[sdiy] seek help on Jupiter 6 repair

[jdm at synthcom.com]

At 03:53 PM 9/13/2002 -0400, Robert O Anderson wrote:
> That the
>problem could be digital or analog is an important point, unfortunately it
>expands the possibilities.

You got that right.  But the tuning is probably the best place to start.
The VCO outputs are fed into an 4051 (IC20), where they are sequentially
selected to be fed into a comparator whose output goes to the INT0 line
(pin 12) of the CPU.  The interrupts are timed to figure out what the
frequency is.  The 4051 inputs are as follows:

pin 13 - VCOA1
pin 14 - VCOA2
pin 15 - VCOB1
pin 12 - VCOB2
pin  1 - VCOC1 (4 voice only)
pin  5 - VCOC2      ""
pin  2 - VCOD1      ""
pin  4 - VCOD2      ""

Put a scope on 'em and see if any look significantly different.


>I've repaired Oberheims where the culprit voice can be quickly determined.
>Does the Jupiter 6 provide any similar indication of which voice is being
>tuned? 

Unfortunately, no.


>Which board is tuned first, the 2 or 4 voice?  

They're both tuned at the same time.  Each board has its own CPU which does
the tuning.  If it tunes ok, the CPU changes the state of its TX line (pin
11).  These are wire ORed together and read by the master CPU.  If the
state never changes, then the master CPU hangs.  Pretty stupid design if
you ask me.  Why not use the UART to return a failure code?


>Does anybody have a voice card schematic.. 

Sorry, I have to keep my nose absolutely clean in regards to Roland and
their copyrighted materials.

JDM