[sdiy] Saw vs Triangle Sync

[Colin f]

> Oh, then you really *MUST* read the article(s) relating the 
> EM&M Spectrum. It uses the CEM chips, but hacks the sync and 
> has two different sync modes, and a knob for 
> soft-to-hard-sync tweaking.

Spectrum owner chiming in...
The definition of hard and soft sync varies depending on who you talk to.
The common theme is that hard sync does something more noticeable than soft
sync.
The CEM3340 does have hard and soft sync inputs on the chip, but I would
really call them soft and softer.
IIRC the CEM datasheet shows how to add an external transistor to force a
reset of the timing cap to give true, saw-core type hard sync.
That is what is implemented on the Spectrum, and also the Pro One.
The 'soft' sync on the Spectrum is really just a reduction of the window
that the hard sync occurs in.
There is no fancy tweaking of thesholds or anything - the sync depth
controls the pulse width of the synced oscillator, and the sync signal is
ANDed with the pulse wave. So the sync depth control sets the window during
which hard sync can occur after the start of the wave.
With a small window, sync occurs only when the two oscillators are close to
an integral frequency ratio.
There's an mp3 of how it sounds here:
http://www.colinfraser.com/spectrum/SpectrumSync.mp3
It starts with VCO 2 being modulated by the LFO; then VCO1 is added; then
sync is turned on at 100%; finally sync is set to 50%

Cheers,
Colin f