I wonder if this efficient driver from Fonarevka has been tested in our forum. It has been around for two years now, I think. It is praiseworthy that they show the schematics (unlike some recent developers).
I don't think the driver uses PWM, but I can't tell for shure.
No but with schematic…and some AWESOME Eagle PCB geniuses and incredible coders…we might be able to 
Google translate
Drivers are no longer produced
(I did get the schematic, images, and source code and .hex file though)
PM if you would like a copy
Looks like a home etch job. Amazing that he got it to 17 mm and he shrunk it even further to 15mm.
oh yeah…and he had a stack of em! That was alot of work…too bad the guy quit making em
Very interesting find. Im surprised that this is the first we’ve heard about them on this forum.
Can’t read Russian 
No longer produced probably because it was unprofitable, it’s almost impossible to beat nanjg’s price.
Would be nice though, if it were available from OSHPark with parts list and some rough description (at least).
Might have had shipping issues as well from there. A 19mm version is the best available right now at >$10 a pop and people have been asking for years for a 17 mm switching buck driver. I’m curious what he’s done that was so inconceivable. Even the 19 mm ones have pretty mediocre efficiency.
not sure…
Datasheet on the ts5a3159
We know what the ATtiny58 does (a larger eeprom version of the ATtiny13A)
PWM pinout from the Tiny58 runs the input of that (pin 6), and that is connected to the FET thru the choke maybe a high frequency switching power supply of sorts? (or the output runs the electronic switch that runs the FET above it as a buffer to prevent the PWM whine from a Amtel driven FET [and we know what a proper MOSFET is capable of ])
IRLHM620 datasheet
http://ec.irf.com/v6/en/US/adirect/ir;jsessionid=D33DAE1EF6A107EE4529BE3FE856A436?cmd=catProductDetailFrame&productID=IRLHM620:
Doesn’t look like it has low voltage protection as pin 7 (thru the voltage divider resistors) is the pin for the voltage sense (I believe)
The circuit seems require a differential inputs which the Attiny13 doesn’t have.
Yeah, it uses he very nice 20x gain differential ADC which combined with the high-speed PWM makes those controllers almost perfect for such simple makeshift converters. (Others have even 16bit high speed timers).
Since the price is virtually the same, no reason to stick to the tiny13.
The analog switch is just misused as an gate driver for the FET. Would not be my first choice, but if it works . . .
It’s as simple as it gets for an diy microcontroller bases switching converter, I like it.
And no, no PWM besides the actual power switching. That’s one of the benefits if you directly regulate on the current.
Nothing inconceivable really. But he actually made the effort to build it.
Efficiency claims seem a little optimistic, my gut feelings would place it somewhere in the middle-to-high ~80% range.
But still nice and simple.
Thank's for all the comments. I can see now that it must have PWM but this is more or less filtered out by the coil and C2.