If the button is actually a mechanical switch, then replacement is an option. If it is truly a touch sensor input type device then the problem is not the button, and the diagram in post #37 is very misleading.
Can we get an explanation of how the button is used to select the brightness?
If it is truly a touch sensitive control arrangement then someplace there is a circuit that does more than OFF/ON control. But if it is a mechanical button switch then replacement will be simple.
But from what I can seein the picture of the PCB, the device "U3" is an IR reflective sensor that may be intended to detect touches from behind a clear spot in the mirror. Not at all a touch sensor. So if somehow the assembly had moved from the intended position, the sensor would not be in the location to see the reflected IR light. In that case the problem is mechanical and the solution is putting the sensor back in exactly the correct location.

And the actual button is just a location..
Others need to look at the pictures that show U3 in more detail and see what they can see. I see an LED and a photodiode in U3.
And now, having read that moving a finger makes the LED blink, that has verified that it is the photo sensor. Theelectronic switching transistor, T1, is functional if the LEDs can be switched on.But the current requirement may be excessive.

I don't understand what can be misleading...
I literally have a picture posted of the button, and I said that's the only PCB between the simple LED strip and the 12V DC transformer (LED driver 230V AC to 12V DC). The LED driver works and the strip works too because by default it's turned on and that's how it is. So the only thing not working in the circuit is the tactile/proximity button, which PCB is shown in this thread. What's misleading about this?

No, the button of course is not a mechanical switch... it's a tactile (advertised as such), which basically means, and I said that several times too, it's a proximity button, you can clearly see the IR in the board, and also in the 2 videos (check first post, all the info and videos are there).

In the videos you can check how I interact with the button, it's not that it's not located in the position it should be, of course it is, it has worked in the past, so the IR sensor goes through the clear spot in the mirror. Anyways, I was testing it outside the mirror. Check the videos.

My uncle used this mirror, not me, but I guess the button works like this:
1. If you activate the sensor by putting your finger next to the button and put it away, it's an ON/OFF gesture.
2. If you maintain the finger after 1s or so, I guess it will dim the LEDs up or down, or something like that.

 

The misleading part is the implication that inside that rectangle is a simply replaceable switch, while in reality it is a whole circuit board full of components.

 

The misleading part is the implication that inside that rectangle is a simply replaceable switch, while in reality it is a whole circuit board full of components.

Where's that implication you are talking about?
Never stated that the button is a "simple switch", not only that but from the beginning I've posted a clear picture of the button, which happens to be an IR proximity button, advertised as "tactile button", really common button in these mirrors, and yeah, happens to have an IR sensor and some other components. My explanations and drawings are clear as water, IMHO. Also... if the button was as simple as a mechanical switch, I wouldn't probably have made an abstract box, I would have drawn a switch, just like the cables are drawn as cables, lines, not abstract boxes that say "connection".

Anyways discussing this is nowhere near the point of this thread.

 

Okay, I think I'm just gonna buy finally a generic IR button for 12V. These boards are sold, and I don't need the exact same model I have, I just simply need a button that can power and control a 12V LED strip. Since my strip is 30W, and most of these buttons support up to 60W, there won't be any problem!

By the way, I think I know why the strip has 4 wires connected to it. I think it's because of voltage drop, just to make sure 12V are 12V in all the LED, as some strips are long enough that would be a factor to think about. May be here is overkill, I don't know.

 

Fixed it!

Replaced the board by another one compatible, 12V and IR sensor, and now it works fantastic!

However, that new sensor does not work THROUGH the glass/mirror like the original board does. The new IR sensor detects the mirror/glass as an object and it gets constantly triggered (pressed). So I will have to put that sensor on the bottom of the mirror, so the new gesture to turn on the lights will be swinging the hand below the mirror. May be even better than the original gesture.

 

Well done!

 

This is the final result!


It's positioned on the back of the mirror, aiming down, so you simply swipe your finger or hand a bit below the mirror and it turns ON or OFF. Also, if you keep the finger below, it modifies the intensity.

The input of that button is 12V (40W max), and the "case" looks like aluminum but, even thought the board is elevated with a spongey sticker, I was worried about the back of the board's soldered pads touching the case, creating a short circuit. However, I measured continuity and the case is not conductive (as I was expecting by the design). Anyways, even if it was, I isolated the back of the board.

I played around with the case, throwing it at my wood desk, and it really sounded like it was made of metal and not plastic. What material could it be?

What do you think!?

 

There are composite materials that are very hard and at least as stiff as some aluminum alloys. In addition, there is hard-anodized aluminum that has a passivated surface that is non-conductive, like glass. So those are two possible materials. Since his mirror package is a consumer product, my guess is that it is anodized aluminum, possibly with an organic coating on top of the anodized layer.