Rick's b.log - 2024/03/28
It is the 18th of December 2024 You are 18.119.213.225, pleased to meet you!

My laser printer is dead

Using a multimeter, I worked out where I could tap in to get a +24V supply, and where to inject the 3.3V for the electronics.

Adding some hooks (red=3.3V, orange=24V, black=0V).

Once the hooks were in place, I powered up the printer and adjusted the output until it read 3.3V.

The output power set to 3.3V.

That done, time to power down and attach the red wire to inject the 3.3V into the machine. The regulator has the ability to cope with 3A, though, maybe not for extended durations as there's no heatsink other than that built into the chip. At any rate, it ought to be enough to get some signs of life from the circuitry.

Only it didn't.

I measured the little motherboard's input as 3.14V, so I adjusted the power supply until it read 3.5V, which was 3.32V at the motherboard.
Probing around, something was 1.8V so if there is stuff running at 1.8V (not unheard of), then there's power getting into the board. But no signs of life when I pressed the power button.

So I think I'm going to have to accept that it's a dodo.

On the other hand, I have a module that can generate a useful voltage of my choice from a wide range of inputs. I think the input can be around 5V to 40V, and the output can be about 0.5V to 36V. The only limitation is that the output must be less than the input.

Even better, without changing its setup at all, I hooked it to the 12V supply for my keyboard and got a nice steady 3.3V out of it. So the output is controlled independently of the input. Nice.

3.3V on half the previous input.

OLED module update

128×64 OLED showing a slightly corrupt output.

It turns out that two bytes were being sent at the beginning of every pixel data transfer. I'm not sure what they were for, and it certainly seemed as if the original SSD1306 driver wasn't bothered by their presence, as the output looked okay on my old OLED.
The newer SSD1315 that's in the replacement display was less willing to tolerate this extaneous data, and thus corrupted the display by interpreting these bytes as pixel data.

So I removed those bytes and rebuilt the project... and cried a little at realising that I started my OLED driver module ten years ago. Damn, time passes too quickly.

128×64 OLED showing the expected output.

The archive below contains the OLED module, some tests and examples, documentation (don't worry that the documentation is for OLED v0.04, there have been no API changes), and the source code. It is licenced under the EUPL version 1.1.

Just messing around. ☺

Your comments:

David Pilling, 29th March 2024, 03:12

Magic of the 'buck' converter, switch mode power supply, see the inductor on the PCB. Keep switching the current on and off and the inductor generates voltage. There are 'boost' ones too with higher output voltage than input. It is harder work to get a bigger voltage than the input.

Joseps, 29th March 2024, 08:54

Sad to read the buck converter hack didn't work.     Thinking about it now I see the pics, maybe the 3.3 linear regulator that feed that line (as that probably is how it was fed) could be the one shorted and hence, messing with the buck converter supply.     Removing it could help, although once removed just testing it outside the circuit would be even better. If good, then there might be an intermediary power supply in the board being be the one failing, which could also be affecting other secondary voltages.