Deceptively simple USB charger

[Badcappy]

I have a tiny USB charger that decided to stop working completely, so I opened it up and found and fixed a loose supply wire and it works now, but one thing that caught my attention was how simple the SMPS circuit looked --- no ICs at all, only common discretes. I decided to trace the PCB to see if I could figure out how it worked, but the circuit (attached) was a little too difficult for me, someone with moderate electronics knowledge, to understand; I can see the basic parts like the primary rectifier, reservoir cap, transformer/feedback optocoupler, and secondary rectifier (half-bridge only, but not too surprising as the supply is only rated for 300mA), but how the oscillator/feedback actually works is totally not obvious. Noticing how small the reservoir is (2.2uF), does it somehow cleverly use the ripple current to start the oscillation? It sure starts up and gives a pretty clean and steady 5V at the output.

Can anyone here give their thoughts on this circuit, especially the primary oscillator? The secondary side is rather straightforward. Here's what I've figured out so far:

Q1 is the switching transistor; the top coil on the primary is the main switching coil, the second coil on the primary provides a low voltage (<25V) rectified through D8 and filtered by C4 to power the feedback loop
Q2 somehow regulates the oscillator via the feedback loop to control the output voltage
C2, R3, and D6 are the transient suppressor
C3 might be the oscillator cap
DZ1, a 4.3V(?) zener, sets the output voltage
Q3 and Q4 control the charging status LEDs

[POM_MJ]

Re: Deceptively simple USB charger

look like 2-transistors 5VSB circuit in ATX PSU.
Q1 Base-Emitter might be swap to circuit below, I guess.

[everell]

Re: Deceptively simple USB charger

I agree. Looks like a typical 2 transistor 5vsb circuit. Capacitor C4 is most likely the "critical capacitor". If it goes bad with age, the output voltage will probably increase beyond spec and maybe damage your equipment. It may be a cheap solution, but I prefer circuits that use a pwm chip for voltage regulation.

By the way, you did a great job on making the schematic, and your analysis was also good. Keep up the great work.

[Th3_uN1Qu3]

Re: Deceptively simple USB charger

It has an opto and a voltage reference. It's quite unlikely that a dried up C4 will make the voltage go up. Or am i missing something?

[everell]

Re: Deceptively simple USB charger

High ESR on C4 means fluctuating voltage (not good DC) feeding the opto and the two transistors. This is what causes the two transistors to drive the transformer harder. It is popularly called "getting Bestec'ed"

[everell]

Re: Deceptively simple USB charger

Here is my analysis of the problem

https://www.badcaps.net/forum/showthread.php?t=10301

[Th3_uN1Qu3]

Re: Deceptively simple USB charger

Thanks, i get it now.

[Badcappy]

Re: Deceptively simple USB charger

Thanks for the analysis, and the circuit does make a lot more sense with Q1 connected as POM_MJ showed, but unless the pinout on Fairchild's site for the E13003 is wrong (http://www.fairchildsemi.com/pf/KS/KSE13003.html), it really is connected as shown. I've attached the actual PCB layout with the components identified.

[everell]

Re: Deceptively simple USB charger

Check that Q1 one more time - I would have to agree with pom-mj correction. The R1 is a startup resistor. I don't think it would work properly the way you originally drew the schematic.

[Badcappy]

Re: Deceptively simple USB charger

Finally found the time to check Q1 more thoroughly, and indeed their datasheet was wrong on the pinout.

Also replaced the "critical capacitor" with a new bigger one (Nichicon 63uF/50V), just in case. Output voltage didn't change, but this PSU doesn't seem to regulate well with no load --- at less than ~2mA the voltage hovers around 5.5V, but with just a little bit more current draw, around 2.5mA, the output drops to 4.98 and stays there almost perfectly all the way to 300mA. Is this normal for simple switchmode supplies like this?

[POM_MJ]

Re: Deceptively simple USB charger

because of this circuit isn't just 5V power supply, it's output has config to current source for charging purpose.

D8.2,R8.2 and Q4 are current sense.
Q4 turn ON, LED11 lit while charging.

generic SMPS always need a load resister to maintain voltage regulation.