I think the guy who designed this single cell charger for LiPo batteries must have been paid by number of components used. I've no idea what half of this stuff is supposed to do. This model was produced in 2002 for a couple of years. What d'yall make of it?
Only the first 3 pictures in this folder, guys. The rest of the pictures are of something else entirely out of a scope, so just ignore them:
How so? I see, hmm,
TSM104A = quad op-amp plus ref, roughly a low current LM324 + TL431 in one (I wouldn't even be surprised if an x-ray shows two dies). Probably for voltage/current regulation/monitoring?
ST52TO1C6 / MPACE0320D = ??? Can't find anything on this. Can you confirm my reading?
I would guess it's a battery management device. Or some MCU. Not sure what all the transistors are for, but it would seem like they happened to need a lot of level shifting or logic and it wasn't worth putting in gates to do it.
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
ST52 perhaps some kind of 8051 clone/copy/derivative ?
I think it may be ST62, which is an 8bit uC
I'm thinking lipo suction may be in order here.
There's a thumping great fat power diode immediately to the right of the transformer. What's that doing there??
Close! It's actually ST*6*2TO1C6 / MPACE0320D
The chip below left of that is marked: M104AI / 9N3316
The quartic shaped device immediately above the white socket is an STA353YT1
Output side rectifier -- it's a flyback topology SMPS. Guessing B340 or somewhere around there?
Likewise, the capacitor just above the opto (C11?) is the main filter cap. (If you have some of these that are wheezing, that would be a prime suspect, as well as C01 and, can't make it out, but it's right beside the "U01" label.) D07 is the primary side aux supply rectifier. Oh, and D51 seems to be supplying a second voltage rail on the secondary, just noticing that. Which goes right over to U41 (the opamp/ref), which might be for high-side current sensing? (It doesn't have a common mode range that includes the positive supply, so that would be helpful.)
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
Ahh, that explains it. 8 bit MCU, 2k OTP prog, might as well be your bargain basement 8051! Euch, stack levels, maybe PIC would be more apt.
Looks like the STA part (2A 40V PNP, roughly equivalent to, oh, halfway between 2N4403 and TIP31?) might be high side control or sense. Maybe it's got a discrete LDO post-regulator, or they're just turning it off when complete (since after all, the charger can't keep its brains running if it actually shut down the main supply..). R47 would be current sense (0.1 ohm).
Thinking the quad op-amp is for voltage and current sense, the MCU orchestrates everything (possibly including current programming), and the dusting of transistors and resistors is for level shifting and feedback gain ratios.
Current could also be controlled by feedback to the TL431 (or whatever it is exactly), in which case the STA353 would only be needed to disconnect the battery when charging is complete. A current sense resistor would be needed in either case, but being positioned near the transistor, and having enough op-amps to go around, it's probably a current source.
That should be a pretty complete "rough" description of everything; if you want to learn more, I'd suggest picking up some appnotes on the devices (or related, e.g. TOPSwitch SMPS controllers; battery management chips; etc.), and if you want the schematic for further study, knock the components off and trace the connections. For MCU code, you might be able to hack a config pin to read it out (how this might be done, is not obvious, and very device specific), but it's quite possible you'd need to decap it and read bits with a microscope. A behavioral model is probably much cheaper to derive (run it through its paces and see what the controller does; synthesize your own replacement and compare results) than any of these methods.
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
I'm somewhat bemused about the need for this size of diode, too. Having hooked it out, it turns out to be an SR306 which according to the datasheet is a Schottky good for 3A forward current. Given the stated output current max is about 750mA, this does seem very generously specified. :-/
Well I've no intention of trying to reverse engineer this thing. It seems
*absurdly* complex for what it does. It developed a fault which I haven't the time to trace, so I've binned the guts of it now and replaced it with a charge controller chip, a MCP73831 which Artie here recommended. It only needs two caps and a resistor to make a fully functioning charger and away it goes. Tried, tested, works - at tiny, tiny fraction of the component count and cost of the original. Was this VASTLY simpler solution not available in 2002 when the original charger was produced??
The peak current for a BCM flyback is four times DC, so it's up there. Not that peak is what matters in the rating, a 1 or 2A part would be fine. It could even be an SOD-123F or SMA on the underside, who needs THT.
Maybe they wanted to go towards the efficiency. Though using a 60V part seems counterproductive when 20V would do. Noteworthy that schottky diodes have quite a bit of capacitance, and can actually perform worse than PN diodes in some cases; using an oversized one would probably be such a case.
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
Maybe also possible that it was a device they were already buying and in stock? Sometimes you use overkill parts from other products if you get a purchasing advantage.
Possibly! Or it was more expensive than the pennies of parts seen here. Remember, China production, phenolic board (in the >100k production qtys), wave soldered. They didn't even get a chance to orient the parts correctly for wave soldering, and there's a number of suspicious joints on there (like resistors that are right next to each other, with a big solder blob spanning them -- not good for strain on the metallized terminals). Which could be a thing, cold or microscopically cracked solder joints.
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.