I decided it was time to reveal a little information on what is turning 
out to be a rather interesting discovery. I have long wondered if, as 
originally claimed by Yamaha and everyone else since, if the integrated 
circuits that comprise a CS-80 voice were in fact the chip versions of the 
discrete modules used in a GX-1 voice. Since the Yamaha SY-1 and SY-2 solo 
synthesizers, and the electone CSY-1 and CSY-2 combo organ/solo 
synthesizers used many of the same GX1 discrete voice modules and are 
imminently more obtainable than a GX-1, let alone a GX-1 voice board, I 
decided to buy an SY-1 and do some studying. 

This is not as trivial as it sounds. The Yamaha "NE" modules are 'potted,' 
that is, they are placed in shallow plastic boxes which are then filled 
with a liquid epoxy resin+catalyst which then hardens into a solid block. 
The circuit connections from the potted board to the outside are done by 
short lengths of buss wire soldered to the board prior to potting. Opening 
one of these is next to impossible without special solvents, and even then 
the task is not easy. 

In any case, I dismounted the high-pass and low-pass filter modules from 
the SY-1 voice board and used an epoxy solvent to try and loosen the 
epoxy. While it did not 'cleanly' do so, I did manage to uncover the 
circuit trace pattern and enough of the parts to reverse-engineer the 
filter circuits. The results were rather interesting: 

The GX-1 filters are *not* the same type as used in the CS-80. The CS-80 
uses the state-variable filter model, whereas the GX-1 (and SY-1, etc) use 
the Sallen-Key "diode ring" filter, where the diode array is the dual VC 
element used to tune the filter (both filter models are 2nd-order = 
12dB/octave). Now, the GX-1 uses a band-pass filter module as well that is 
not in the SY-1, but I am willing to bet it is also a S-K type, using a 
"fixed Q" resonant circuit. 

I already know the envelope generator modules are not the same as those in 
a CS-80, which leaves only the VCO module. The VCO is called "VCO-III" in 
both the GX-1 (NE11500 module) and CS-80 (IG00153 chip), but the raw 
output waveforms do not look the same. I suspect that, as with the 
filters, the tone oscillators aren't going to use the same model in their 
design as in the integrated-circuit. 

This leaves only the WSC (wave-shaper circuit) which, of all the modules, 
might actually be similar--and the most trivial to determine. The WSC 
creates pulse and sine waveforms given a sawtooth input. 

The end result is going to ruffle a few feathers like Yamaha whose 'party 
line' is that GX1 discrete circuits = CS-80 integrated circuits) when in 
fact it is turning out that the CS-80 and GX-1 are not the same at all on 
the circuit model level. 

At some point I'll have a complete web page on the subject, but I've been 
anxious to describe this research as it is turning out to be rather 
different than popular myth suggests. I am very interested in who was 
using the S-K diode quad filter model first, Korg, Yamaha or Steiner. I 
wonder if this discovery begins to explain why Yamaha seems to have 'lsot; 
all the engineering data pertaining to the NE module set...