Now I know why was the diode ommited: having fun with the CN3761

ShangHai Consonance Elec CN3761 | 2020-12-20, 09:57:00

After I toyed with the CN3791 board, I designed my own.
My version uses the CN3761, which doesn't have the MPP tracker.
The datasheet, in chinese, is almost identical to 3791's one.

I complained in my previous article that the D1 diode was not routed on the circuit and now I know why: it has to be a very fast switching one, probably to switch synchronously with the mosfet.

I've tried adding one and it won't work with it.
This was the test circuit, and I've tried quite a lot of things with it, such as that step-up regulator from another board. I wanted to try two mosfet types, one that comes in a soic package and the other one routed to the bottom.

This board, as it is, works.

The text on the bottom is the same I had on the 4056 board, and I should've had it changed, this IC actually pulsates the leds when the battery is missing.

Unfortunately, this is the only board that works and I still haven't discovered why.
Proud of myself, I was very quick to merge the design from the board I've cut with this one and ordered a new one, the result being this:

I could not believe that the CN3761 wasn't working after the other IC's outputed stepped-down voltages from the battery without any issues.

For some reason, the regulator simply won't charge, at all. It turns on, the charging led stays on, no charge delivered to the battery. Unlike in the working version, it won't pulsate if there's no battery, so it's like the internal logic won't work at all.

I did a third and fourth board, without the step-down regulators, in order to isolate the issue, maybe there was a problem with those two on the board, and, again it wouldn't work again.

I'm not sure whether it's just bad design or the IC has such a high failure rate or I've been sold rejects from QC if any.


Honestly, I'm not interested in further study since it already took me an unreasonable quantity of time.