@Doomedahab I'm glad that there are still some active people in here after all these years and i appreciate all the help and compatibility reports. So: thanks for keeping this thread alive!
Thanks for pointing me at the right direction! I would never ever figured it out by myself... not in a million years!
Here's what happened:
I was looking at C60... lifting it up on one side to measure it properly. It measured around 11.6nF... which seemed okay. So i thought about replacing it with a 1nF capacitor - as recommended. And then i saw it! The solder pads were connected!! I had soldered C60 and R46 this way:
... instead of this way:
That's so embarrassing!!!
So i rotated both components... and now: the AC/DC detection works absolutely fine!
Btw... is there a suitable replacement for the Onsemi MMBT5551 used for Q4, Q5 and Q12? According to the 'G1' markings on my transistors i used the MMBT5551 from Diotec:
... which should work fine (i guess). I don't know for sure, if they are genuine ones (but i always tried to use reputable sources for the UniSolder parts... so: they theoretically should be genuine). I now ordered some Diodes Inc. MMBT5551s... which (according to the datasheet) should be fine too:
I just have a cheapo ESP32 based 100kHz max. handheld oscilloscope to check VINT and Vacdc... but i guess (while not perfect) that should be good enough.
But AC/DC detection still doesn't work. On my other UniSolder (5.2 without c) when i switch the soldering station off, i get a "POWER LOST" message for the last second before the screen gets black. With the new UniSolder the screen only quickly fades to black... without that message.
After replacing all capacitors on the OLED PCB - which (unsurprisingly) didn't change anything - i desoldered the OLED flex cable and measured OLEDVCC on the powered on station. It gave me a reading of 8.725V . There's obviously something wrong with the display. So i've ordered a replacement.
I the meantime i've found out, that i had flashed a slightly older firmware onto the PIC. It had the bootlogo - so it couldn't be THAT old - but there were some differences compared to the latest GitHub version. So i flashed the latest firmware.
I don't know if it'll fix the AC/DC detection problem for me. I'll report back when the new OLED has arrived.
Thanks for now @minkok for taking the time to help!
I even replaced the coil. Then i lifted every capacitor and every resistor on one side, measured them all one after the other and compared the results with the BOM. There was nothing obvious. All the measurements were within the tolerances of the components .
After that i resoldered the resistors and capacitors and cleaned the PCB thoroughly. The result looked like this:
But sadly it didn't change anything !
I don't know what else to do. I even ordered a new OLED display and consider ordering a new OLED PCB and new capacitors...
Which is a 3V one (as it should be). But i checked the resistors for the voltage divider... and while the 1kOhm reads 0.999kohm (which is within the 0.1% tolerance) the 27kOhm reads 26.565kOhm (which isn't within the 0.1% range). Maybe this could be the problem?
I power my new UniSolder 5.2c with a toroidal transformer... but the display says "DCE" :
Don't let the picture fool you: compared to my first UniSolder build the OLED looks very dark and dim (even though "Brightness" is set to 15 in the menu).
Maybe someone has seen something similar and can help:
I've build my first UniSolder 5.2c... the voltages from the back board were fine... so i connected everything and flashed the firmware -> there was no soldering iron connected and the UniSolder SEEMED to work... but the OLED was incredibly dim and there was a dark horizontal bar in the middle of the display.
I figured out, that i had used the wrong value for R68 (3k instead of 4.7k for my 9V display). So i replaced R68 with a 4.7k resistor... and it made everything worse. The OLED was now almost too dark to read under normal light conditions. So i changed R68 back to 3k.
Then i measured the Vin and drain pins of the LM2731 boost converter i've used -> Vin read 3.3V but on drain i was getting around 3.0V. So i tried something: i changed R66 to 33k (instead of 27k) and left R68 at 3k. Theoretically i should have got now 14.8V from the drain pin of the boost converter. But i only got around 3.3V. This improved the display a bit. But still: something was totally wrong there.
After that i measured all capacitors, resistors, the coil and the diode on the OLED pcb (i even removed all parts with unclear readings for testing them out of the circuit) - but everything seemed fine! Even the voltage on the feedback pin of the boost converter measured 1.23V - which afaik should be on point.
So... did i have a broken boost converter? Or could it be something else? Any ideas?
Maybe because mouser has very high shipping costs for small orders from outside the US. So most people rely on local sellers... and not all of them have the PIC32MX534F128H in stock.
Because i have a leftover vinyl sticker for the front of my 5.2 i want to use for the 5.2c too. Those were laminated from a shop specialized in vinyls for arcade machines - so they weren't cheap.
Thanks for all the input and suggestions. I've had some delivery problems regarding the MJD122 i bought. But i received them in the end... and now everything works perfectly fine.
I'm planing to build a 5.2c too. But i want to design an addon pcb for replacing the tactile switches with touch buttons (probably using some TTP233 ICs and an acrylic front plate... or maybe some sandwiched thin acrylic plates with cutouts for getting the touch pcb closer to the front). Should be possible... shouldn't it?