Hello Holger,
The transparent blue casing of the adapter is [url=http://www.conrad.de/scripts/wgate/zcop_b2c6e13574c/~flNlc3Npb249UDkwV0dBVEU6Q19BR0FURTAzOjAwMDAuMDJkZi4yZDA4NjIzOSZ+aHR0cF9jb250ZW50X2NoYXJzZXQ9aXNvLTg4NTktMSZ+U3RhdGU9MjU5NDM1MDc2Ng==?~template=PCAT_PRODUCT_DETAILS_DOCUMENT&object_id=0000000000000000000000000000000000522471&object_guid=E42DD83B3952E67CE10000000A010221&p_load_area=203036&text_ztab=Typ%401591+BTBU+%7cFarbe+%28transparent%29%40Blau+%7cAbm%2e+%28L+x+B+x+H%29+mm%40112+x+62+x31&page=5&hmmh_alt_image=%2fIMS_Docs%2f76%2f76C5953EECAEA88DE10000000A010251%2eJPG&p_selected_description=Geh%e4use+Bis+120+Mm+Kantenl%e4nge#l] here, which I bought from Conrad in Cologne. The OBD connector, which I bought quite expensively on eBay based on a tip I found, I mean it cost 15 euros for this plastic part.
The curves I recorded were all obtained using freeware versions of VAGCOM, and they include at least the first 20 data blocks from the engine control unit.
The circuit is the one designed by Maik.
minus
The Golf IV does not have an 'L-Line' trim level.
The driver transistor for data reception is not necessary in this case.
plus
2 specialized optocouplers and a small DC-DC converter (a great piece of equipment, a bit more expensive) to power the MAX232.
A different LED configuration is needed so that the receive LED only indicates data reception.
and
not to mention.
the blue voltage LED in the transparent blue housing (a must-have for the Golf IV)

.
To do this, dust off an old laptop (mine has a Pentium 75), a VGA screen with a resolution of 640*480 is ideal, and you'll need a power supply for the car and VAGCOM for automatic startup

.
Well, of course, something has to break on the car so that men have something to tinker with.
At least, I once properly cleaned my Mass Air Flow sensor (picture here), after 115,000 km. It was an ALH engine, specifically the side that is exposed to incoming air, i.e., the area where you can't reach without removing the sensor, using a Torx screwdriver with a hole in the middle, from the pipe fitting. I recorded the before and after curves. Afterwards, the values were back within the normal range

, meaning that at full load, the smoke limit was again above the torque limit.
I used isopropyl alcohol for cleaning, but I think it's too aggressive. It's better, as Diesel-Gustav (I think his first name starts with H

) described, to use Pril diluted in water, because I only saw the finest soot, that stuff that we all have in our lungs

.
@Everyone
I've been reading here for a long time, so I'd like to take this opportunity to thank Ulf and Michael II, and of course everyone else, not forgetting Rainer, the owner of the website. The interested layman (apart from Diesel-Gustav, who is probably an expert) is capable of amazing things, as I can see time and again when I read here (the last entry in the error database, written by Ulf, about the ticking sound when accelerating, is brilliant).