Seem it's way to go good old way, if want something done - do it by youself
Will try some usual "try put some parts and see what happens" way later.
For now teardown, as whole internet don't have a single hi-res photo of 4263B internals...
So here we goes, all photos click-able for hi-res.
Front panel
Back I/O
Unit is not deep, but wider than usual Keithley/HP bench meters, so it will not match in rack as for size
Remove cover by two screws from rear:
Bird's view
Disassembly pretty simple and user friendly. All can be done in few minutes. Well thought design layout.
Input BNC connectors.
LCD is standard 40x2 char type, with very well known 44780 controller. Maybe later I'll change LCD to something with better contrast, if repair will be successful.
Protection on inputs?
Mains protection and filters.
Unit low voltage power supply in it's cage. Two 22000uF 25V KMH capacitors and lot of small 22uF 50V electrolytes.
Here goes digital board, mounted on standoffs above signal acquisition bottom board. MC68HC000FN10 CPU, two Sony branded CXK581000AM 70ns SRAMs, Xicor X28C64 ROM, AM27C2048 UV ROM with firmware (04263-85041 on my unit) and some big National Semi S9652AB SCX6B65AHI ASIC, and bunch of 7400 series logic and AMD-branded PALs around.
No batteries to store cal/configuration constants or NVRAM, only supercap there. So this unit pretty safe from dying from dead batteries, like some other gear
Some sections done with PCB-PCB modules, one of which near standoff had burnt resistors and SOT-23 resistor.
It's similar to other presumably alive module, so should be not very hard to find out what it does.
Good module:
It have PNP and NPN with small heatsinks on another side.
Input terminals protection and switching. Probably some stuff is borked here...
I already put some resistors to check basic functions, but it's not working right, readings sometimes, all over the joint often, sometimes seem ok.. Need debugging
Analog/switching/measurement components