Cascading inverting stages

[Max9999]

Dear Diystompboxers,

I am building this virtual mixer followed by a phase flipper. I don't think I need C3, but maybe there is something I am missing .. Any reason to include it?

[antonis]

In theory, C3, isn't needed..
(but it makes no harm, also..)


P.S.
Beware of R3/R2 DC gain..

[ElectricDruid]

Just a comment about LTSpice:

You can put the power supply voltage source and Vref supply off in one corner, and then connect them to a "Net Name" tag (Draft->Net Name), so "9V" or "Vref"/"4.5V" or whatever you like. Then you don't have to put multiple 4.5V sources all over your diagram, you just copy the net name tag and connect them to that. So all your "Vref" points connect to a little thing that says "Vref" and your 9V supply points connect to a little thign that says "9V", etc. Much neater and easier to deal with. It makes it more like a typical schematic and less like a paranoid spider's web.

Took me ages to find out about this too!

[PRR]

> I don't think I need C3,

Put it in front of R2, because you need a DC block there.

[Max9999]

Thank you all very much. Ow yes I forgot a cap before R2.

So I was happily creating a pcb for this and then I got to the point of doing a Vref for the virtual earth mixer.
I read somewhere that there should not be capacitance connected to the + node of the mixer, because it could create oscillations at very high frequencies.

Is this really a thing? Would it for this reason be safer to use just a divider without a cap at mid point or a buffered vref? I don't own an oscilloscope to check.

[antonis]

I read somewhere that there should not be capacitance connected to the + node of the mixer, because it could create oscillations at very high frequencies.

True but not in your case..!!!
Even if your (linear) power supply is unregulated, higher ripple frequency should be limited at double the mains frequency..

It worths studing:
Single Supply  Bias & Decoupling

[ElectricDruid]

So I was happily creating a pcb for this and then I got to the point of doing a Vref for the virtual earth mixer.
I read somewhere that there should not be capacitance connected to the + node of the mixer, because it could create oscillations at very high frequencies.

Is this really a thing? Would it for this reason be safer to use just a divider without a cap at mid point or a buffered vref? I don't own an oscilloscope to check.

I don't know whether that's really a thing, but I've never had any problems with doing it. With the typical inverting mixer, the +ve input goes to the Vref, either directly, or via a resistor, which you can faff with if you want to try and minimise DC offsets, but you've got a cap on the output anyway, so you probably don't care. Gain for an inverting mixer is usually low, often less than one, since you're adding several signals together and you need to ensure you have enough headroom, so that helps minimise the chance for oscillations. I would always add a small cap in parallel with the feedback resistors (so R3 and R7) to limit the high frequency response. Since we don't want radio frequencies or ultrasonics, we shouldn't amplify them! With 10K, 470pF gives 33.9KHz cutoff, 680pF gives 23.4KHz:

https://electricdruid.net/rc-filter-calc/?f=25k&r=10k&r_series=3&r_errors=1&c=&c_series=1&c_error=10

[Max9999]

Thanks Antonis, I was aware of that document .. but reading it and understanding it are two different things unfortunately.

I will add the small capacitor Electric Druid, thank you for the tip.

[antonis]

Thanks Antonis, I was aware of that document .. but reading it and understanding it are two different things unfortunately.

ΟΚ.. Go to figure 3 and make RA/RB 22k/22k and C2 47μF..