Sum bus theory and practice

[JR.]

I don't know that heavy lifting is required for the current to voltage convertor. I did one recording console using a TL072 and VCA in feedback for the master bus amp, since with current source summing, the noise gain of the bus amp is so reduced that the opamp's noise is generally not significant,

JR

[juniorhifikit]

Yes, I think so. You might could leave C1 and C2 if you wanted to.

if there's little or no dc at the output of the vca's, why bother?

That's a biased bipolar capacitor with the middle biased up to about 1.6V. Trevor also did that in the 82E149 mic input. In the 82E149 it makes a lot of sense.

I guess I was wondering what purpose it serves in the circuit...

[JR.]

I haven't been following the details of this particular veer, but in general the output of current ratioing VCAs while they don't contain significant DC, they do require a DC path to ground, to avoid whatever slight DC error term may be present from building up to a high enough voltage that it interferes with internal transistor operating points.

JR

[mediatechnology]

I had wondered about eliminating the coupling caps. I think "might" might have been the wiggle word. I suppose you could say that the virtual ground needs to be "virtual" at DC too.

[juniorhifikit]

I haven't been following the details of this particular veer, but in general the output of current ratioing VCAs while they don't contain significant DC, they do require a DC path to ground, to avoid whatever slight DC error term may be present from building up to a high enough voltage that it interferes with internal transistor operating points.

JR

I've been thinking about this in between other tasks, and I'm not sure I understand what you mean without seeing it in more detail. I read through your article several times, but I don't think I have enough basic theory to envision it; and I can't seem to find any sort of detailed drawing. Everything I learn here is compared against other examples that I've seen in practice, but unfortunately, this nifty idea of current summing has not been practiced that I can find!

If DC is blocked with a cap at the VCA's input, is there enough residual DC at the VCA's output to cause a problem? I don't know what level of DC voltage buildup would cause a problem.

Thanks for hanging in there with me!

[JR.]

there is another thread around here where I describe the control voltage side of making fader and pan with two VCA, is don't think I've published a schematic for a full mixer using VCAs because I never built one.

I have sold consoles using synthesized current sources, and the Peavey (AMR2400) consoles and documentation are out there.

-----

Regarding the DC at VCA output. And this may be difficult to visualize if you aren't familiar with discrete design. The output is like two different current sources connected together. They are very close to identitcal, so the net DC current will be very small, but if cap coupled with no DC resistance to ground, any small DC current error will charge up that capacitance until it saturates the one current source transistor that is sucking slightly more current. The will cause the VCA to behave poorly and probably distort on the leading edge of signals, until there is enough AC current to center it for a while. Or maybe worse...

I only know it wouldn't be happy, so DC couple it.

JR

[Gold]

I've been kicking around the idea of making a multiband Pico. I have my head in Lancaster's Active filter Cookbook for the front end. For the back end I would need a summing circuit. Since it would only be summing three stereo inputs would a different approach be better than if one was summing many channels?

[JR.]

Modern opamps are quiet enough that there is no need for killing brain cells over simple combining circuits, for a handful of signals.

For multi-band compression the heavy lifting will be the dividing network in the front end.

JR

[Gold]

Okay then. Thanks John. I guess I have the right book. Building an HPF/LPF set is on the list as well.

[mediatechnology]

Paul - Use a "derived" filter so that they re-combine properly. The old NSC audio handbook has a discussion on that in active speaker crossovers. We should probably start a thread on that.

Wayne