[sdiy] twilight zone question

[Jason Proctor]

i have a twilight zone situation in a module i've been 
building/fixing for seemingly aeons and i'm hoping the assembled 
experts can throw some light on it.

to cut a long story short, i made a midi sync / pulse divider module 
from a combination of the blacet midi sync pic chip and the cgs 36 
pulse divider board.

the pic demands +5v so that sits on a protoboard with a cmos 4538 
chip to stretch its pulse output. the clock output is boosted from +5 
to +15 by a tl072 based comparator, so that it can drive.....

.... the CGS board, which runs on +15/0v and which provides power to 
the 7805 regulator on the protoboard, does all the division and 
output buffering. i separated the /3/6 and /2/4/8 stages in order to 
cascade them, but that's not that significant here.

the problem in a nutshell is that +5v in the +5/0v protoboard 
environment is not +5v in the +15/0v environment. i had to trim the 
reference voltage on the aforementioned tl072 comparator so that the 
clock "low" output of 0v wouldn't trip the 2v comparator. and 
similarly now it seems that the 0.3v "low" reset line is tripping the 
2v comparator on the CGS board so with the line connected the 
counting never gets going.

what the **** is going on? i've measured everything under the sun. 
the grounds on the boards are physically connected and are within 1mv 
of each other. i had this crackpot theory that the 7805 regulator was 
giving me the "top" 5v instead of the "bottom" 5v, but the +5 coming 
out of the 7805 is exactly 10v away from the +15 measured at one of 
the counter chips on the CGS board.

any ideas with this mofo appreciated. this is design #7 on this 
module and i'd like it to actually fully work sometime in my 
lifetime....

j