Archive of the former Yahoo!Groups mailing list: Yamaha CS80

Concerning the 0.01µF caps: This part of the circuit is to create the glissando from the portamento: the low-pass filter just smooth the steps of the glissando to produce the portamento. It is explained on pages 24-25 of the CS-Series servicing guide.

Depending on the Portamento Short-Long slider position, two values of the resistor are used, whereas at least in theory, the resistor value should vary continuously with the slider. So resistors are already approximations and certainly a precise value of the capacitor is not that important. 

The change of the value of the caps was probably a last minute improvement. If the portamento sounds good at any position of the Short-Long slider, the value of the capacitor is certainly correct.

D-oh!   Thanks for the reminder Joachim.  I knew that but had a mental block.  Still not sure whether it’s this particular CS80 or the docs.

To remind myself - and for the amusement of others - the glide/portamento is based on the glissando setting.  Basically, the DAC chip gets the portamento clock (PC) which varies with the slider.  This tells it how fast to transition, in semitone steps, from one note value to the next.   To turn this into a smooth glide, as Joachim wrote, the individual semitone steps are smoothed out with a simple resistor/capacitor circuit that’s switched in using analog switches.  As he wrote, this RC time constant should really change with the glide time.  It doesn’t and Yamaha cheated by setting it to a “good” fixed value.

However, they are doing something a little better - and kind-of weird.  I never looked at this part of the logic very closely, but if you check the big, fold-out schematic ( http://therogoffs.com/cs80/manuals/CS80_Service_Manual/cs80_foldout_schematic.jpg ), they are doing some odd stuff with comparators and logic gates to basically turn the port/gliss switch value and the port time slider voltage into a 2-bit digital code that controls the 2 analog gates for each voice and can therefore switch the smoothing time between 4 values, including basically instantaneously for glissando and 3 different RC time constants based on the portamento time slider.  

So, the smoothing time doesn’t constantly vary with portamento time - which would be the best case.  But it does switch between 3 values to at least roughly track with the portamento time.  This doesn’t answer my question and actually makes the situation more complex :)

 David