Got this unit in for some power supply issues and did a few specific boards per customer.
Initially the relay was engaging longer over time about 12 seconds, verified voltages and the H805-807. Turned out to be a leaking cap that tested bad under duress. The relay is on an ball, a 4dpt type. This particular one was pulled and the contacts treated and cleaned. Used here were Panasonic FR's and FC's. Also the 131B diodes were replaced with MUR1620 common cathode diodes. Monitored and regulator adjusted to 35.5VDC, relay sits at about 5 seconds.
The axial was replaced with a Nichicon TVX and Silmics and WIMAs used for the tantalums. H409 diode was replaced with two 1N4148 in series and heat-shrinked. H405/406 (2SC1175A) were replaced with matched hfe KSC1845's.
This board has the nice fat browns and the skinny yucky raw films. The gray films at CE05/06 were replaced with Panasonic ECW PP type. CE37/38 are bipolar lyrics, replaced with MUSE's. The rest are Silmic and WIMA's. Mind 5mm spacing through for lead spacing on the WIMA's
Each boards differential pairs (Q701/702) were replaced with matched KSC1845's with compound and heat-shrinked together, the KSC1845 has the appropriate pinout BCE. The bipolar were replaced with WIMA MKS2's and Silmic used throughout. New thermal compound and MICA insulators installed.
The biasing procedure is extremely easy on this unit with the test points mounted specifically to the sides, top mount. There is a service error in the manual, Bias is adjusted at R727 and DC offset is adjusted at R726.
A new eg1018-nd switch and X1Y2 safety capacitor installed, along with both dubbing jacks, both shaft had been broken from the body of the assembly, I added an eyelet to the grounding jack of the jack to make the common ground easier to share.
To be honest I didn't expect much out of this unit, I was already convinced the 2325 is as good as you can go. Turns out I was wrong, the 2330 has that nice Marantz depth to the soundstage yet this modern…clear overtone I haven't heard in other models.