Hello, time for a new report.
My friend removed the C11 and measured it with a tester and multimeter and he got:
tester:
C11 ( 10uF 50V electrolyte)
10.68 uF
ESR 1.1 Ohm,
Vloss 1.4
Multimeter:
10.03 uF
Judging by the measurements this should also be OK.

- 1.JPG (77.27 KiB) Viewed 954 times
0.793A measured across the fuse. I think this is in spec.
The yellow channel is the voltage on C11, the blue is the output from 4070B (both values correspond to the 5V, they are just offset). All was measured without the keyboard.
The voltage on R18 is a stable 5V.
But he changed the C11 nevertheless for another brand new one and this resulted in stable reset (yaay

) (for now at least

) , so Mark was spot on once again!

(thank you Mark)
The new C11 is 10 uF 25V, ESR 1.5 Ohm
Now the image was stable for at least 20 minutes as far as he tested.

- 2.JPG (73.27 KiB) Viewed 954 times
Now this looks like another bad 4116, which I can handle as they are all socked now.

(the image is black and white, because we've changed the link on the video board to B (bypassing the mixing chip) during the testing. Does this sound reasonable? As we've tested all of the video ram connections for continuity and for shorts.
Any thoughts? Was it all because of a possibly weird C11? He did report, that he got a stable picture with the old C11 after he soldered wires to it for measurement (it's possible that he fixed a cold join with this? Or changed some properties of C11 with scope probes attached to it?), but after manually resetting on the keyboard connection circuit exempted the old symptoms again. He doesn't want to try the reset again with the new capacitor

so this is it for now
Any other ideas and suggestions?
As always thank you for your help.