The LED strip stopped working.

[AleksandrYanovskiy]

LED strim and TFT worked for a long time on PROXMOX without any problems.
Then I noticed the LED strip stopped working. I tested the LED strip itself on another PC – there are no problems.
When turned on on another PC without disks and memory, the LED strip starts working immediately.
On this PC, it doesn't work at all. There is no signal on the middle pin of the LED strip connector.
I couldn't find the signal path on the board and I also couldn't find the circuit diagram.
Can you tell me which component on the board controls the LED strip and what sends the control signals even when the PC has no memory?
Or is this a software issue and can I somehow enable the LED strip?
Thank you.

[tjaworski]

Both the LCD and the LED controller are connected to the USB bus:

Bus 001 Device 004: ID 1a86:7523 QinHeng Electronics CH340 serial converter  <-- LED
Bus 001 Device 005: ID 04d9:fd01 Holtek Semiconductor, Inc.  <-- LCD

By default, the USB bus provides power to both devices. When powered, the LCD initializes its firmware, enables the backlight, and displays the clock. The LED controller also powers up and initializes to its default state. Therefore, even without any software intervention, the LED should turn on.

The LED controller has also failed on one of my devices. the design does not appear to prioritize long-term reliability. Anyway, can you provide a photo showing the traces that were tested?

[AleksandrYanovskiy]

Yes, all the USB devices are present, and there are no boot errors.
I only found where the +5V and ground from the connector go.
I checked the signal path across all the accessible transistor switches, as well as other components on the back of the board, but I still haven't found where the signal goes.
I suspect I need to remove the heatsink and look underneath.

[AleksandrYanovskiy]

I managed to find where the signal goes. As I thought, it's a microchip under the heatsink, near the power button. U31, pin 2. The microchip name is erased, so I couldn't read it.

[tjaworski]

Have you tried jumping that pin to +5v, does the LED come on? I wonder if this could be a CH340N USB-to-UART TTL converter, something like this: https://uge-one.com/product/ch340n-usb-to-uart-ttl-converter-ic-smd-sop-8. Maybe you could hook it up to an oscilloscope.

[AleksandrYanovskiy]

Oh, no, it's not a CH340N.
The CH340N chip in the picture above - is U3. There is a marking on it.
It sends a signal from pin 6 (TXD) to pin 7 of an unknown chip.

I was short on time, so here's what I found after examining the pins of this unknown chip:
Pin 1 - VCC
Pin 2 - OUTPUT to LED strip
Pin 5 - Q98 MOSFET DRAIN (smd 72k - to the left and above the chip)
Pin 6 - Q19 MOSFET DRAIN (smd 72k - to the left of the chip)
Pin 7 - INPUT from CH340N
Pin 8 - GND

The transistors seem to be working properly. I haven't found where the Gates are connected yet.

When power is applied, a single pulse appears on pin 7.
The same signal is also present on pin 6 CH340N.

+5V is present on both chips.

There is no signal on pins 1(UD+) and 2(UD-) of the CH304N or I didn't see them.

There is no signal on the 2nd pin of the unknown chip, although on a working board there is a series of different pulses.

So, the cause of the malfunction has not yet been determined.

I can't take measurements for long — the CPU gets very hot without a heatsink.
I need to attach some kind of cooling plate first and continue my investigation.

[AleksandrYanovskiy]

Here's what I've found so far.
Signals on the working U31 pins:
1 - +5V
2 - OUT LED
3 - +5V (some time after on)
4 - ?
5 - IMPULSE 250ms (some time after on)
6 - IMPULSE 250ms (some time after on)
7 - INPUT TXD from CH340N
8 - GND

When turned on, the working chip automatically generates a signal for the addressable RGB LED strip.

The non-working chip has no signals on the following pins:
2, 3, 5, 6.

All pins have the same voltage drop in diode test mode as the working chip.
The 3rd pin raises a question: is it an output or an input? I couldn't find where it's connected on the board.
Perhaps the presence of +5V on it turns on the chip.
I haven't checked it because I've already wasted too much time.

Perhaps someday a schematic of this board will appear and we'll be able to understand how the backlight works. I couldn't find the chip itself. There's no marking on the working board either. It's most likely either a special chip for addressable LEDs with programmable capabilities or a microcontroller.

I currently see four options:
1 - сontrol of the U31 chip doesn't work due to the lack of the necessary signals upon power-up.
2 - the U31 chip itself isn't working.
3 - there's a software error and the chip simply needs to be reprogrammed.
4 - build your own LED strip controller on an ATtiny.

Any information on this matter would be greatly appreciated.

[tjaworski]

this sounds like a tiny mcu but without a part number it is just a guess. I doubt we will ever see any schematics for this. you can try asking on the https://www.minipcunion.com/viewforum.php?f=110

[AleksandrYanovskiy]

Good news. I was able to connect the LED strip to the Digispark. It's expected that the microcontroller will be able to be connected with minimal intervention to the PC board. In the future, I want to add the ability to connect it to the built-in UART CH340N and programm control the LED strip.