Looking for: 17mm, single AA driver with programmable ATtiny. Does one exist?

Indeed :bigsmile:

RMM has just developed a very cool 4A / 6.8A buck driver and if he would come up with something similar in the boost department I’d be very happy.

My contribution to wight’s WIP boost driver originally was meant only as a support for wight, some troubleshooting, some testing to see if it’s worth for him to prod further. Then he went out to lunch… miss him, too.
Status is: concept seems to work for 1AA; 2AA killed me a boost controller; I’m testing some different diodes and inductors to find smaller components as the Nanjg110 ones dont fit 15mm. I know why RMM uses these expensive CoilCraft inductors for his buck driver, they are really good at a small footprint. I already have what seems to be an ok 17mm Eagle board and a possible 15mm one, but I need a final component decision yet for the layout.

As an additional idea for OP: If we can’t get an ATtiny boost driver going, take a look at the LD-30 boost driver as a possible (but not programmable) alternative. More expensive, but for 1 or 2 AA and with modes (sometimes 3, sometimes 5). HKJs test of the LD-30 boost driver. IOS and Fasttech do not stock them anymore, but Banggood has them and LCK-LED seems to be the origin.

If I can find my spreadsheet I found a few cool boost ICs a few months ago that might work out well for this type of application and allow for higher amperage.

That said, the solution of piggybacking or stripping existing drivers for parts is the lowest-cost solution available. Even buying in bulk it is often cheaper to buy complete drivers to strip for parts than it is to start fresh, so working with that in mind is probably the best way to go for a budget driver. When you're building a lot of drivers this can be time consuming, but for a few it can be a good deal.

The Coilcraft inductors are the best in the smaller sizes, but they're really expensive. Almost as much for an inductor as I would hope to be able to build this entire driver for.

Hi occulai. Welcome to the forum.

It should be quite easy to swap in an Attiny13a into the driver you linked to in Post 2. You need to be able to remove the stock MCU (I recommend a hot air gun) and air wire in the Attiny. Only 3 wires (VCC, GND, and PWM) would be needed unless you go momentary switch.

Once you remove the stock MCU, you power up the driver and probe with DMM to verify the VCC and GND pads. Then short the VCC pad to the pad that appears to be for PWM to verify that the led lights up. You now know the 3 pads that need to airwire in an Attiny.

The MCU is the largest component at top of the below picture. Counting counter clock wise starting at the divot (Top left pin), the pads appear (guessing as the traces are not clear. Could be totally wrong.) to be:

Pin 4 VCC

Pin 5 PWM

Pin 8 GND

EDIT: Here is the pinout for the Attiny13a (link to source in picture). Most of the BLF FW has PWM going to Pin 6.

Yeah, it looks like they're bootstrapping the MCU's power off of the output.

If that’s the case then it should be possible to come up with a board that can be populated with borrowed components plus an Attiny. This should be fun.

Anyone with a dead board they can strip and map traces?

These pictures may help,this is basically the same driver,without MCU and transistor?

https://www.fasttech.com/products/1612/10001535/1114501-1-mode-led-driver-circuit-board-for-flashlight-diy

^ Sure looks like it. Good find Iraksat.

Combining all the pics, one can see almost all the traces except for what is going on under the large diode. Appears that one trace is going under is from the Inductor/Capacitor to the bottom right leg of the boost chip there.

So MCU pads 1 or 4 can be used for VCC.

EDIT: I searched my driver stash, but could not find a driver like Post 2. I did a quick sketch of what the circuit appears to be based on the product pictures in the above links. I haven't tried to double check it or anything. Just a start.

EDIT2: Based on the purposes and positions of the boost chips pins, it appears to be a PAM2803.

EDIT3: Reading the discussions on the product pages, someone that appears to know what they are talking about says its a PAM2803. The data sheet says max input of 5v (min .9). So the potential applications are limited to low voltage, but it can be up to 90% efficient. Ordered some of the ones linked in Post 2 to play with. Also ordered this.

Thank you for the schematic!

Following the datasheet of the PAM2803, the trace under the diode will surely be pin6 to Inductor/Bat+/Cout, as you drew it. In all probability pin4 (on/off), which you drew unconnected, will be connected with pin6 by a trace under the controller.

There are 2 points I find interesting.
First thing is that the MCU (the 8pin microcontroller) is not connected directly to GND, but to the other side of the current sense resistor (R150), which is the LED- of the controller (pin3) before the FET. Is there any reason for that?
Second thing are the 2 additional resistors (PWM to MCU-/pin3 (333) and LED+ to GND (180)). Could anyone explain what they are for?

As the controller of my Nanjg110-with-Attiny-and-FET blew up at first use with 2xAA, I’m still looking for answers why this happened. So I assume the mentioned details on the above driver might help avoiding current spikes. Ideas, anyone?. Thx.

^ You're very welcome. Thank you for helping me with Eagle in our PM's.

Thanks for the info on the likely tract to Pin 4. When I get the driver, I'll strip it down and post pics of the board.

I have many gaps in my electronics knowledge because I learn enough to accomplish a particular goal as I go along. Hopefullly, I'll eventually have a semi-rounded electronics knowledge base. I haven't learned how boost circuits work yet. So I have no substantive input on your observations and questions. All that is said so that folks take my statement with a grain of salt. Your assumption that the R150 may help in reducing the magnitude of spikes makes sense.

Regarding the 33kΩ that connects PWM to "ground", that is pretty typical in buck and linear drivers. I think it may be referred to as a pull down resistor as it pulls down residual charge between the MCU and FET in this case.

I think resistor "18D" is a EIA-96 code which translates to 150kΩ of resistance. It is not on the non-MCU driver that member Iraksat found. So maybe it helps the capacitor smooth voltage spikes or maybe speeds up voltage drop for mode changing. With such a high resistance value, it doesn't seem like it would to much to tame spikes. So it's more likely the latter or so other purpose.

Hopefully, someone knowledgeable with these types of circuits chimes in.

EDIT: Fixed some typos and added more content.

I thought the fasttech link said it only works with 3-4.2v?
hmmmm

Sorry a wrong link,here is the correct one…. https://www.fasttech.com/products/0/10001750/1127402-12-aaaaa-1-mode-800ma-linear-booster-led-flashligh

Any updates here? I am looking to mod my Thorfire TG06 and found this thread…

http://lygte-info.dk/review/DriverTest%201-2%20AA-AAA%205-Mode%20800mA%20Linear%20Booster%20(YJP)%20UK.html

I got a few of these drivers but haven't had a chance to play with them yet. I'm hoping to be able to piggyback an attiny on them to get some good modes with hopefully little effort or cost! Yes, I know those are often lofty and unrealistic goals...

So, can any one copy and test this one? I have one left. I don’t think they are sold anymore. It uses an Attiny 13.

I’m just going to wait patiently for RMM.

I’ve bought several cheap 1-2AA boost drivers

(put aa boost into the search box to find various mentions of such)

I put those into older lights to gift them with 18650 size NiMH batteries (thinning the herd).
None of them are flashaholic-pleasing drivers though.

Maybe I can send mine to Toykeeper or someone else to see if they can do anything with it?

I would love to get something made for everyone.

+1 As this is a pending project for me too.

18650 NiMH batteries? Where do you get them from? That's a pretty cool idea.