47k are output resistors (signal to ground) two small SMD after the 10uf. the last components before signal leaves the board.
I found this values completely wrong.
Concerning the capacitor substitution.
I proceed as resistor explained before, with two more wires soldered on capacitor holes and alligators, I test the whole range of Elna silmic-II.
Start from 10uf, which betters the stock Russian, then 4,7uf. fount it more coherent, then 3,3uf. further coherency and balance, finally 2,2uf. best of all, highs opened without loss on bass depth. Easy. Now I have on the unit this litlle Elna Silmic-II and Vishay to ground for easiness in faultfinding, but when finish the whole upgrades of this good thread, back to bulky PIO k75-10 and choke.
Modification of K75-10 capacitor is a must to juice at full potential:
Ah! concerning the repair, thanks so much LosNir for help
, I agree for damage on TDA1543 IC's.
By chance, do is someone able to test on a healty IC, the diode value (from ground to other seven pins)?