I updated the question with a final, correct schematic and derived R27 value.
Also solved the (now funny) “enigma” for the SMD marked ‘390’:
- First I thought it was ‘398’ (nonsense way huge R value)
- Then thought it was 39R and I had a “dotted” zero as last digit
- Turned out, final “digit” was not any digit but a ‘B’ !!!
Finally I noticed that this is an EIA-96-coded resistor (another nomenclature standard), see e.g.:
EIA-96 SMD coding standard
So marking was really ‘39B’ which means a 2k resistor (measured it)!!! Why on earth -No pun intended! :)- does this not follow the usual marking convention, i.e. ‘202’ for a 2k value? I never faced this before (then I’m not totally used to SMD components in my ideas and designs).
If someone can explain this marking’s logic, I’d appreciate.
Anyway, R27 must be a 2k, too, for the voltage divider to feed about 2V to the FB feedback input on the switcher IC. This “reference” FB value is mentioned in the Datasheet linked above, although there it’s shown the case for DC 12v output and quite arbitrary R values are given there.
The R values here for the 5V output case are quite convenient as you can see.
Here you can see the Diagram included in the "Application" mentioned in the Datasheet for the switcher IC, which inspired me to solve my own design's determining my R27 value (remember, this is for 12V output, so not entirely similar to my 5V case, and the labelling for the resistors and other components differ also from my schematic's):
OH! BY THE WAY! the right mode in which this design is operating is NOT the CCM (Continuous Conduction Mode) as I said initially, but the MDCM (Mostly Discontinuous Conduction Mode). Reason (according to the Datasheet) is that, D4 -in my schematic- being a "fast switching" UF4005 diode, 75ns Reverse Recovery (RR) time, is not "super fast", meaning that if you want the IC to work in the CCM "region" then this diode would have to be even faster than about 35ns (I don't know of any available replacement in this sense, and in any case this would only get us about 130mA maximum output current).
So, case closed I guess !!
Now it remains for me to figure out an appropriate "entry point" in which I should disconnect this design from the mains, in order to replace fully the Power Supply section with my own, "black usual brick" AC-DC adaptor. This is because, as I mentioned in another Question of mine in this website, I plan on fully controlling this Coffee Machine with ESP01+AtMega Tiny85 (i.e. project for a WiFi commandable machine), and the ESP01 module alone can draw as much as 750mA during WiFi-searching periods!!!
I guess that I could "cut" some connection in this schematic, to be able to add my own 5V supply with enough "beef" for this. BUT, I need to figure out:
a) which connection is going to be the GND common to the microcontroller's logic, since this also affects the gate voltages for the TRIACS in the original board, and
b) how to deal with the design being (oddly in my opinion) such that, it's the NEUTRAL which acts as positive "rail" in the AC-DC conversion here depicted, and how does this affect proper isolation for my newly (replacing) external 5V power supply.
This quirks I have yet to investigate deeper, no hurry.
Any further info/ advice/ remarks will be hopefully made in a separate thread, since this note closes more or less fully the original question I had.
Thanks to all you folks!