Circuit diagram for Amazon/ebay/aliex 15W USB load

[Lupin III.]

Hello!

Has someone reverse engineered the circuit for the 15W USB load that's available all over Amazon/ebay/aliexpress? I'd like to use it to discharge li-ion cells (small pouch cells or 18650s) to storage charge. But it is only constant current and has no lower voltage cutoff, which I'd like to add.

[DavidAlfa]

Why bothering when there are cheap devices in aliexpress that to do exactly that?

[Lupin III.]

Do you have an example for that? The only thing in the same price class (I will need multiple of them) to discharge are the ones in the picture below. But they are not what I want, because it needs external power (the main downside), the current can't be set easily (you have to exchange the power resistors) and it isn't constant current either. I also don't want the "smarts" of those. I want a "dumb" device that can just be connected to a cell and off it goes without having to push any buttons or worrying if the cutoff voltage is still set to the right value.

[kripton2035]

something like this : the FZ35 ?
https://www.aliexpress.com/item/4001089537342.html
https://www.eevblog.com/forum/testgear/fz35-adjustabe-electronic-load-questions/

[Lupin III.]

Too "smart" and too expensive. As I said, I will need multiple of those (5+).

The original device is exactly what I need (easy of use - no buttons to push, adjustable constant current range up to 3A, powered by the battery itself, integrated cooling with temperature sensor, cheap, on a small PCB). It lacks the voltage cutoff though, which should be easy to add since it's all analog components. That's why I was asking if someone has seen or reverse engineered the circuit of it, which is still the point of this thread (unless there is a device that does exactly the same as the USB load, but with voltage cutoff). If none has that, I will have to do it myself. I just wanted to avoid that as there are quite a few tracks hidden unter components and the silkscreen is covering a few of the thin tracks as well.

[DavidAlfa]

That seems just a simple circuit using an operational amplifier with feedback from the shunt resistor and a potentiometer to set the gain ratio.
If you want to add voltage cut-off, you will have to add some circuitry.

You could disassemble it and take pictures, it's not a big deal, only 3 screws.

[Lupin III.]

You could disassemble it and take pictures, it's not a big deal, only 3 screws.

I never claimed that it's a big deal. But it is an annoying one, that someone might have done already  . Since none has apparently (well at least none who read this forum; how dare they  ), here it is.

The circuit I added is the blue block, set for 3.85V cutoff (the power transistor of the load will turn off). No guarantees for completeness.

Thanks for trying to help, even though the actual question was, as is so often the case in forums, replied to with "why don't you..." instead of answering it

[ledtester]

Just noticed in your blue-box area that you have a NPN and PNP both designated as a BC857.

[Lupin III.]

Just noticed in your blue-box area that you have a NPN and PNP both designated as a BC857.

Ahh, right both being BC857 (PNP) is correct. I fixed that symbol in my small schematic, that I used to lay out the components for the proto board, but I forgot to change that in the full schematic. I will upload an updated version later. I made a few other small changes as well, but they didn't even make it to the small schematic yet, only to the real proto board .

[DavidAlfa]

Nice work there!
Yeah, what do you expect, if you already searched and found nothing, I doubt anyone had reverse engineered it and not publishing anything

[Lupin III.]

The slightly improved (and corrected) version. I added a capacitor to the 431's reference pin, because the whole circuit was oscillating at a few kHz. It still does, but only at a few 100Hz now. When the circuit reaches the cutoff voltage, the power transistor closes, the voltage on the li-ion cell rises above the cutoff voltage again because of the removed load, the power transistor opens again, the voltage drops, aso, aso... I don't mind the oscillation itself. After all the only purpose of the circuit is to discharge one li-ion cell to 3.85V, and it does that oscillating or not. But at the higher oscillation frequency the MLC caps on the board caused an annoying audible whine . Adding hysteresis somewhere could completely stop the oscillations, but as I said I don't see any point in throwing more components at it.
The LED I added turns on btw. when the input voltage is below cutoff voltage to indicate the discharge has completed.

[DavidAlfa]

Had a moment and I decided to try designing something similar.
This is what I got. It has self-latching, so once it triggers, only removing the power will reset it.
Will probably need some tuning when doing the real thing.



C1 will will ensure that Q3 doesn't receive any power at power-on, so Q2 forcefully turns on before Q3.
Also will filter the noise when connecting the battery and prevent too easy triggering.
D1 will quickly discharge C1 when power is removed.

At first, Q2 will be always turned on, until the input voltage gets low enough to release R2.
R2 will slowly charge C1, until it turns Q3 on. At the same time, Q3 emitter will turn on Q1, permanently latching Q2 base to gnd, no matter the Vinput.

I omitted any zener or regulation circuit.
Since a bipolar transistor base-emitter union has a 0.6V threshold  (Yeah, not amazingly accurate, but I think it's okay in this application), a simple voltage divider might work.
Adjusting RV2 should be enough to correct any transistor tolerances.

In simulation, the circuit cutoff range was between 4.2V(RV2 set to min.) and 2.9V (RV2 set to max), the complete voltage range of Lithium batteries

I'm not a designer engineer so don't blame me too much! 

[John_doe]

Sorry for the necroposting, but I just wanted to mention that I used the schematic by Lupin III. in my thread here:
https://www.eevblog.com/forum/beginners/a-question-about-a-commercial-15w-electronic-load/