QUAD VCA Adapter Boards For SB4000/GSSL - Available (Soon)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ruckus328

Well-known member
Joined
Apr 14, 2010
Messages
906
Location
Philadelphia, USA
Finally got around to wrapping this thing up.  Managed to cram it all on there without having to stand parts up.  Prototype boards are on order, production boards will be available on my white market thread in a few weeks.  If you're interested please post here.

Multi_VCA_PCB.jpg
 
Looks good!

Just for clarification... there is a label saying OPA604 by the socket. I am guessing a standard NE5534 will still work there, right? Seeing as it only buffers the control voltage...

Cheers!
 
I think the theory is that you get 3dB (or is it 6dB?) lower noise for each doubling of the number of VCAs used in parallel? Perhaps better distortion figures too?
 
MikeClev said:
I think the theory is that you get 3dB (or is it 6dB?) lower noise for each doubling of the number of VCAs used in parallel? Perhaps better distortion figures too?
The theory is that there is a difference between how you add correlated and uncorrelated distortions and noise.

Correlated voltages and currents add up as you'd expect. So 2 VCA's in parallel give twice the signal current out.

Non-correlated Gaussian noise voltages and currents add up as power (due to the so called "random walk" effect).
Since power is proportional to current (or voltage) squared, 2 VCA's in parallel will only have sqrt(2) = 1.414 * the noise current of one VCA.

So any systematic problems and distortions will not improve (if all chips have a hump in the frequency response at 300Hz, or a non-linearity in the compression at a particular control voltage, so will the combo).

Random manufacturing distortions and noise problems will improve by 2/sqrt(2) when doubling the number of parallel units.

So four chips in parallel should increase the signal to noise ratio by a factor of 4/sqrt(4) = 2 = 6dB all other things being equal.

This is more of a replacement for a 202x (8 * 2150 VCA chips in parallel) than a 202C gold can (which was built on discrete transistors)
 
btw i see there is only one trimpot .. is it the magic one or the symmetry? i been trying for a bit to get all of it in one lil board like that damn good job!! :D
please keep us posted !!
thanks dude! u rock
 
ptron said:
btw i see there is only one trimpot .. is it the magic one or the symmetry? i been trying for a bit to get all of it in one lil board like that damn good job!! :D
please keep us posted !!
thanks dude! u rock

It's the Rmagic Pot.  The Symmetry Pot is already located onboard on the GSSL, but there's no rmagic pot located on the gssl, so in order to use a VCA expansion board it has to have the RMagic Pot on it to work, and as far as I know this is the only version that has it built in.
 
MikeClev said:
Looks good!

Just for clarification... there is a label saying OPA604 by the socket. I am guessing a standard NE5534 will still work there, right? Seeing as it only buffers the control voltage...

Cheers!

Yup, or course.
 
Back
Top