Just spent the past couple of hours cleaning up an running some preliminary checks on my latest toy - a 7904A 500MHz mainframe oscilloscope with plug-ins. Half a can of cleaning solvent later and all of the controls are now noise free and everything (with a minor triggering fault on one of the timebase plug-ins) appears to be in order. Just checking everything out before I go to town on it replacing all of the electrolytic capacitors and giving it a full factory-spec calibration.
Second only to the 1GHz 7104, this model is right up there at the pinnacle of analogue oscilloscope design/performance. The jitter performance of the 7B85 timebase is pretty impressive, as shown in the 1nS/division screen photo. The 7B85 can trigger up to 400 MHz and has a maximum sweep speed of 10nS/division. 1nS/division is achieved with the x10 multiplier switched in. The majority of the jitter shown can be attributed to the phase noise performance of my ancient vacuum tube Advance model E2 signal generator, providing the 100 MHz test signal. The intensity control for the 1nS picture is only about half way up. This scope has a VERY bright CRT.
Like the 7104A, the vertical deflection amplifiers feature custom amplifier IC's mounted on micro strip printed circuit boards with custom chip carriers. This scope and the plug-ins have been given the "option 3" treatment, which is EMC screening.