[sdiy] Jp-8 head scratcher, solved or....need solvent

[Maarten Halmans]

Hello all,

a little update for the archives and for everyone who helped me on the 
tuning issues of the Jp-8.

I repaired the jp-8 some time ago. The problems in the 2 oscillators where 
not related to each other, altough.......
When I looked more precisely at the waveform I found some strange 
oscillating effect just after when the cap was reset by the JFET. I have 
seen problems with FETs in oscillators of a JP-8 but usually this was low 
drain source current extending the reset pulse or leakage of the gate. 
Replaced the FET and retuned the oscillator and the auto tune also tuned it 
spot on.
The other oscillator had a leaking JFET input of the integrator opamp. 
Replaced it, retuned it and voila, worked out fine.
My conclusion was: The error of the oscillation has more effect on high 
frequencies and the lower input impedance has more effect on the lower 
frequencies , therefore the autotune miscalculates the values for fine tune 
and scaling. Tuning by hand you can sort of make it work because you also 
have the high frequency compensation.

Now I have an MKS-80 in for repair that makes me think again about the 
problems I had with the JP-8. The MKS-80 had large dc feedthrough of the 
envelopes of the filters in 3 voices. I replaced one filter chip, worked 
fine, replaced the two others with no result. Tried serveral chips but no 
change. Replaced caps on both filters. One solved, the other still 
unchanged. Tried all kind of things with no result. Then I cleaned both 
sides of the board with flux cleaner. Problem gone.

Are the more people having troubles with the Roland flux?  There boards are 
always covered with a thick layer of flux and I have never had problems with 
it before or not that I know at least.

Cheers
 Maarten.