GGG Minimixer + Tone Control...?

[cheeb]

Is it plausible to insert the tone control below between the board input and the volume pots of the GGG Minimixer, after changing the input resistors to 10k for use with mics?

[Mark Hammer]

A circuit like that bleeds a lot of signal to ground, so you either need to: a) start out with a hot signal, b) stick a small preamp front-end between the input jack and this circuit, or c) be prepared to seriously goose the gain on the mixer stage and live with the ensuing hiss.

[Ben N]

Or reverse all the pots and put the whole kit'n'kaboodle in the feedback loop of an opamp?

[cheeb]

A circuit like that bleeds a lot of signal to ground, so you either need to: a) start out with a hot signal, b) stick a small preamp front-end between the input jack and this circuit, or c) be prepared to seriously goose the gain on the mixer stage and live with the ensuing hiss.

Since I'm using mics only, option A isn't possible for every input. Option B isn't really an option for recording, nor is it a good one for live applications. So i guess B is my jam.

By small preamp (in the DIY realm), any simple cleanish boost put between the circuit and the inputs would do, correct?

[Mark Hammer]

Here's an example of what you can do: http://www.aaroncake.net/circuits/mixer2.asp

In this circuit, R9 and R11 are the tone controls, and make a somewhat simpler network than your 3-way tonestack.  What is central, though, is they too bleed signal to ground, so we can generalize from this arrangement to what you want/need.  U1 is an inverting op-amp, whose gain is set by the ratio of the feedback resistance (R7) to the input resistance.  In this case, the unit appears to be designed for a substantial amount of gain in U1, and employs U2 as a unity gain noninverting stage.  (Myself, I'm not comfortable asking that much gain out of a single stage, so I'd rather split the gain up.)  You will note that the level pots are wired up "backwards".  The idea is that two things happen at once as the level control is increased: 1) the input resistance is decreased, and 2) the amount of attenuation resulting from the combination of the ground leg of the pot and the input resistor (R1, R2, R3) is also decreased.  Both of these will result in MORE signal coming out of U1.  personally, I would probably not go above 100k for R7, and would stick a 100-150pf cap in parallel with it to keep hiss under control.  In U2, I'd replace the wire link between pins 2 and 6 with a 22k-33k resistor, and stick a 10k from pin 2 to ground, for a gain of 3-4 in that stage.  Depending on how much hiss you get, you can also consider sticking a feedback cap in there as well, say 330pf.  naturally, a DC blocking electrolytic cap hangs off the output of U2 and goes to a master volume pot.  Near as I can tell, your tonestack can easily fit between the output of U1 and the input of U2.


Here's another: http://www.zen22142.zen.co.uk/Circuits/Audio/6ipmix.htm
I won't post this schem here because it's kinda of big and splits up the text too much.  This one is a little different in that it uses individual op-amps to provide individual gain stages for each input, where the one above uses a single op-amp and provides different a amount of boost/gain for each input.  The values shown for this second one provide a fixed gain of x100 for each mic, and the 10k pots after the op-amp attenuate down from there.  Those attenuated signals are then mixed down to mono and fed to a common inverting op-amp with a gain of around x2.  With any given level pot maxed, the input signal is boosted roughly 200 times, which is quite a healthy degree of oomph to feed a subsequent mixer.  You will note that inputs 4 through 6 on this second circuit have no x100 front end, merely a passive attenuator before going to the x2 mixer stage.  You will also note that because the 3 mic inputs use an inverting stage, and the output/mixer op-amp is also inverting, the final output signal has the mics in phase where the line inputs are phase-reversed because the mics are inverted twice, and the line inputs only once.

From your first question, I take it you'd like to have tone controls for each input?

[cheeb]

Yes, tone control for each input is my aim.

In reviewing the second link, merely leaving off the line inputs and adding another mic input identical to the others would be plausible? And is x200 gain really necessary?

As far as the first link, I like it, but in having a separate tonestack for each input, it seems to me after some more thought that a separate gain stage is required for each input, followed by the tonestacks for each input, and the gain recovery by an all-encompassing gain stage just before the output, i.e. the second link.

So am i correct in assuming that I can merely insert this (or another) such tonestack after each input's first gain stage?

Or is this asking too much gain from the final stage?

[Mark Hammer]

Yes, tone control for each input is my aim.

In reviewing the second link, merely leaving off the line inputs and adding another mic input identical to the others would be plausible? And is x200 gain really necessary?

Certainly not for making up the signal level lost from a passive tonestack like the one you're intending to use.  On the other hand, voice mics have a much lower level than a guitar, and if you plan to lose even more signal via a tonestack, then mere hiss-management demands that the initial input goose the level.  A x50 stage is not unreasonable.

As far as the first link, I like it, but in having a separate tonestack for each input, it seems to me after some more thought that a separate gain stage is required for each input, followed by the tonestacks for each input, and the gain recovery by an all-encompassing gain stage just before the output, i.e. the second link.

Correct.

So am i correct in assuming that I can merely insert this (or another) such tonestack after each input's first gain stage?
Or is this asking too much gain from the final stage?

You are correct, though you overestimate how much gain is needed in the subsequent stages.  Remember, it's multiplicative.  A x50 input stage, is enough to get you safely past a passive tonestack, and a x10 stage after that is enough to feed subsequent devices.

One consideration, though, is the degree of isolation needed for each input.  The tone stacks CAN be inserted between the input and mixer stage, but they are not isolated from each other so adjusting one will always provide an altered path for the others.  In other words, they will likely interact in a big and largely unpredictable way.

So what can you do?  Well, one decision point is to ask how much individual control you want, relative to group control.  So, do you really want/need midrange adjustment at the individual input level, or can you be satisfied with some very basic tonal control of individual inputs and save the tonestack for the mono mix?  It is a very simple affair to calculate the 3db down point in the input to an inverting op-amp.  In the second circuit, the 10uf/1k combination yield a 3db rolloff point at 16hz.  Drop the 10uf cap down to 1uf and its 159hz.  Drop it again to .47uf, and its a 6db/octave slope below 339hz.

At the other end, caps in parallel with the feedback resistor set the high end rolloff.  If you made the resistor 47k instead of 100k (gain of x47), then a 220pf feedback cap would score you a rolloff above 15.4khz.  A 330pf cap moves that down to 10.3khz, and a 470pf cap moves that down to 7.2khz.  So, would an input stage with a level control and two 3-position toggles (low-cut and high cut) be sufficient control for each input if there was a 3-band tonestack after that?

If you want/need more than that, then the optimum strategy is to run each input through its own tonestack into a second op-amp stage made up of a noninverting unity-gain stage (output pin goes to inverting input pin).  That second op-amp will isolate the tonestacks from each other, and you can feed THAT op-amp output through a summing resistor to a mixer stage with some gain in it.  You can then either use a variable feedback resistance to adjust the gain of that stage, or use a fixed gain and stick a volume pot on its output to use as a master volume.

[cheeb]

So, would an input stage with a level control and two 3-position toggles (low-cut and high cut) be sufficient control for each input if there was a 3-band tonestack after that?

Yes, that would suffice just fine. So at what point would be the appropriate landing spot for the final summative tonestack? And how much signal will be lost with four sets of high/low cut toggles?

[Mark Hammer]

The suggested toggles would actually lose very little signal, unless what you chop off is a substantial portion of the spectrum.  For instance, the rolloff switches often found on voice mics only lose a little signal when you engage the bass rolloff.

The "appropriate landing spot" lies between a final and next to last op-amp.  If you had 3 inputs, then the total chip count would be 2 duals and a single, or a quad and a single.

[cheeb]

Sounds great. I'll get to working on a layout for it and I'll ask you to check over my shoulder (if you don't mind) when I get it finished. I'm pretty excited about this. Thanks for all your help so far.