls at lower battery voltages where the drive becomes direct with no regulation.

The driver comes complete with a (rather thin) red and a black lead to connect to the LED.

Summary: ★★★★

The group no. 2 has a very usable spacing of 100% - 35% - 4% and maintain a constant output as long as the battery voltage is higher than the LEDs forward voltage, all the time with good efficiency. Then there are the possibilities to make just a 2 level light or include the normal blinking stuff with a really irritating strobe and a nicely timed SOS.

Incredible review. Thanks for sharing this.

I will purchase it for a future XM-L light

Thanks for posting a review on this new driver and AK-47C that you wrote earlier. It's good to have this information available.

Thanks and glad you like it. Sometimes something good comes out of an unexpected delivery from the vendors.

Thanks for sharing your experience, sixty545!

I was just about to start a thread about this driver when I saw your review because I am thinking about getting one.

May I ask you (and all the other electronics experts here on BLF): On Aqualab's driverlist which I found very helpful and trustworthy in the past the Vin for this driver is listed as 3,3-6,6 V. Now I am wondering: Is it possible to run it safely with 2 x CR123A (primaries)? Is there any alternative that can handle both 1 x Li-Ion and 2 x CR123A that gives >2A and has at leat one low mode and preferably no strobe?

TIA, Huny

Shiningbeam says its drivers can be run on 6V, but not 7.2V. They are NANJG linear regulators too, so I think you can do it, but you will lose a lot of the efficiency because all of the extra voltage will be burned off by the linear regulators. What you are describing is more of a buck driver, but I don't know of any good cheap 3-mode ones. A buck driver would work great on 2 cells and would probably go to direct drive pretty quickly on 1 cell.

The limiting factor is the MCU (Micro Controller Unit) which is specified running from 2V to 5.5V. Add to this about 0.3V for the schottky diode that feeds the MCU (polarity protection). Then you have the range 2.3V to 5.8V for guarantied running. The absolute maximum voltage is specified to 6.5V, that becomes 6.8V battery voltage. Above that you could destroy the MCU. The low voltage is given by Vf of the LED (plus a small voltage drop, say 0.15V in the 7135's) or about 3V.

So it is possible to run the Nanjg 105A with 2 x CR123A but not very safe/not recommended to do so over any longer period of time, if I understood correctly?

Haven't found one either. Well, with all the XM-L lights flooding the market let's hope that more different drivers will follow.

Thanks for the replies, brted & sixty545!

[note to myself: find out how to quote multiple persons in one post]

sixty545 if you can be bothered to do current draw for hi med lo at tailcap that would be sweet.

Spendid review. 4.70$ of mine are on their way to HK.

I first considered this as it works fine with my MC-E:

I will be happy to measure the current draw, but am not able to do it yet. Thanks to our dear government I could not buy a drop-in with XM-L which costs way over our import limit, so I had to make two orders of one XM-L star each at DX. That happened dec.23 and I have heard nothing from them since. I have a Skyray P60 host and reflector with pill ready to assemble my first XM-L light.

The percentage numbers in my review are from oscilloscope measurements of pulse width's.

I think it will be safe (but expensive) to use 2 x CR123A 3V primaries if you only look at voltages. But there will be generated lots of heat. The two extra Volts dropped in the regulator at 2.8 Amp will generate 2 x 2.8 = 5.6 Watt on top of the 10-11 Watt with a single Li-Ion cell. That must be a no-go.

Thanks for a great review. I especially like being able to eliminate the irritating strobe and SOS modes.

Yeasterday I got this XM-L from KD

I put it in a P60 with this driver

I put 4 volts direct to the led and measured 3.4 amps on my lab power supply.

With 4 volts through the driver I measured 2.7 amps. Installed in the light I measured 2.4 amps

Very happy with the lightoutput - amazing. The led was ordered from KD the 23rd and arrived on the 8th.

KD has improved a lot lately!

At how much is your import limited to? Mine is 22€ which is 29,xx USD atm acording to google.

I think the free of duty limit is set by the EU. The UK version is £18 which is roughly the same amount. But our Post Office doesn't seem to care much about stuff from China. Up until now, I have never had any order queried or opened. And I doubt Customs here are all that bothered.

Looks like a good driver. Was that measured at the LED or tailcap?

He is from Denmark and he has troubles getting a dropin which is roughly 17usd... that picked my interest.

Since the values on the packages are all incorrect, what our customs know, they check them from time to time. Sure they are unable to estimate the real value either way. But when they open one and they feel like asking for a proof, they usually charge 11 euros for the procedure, regardless if the receipt shows, it was less than 22 euros or not. When I order goods from a store, where I have to pay the shippment I don't split the orders. Usually, three out of four of them go through anyway without to get checked. So finally, when I count all together I still win.