NOthing analog here boys..but a great sounding unit. All hands on editing and a lot of similarities to D50 I guess with enhanced multi-mode filter modelling. I was blown away by these when they came out and hope to get one someday. (well in 2013 I think I finally did see one crying to be repaired and snagged it. )
RetroSound Has a great page with diagnostics. My friend has a JD990 rack version that acted very bizarre when the battery went down so always start there I suppose and do the reset. His actually just reset itself after he got the battery in or while it was out :-). Other than watch for dirty sliders obviously :-).
The keyboards are subject to getting too warm and having the key weights drop and adhesive get all over everything of course. I launched into trying to swap a JUNO-D keyboard in since they're relatively cheap (Roland has a couple keyboards left as of 3/2014 anyway but they want 330 some dollars plus ship which makes it not so good when you can get a whole keyboard sometimes for maybe 500 working? Anyway it's no money maker for techs that's for sure) So the problem with the JUNO-D is they arranged things quite differently I found out. Here is the pattern for the scan lines in JD800:
Coming off the junction board to the cpu, pin 1 and 2 are the aftertouch sense strip. Pin 3 begins the linear sequence of T0 through T7 lines, and then pairs of Brx and Mkx (where x begins at 0 and goes up to 7 also) lines which I assume stands for make and break? Probably their convention of expressing 'making' the initial contact that STARTS the cpu timing the note event, and then 'breaking' that timing cycle when the second contact hits. My wild guess. If anyone knows better let me know. Anyway in the JUNO-D they've changed this designation to PR (primary?) and SC (Secondary?).
Ok so the "T" lines on the JD800 work this way. Each "T" line connects to dual diodes that have their cathodes tied at the point of the T line. On the other side of the diode are one side of the primary and secondary contacts of the key switch. T0 is associated with the low C. T1 with C# and so on up to T7 on the G of course. Then the G# above that is associated with T0 again. This pattern continues up the keyboard to the 56th note where they only 5 more notes remain and so the T0-T4 lines are used in a short sequence there.
Now what happens on the other half of each contact pair? The primary and secondary contacts of the first 8 notes are just tied together to MK0 and BK0 for the primary and secondary contacts in those first 8 notes. And on this pattern continues to the top until agan there are only 5 in the last sequence for MK7 and BK7. SADLY the JUNO-D board does not exactly follow this pattern. The top C wraps around in fact unbelievably... to the bottom octave! Quite amazing that they would do this. I believe most Roland keyboards prior had followed the simpler pattern.
Roland still stocks many parts for older instruments! You can always order them through your local Service Center and see what happens! Otherwise I have some parts also so don't hesitate to ask.
Contact Sound Doctorin'