sync problems:soft

[Scott Gravenhorst]

mozepy at earthlink.net wrote:

>julian higginson wrote:
>> How the hell is the fatman designed for this to happen? If it is running
>> 2 oscs from the same CV, I can see you would get "identical" frequency
>> output from them, but surely you would still expect a phase difference
>> between the oscs?
The synchonization occurs due to coupling through the power supply.  
Technically, it is because of how the 555 timer is designed.  The pin 3
output, which is TTL compatible when powered at 5 volts, is a typical totem
pole config, ie two transistors, the top one's collector to Vcc rail, the
bottom transistor's emitter goes to ground.  The top emitter and bottom
collector together provide the output.  When quiescent, only one of the
transistors is conducting, however at the moment that the device changes
state, both transistors are on !!! at the same time !!!  Only for an instant
but long enough to induce a nasty little glitch.  On my website there are
instructions for adding diodes (LEDs really) and caps to isolate the 555
timers from the power supply.  In essence, they make up a mini supply from
which this transient current can be drawn.  
This mod (attributes to Scott Lee of PAiA) dramatically reduces soft synch
and allows much closer operation of the VCOs before they lock.  I got mine
as close as about 1/10 Hz.

-- Scott Gravenhorst
-- FatMan Site: http://www.teklab.com/~chordman
-- RedWebMail by RéÐ ¤ Š†år ¤ Wãrê
-- Linux Rex, Linux Vobiscum