Preamp/mixer design help

[ShutterCat]

Preamble: I've never done this before, so it will be rough.
I'm putting together a 41hz.com amp6 kit, and I'm adding a preamp/mixer stage to share the same enclosure. I sketched up a circuit in TinyCad. There will be 6 stereo inputs in the final revision. This drawing has 2.

The input level control uses dual ganged linear pots per
http://sound.westhost.com/project01.htm

The opamp configuration is per Figure 1 on
http://sound.westhost.com/project94a.htm

The opamp is a dual-supply type powered from a single supply with the non-inverting input biased to half of V+ by the resistor network.

I'm not sure about the placement/selection of filter caps.

DIY-ists, am I on the right track?

My Preamp/Mixer Schematic:
http://i3.photobucket.com/albums/y65/FuzZyXL/preamp1.png

 

[PigletsDad]

I don't think it will work as drawn.

You need a coupling cap in series with R4. R4 is not actually needed.

You have two resistors labelled R3. One goes from V+ to ground, as does nothing useful (lets call this R3A). The other (lets call it R3B) goes from V+ to the noninverting input on the op-amp. You need to move R3A so it goes from ground to the non-inverting input - this will then sit at half the supply voltage.


Because this is a single rail design it will have a horrible turn-on thump, so you may need to start the power amp up after the mixer/preamp.

 

[ShutterCat]

My mistake. Drawing a voltage divider while not paying attention ><

Fixed:
http://i3.photobucket.com/albums/y65/FuzZyXL/preamp1_1.png

So, the coupling cap you said to add is to compensate for DC bias in the input signal? Inputs at system ground and op-amp circuitry at virtual ground?

I plan to use the AMP6's mute function to mitigate the startup thump. (switch the amp on with the mute on, then switch mute off)

 

[ShutterCat]

Could I use an OPA2134 here?

 

[AndrewT]

I plan to use the AMP6's mute function to mitigate the startup thump. (switch the amp on with the mute on, then switch mute off)

Hi,
the active pre will still send a massive pulse into the power amp input. The mute, with a long enough delay, will silence the output if the input survives the mis-use.
Can an amplifier accept a half rail voltage pulse?

 

[Nigel Goodwin]

My mistake. Drawing a voltage divider while not paying attention ><

Fixed:
http://i3.photobucket.com/albums/y65/FuzZyXL/preamp1_1.png

So, the coupling cap you said to add is to compensate for DC bias in the input signal? Inputs at system ground and op-amp circuitry at virtual ground?

I would disagree with PigletsDad, you don't need that capacitor to the inverting input - although R4 (as he said) shouldn't have been there.

Where you need to add capacitors if on the feed in to the faders, to prevent any DC to the sliders (which would feed through to the opamp). This prevents not only any DC to the opamp input, but also prevents the pot being noisey - DC through the pot makes it noisey as you move it.

 

[Nigel Goodwin]

Apologies! - didn't notice it was fed from a single supply

The capacitor is required, but needs to be the other way round (positive to the opamp), also there needs to be one decoupling the two resistors generating the centre point.

 

[PigletsDad]

You also want bleed resistors (say 1M) on the outputs, from say the junction of C1 and R6 to ground, so the output doesn't float up if the load is disconnected.

I agree that a bypass on the non-inverting input is a very good idea, as it will vastly reduce the sensitivity to noise on the PSU.

 

[ShutterCat]

Thanks for all the great input! I've turned C2 around, added C3 on each of the inputs, and a 1M bleed resistor on the output.
Is it necessary to have both C2 and C3? Won't the first cap stop DC on its own?

Here's what it looks like now (right channel omitted for clarity)
http://i3.photobucket.com/albums/y65/FuzZyXL/preamp1_2.png


I may have to make the mute switch disconnect the speakers in the final design. It turns out that the "mute" pins on the amp6 are actually connected to the tripath chip's "sleep" function so that it thumps anyway when you turn off the mute/sleep switch. My concern then is the life of the mute switch. Arcing may decrease the life of the switch if I'm not careful

I know that my labeling is a little lax. Components with the same values have the same designation. I'll fix that in the next revision and as I'm picking out values.

 

[Nigel Goodwin]

C3 (and the other one) are required to stop any DC feeding the sliders, which will make them sound noisey - but I would suggest they should be the other way round.

You also haven't added a decoupling capacitor across the bottom resistor feeding the non-inverting input, and also aren't showng one across the supply rail either.

 

[ShutterCat]

You also haven't added a decoupling capacitor across the bottom resistor feeding the non-inverting input, and also aren't showng one across the supply rail either.

I relabeled the components. and turned the input caps as you suggested. I'm not sure if this is standard, but I used
Type/Channel/Input/Number
R123 is a resistor on the first(left) channel, second input, and it's the third resistor. C101 is the first cap on the left channel that is not connected to any one input
http://i3.photobucket.com/albums/y65/FuzZyXL/preamp1_3.png

Okay, so you mean a decoupling cap in parallel with R103 and one from V+ to ground?

(the TI reference refers to it as "AC coupling" I have to assume the difference is regional nomenclature)
Should I post my .dsn files for tinycad somewhere?