Help me out making a simple mixer

[soggybag]

I want to add a mixer in a box as part of another circuit. This is what I came up with in my head. I'm not sure if it will work. IF anyone can tell what I'm gettign right and what I've got wrong that would great.

[CynicalMan]

Your volume pots will vary the input impedance of the circuit a lot, which could cause problems. You'd probably want the inputs to be wired like a volume pot, then series resistance, like this.

[Transmogrifox]

+1 on RG Keen's mixer.

Short answer is, your plan will work. 
The caveats are this:
You will have a hard time getting any of the channels down to even 20dB attenuation without using really large-value pots. For example, if the feedback resistor is 10k, then each input pot would need to be >1M to get better than 20dB attenuation. 

The capacitors would need to be pretty large since the varying pot value changes the high-pass cutoff frequency.  With this much resistance change on the pot, you would find some significant DC offset as you tweak the knob.  Then it would settle out once the capacitor charged to steady-state.  It would make some scratching and thumping noises as you turn the pot.

[Minion]

Why not make the opamp non-inverting so that the input impedance doesn"t affect the gain as in inverted config ... or use an opamp or transistor for each input and sum the outputs ...

[soggybag]

Would this work as a non-inverting mixer?

[frequencycentral]

If I were you I would ditch your design and do R.G.'s mixer posted above. It's non-inverting and looks how a mixer should look re the pots configuration.

[soggybag]

I would use the R.G. mixer, but I don't think it will work in my case. Since the signals I want to mix are fed to other parts of the circuit have a pot to ground as the mixer would attenuate the signal. I guess I need a mixer with a high impedance input.

[soggybag]

Here's a block diagram of what I have in mind. The buffer on the left feeds the signal to two other stages and the mixer. Maybe I need to add another buffer between the mixer and each block. I was hoping I could make this work with one dual op-amp, half for the input buffer and half for the mixer on the output.

[CynicalMan]

Here's a block diagram of what I have in mind. The buffer on the left feeds the signal to two other stages and the mixer. Maybe I need to add another buffer between the mixer and each block. I was hoping I could make this work with one dual op-amp, half for the input buffer and half for the mixer on the output.

Using pots in series with the opamp input won't do anything. You need pots wired like volume controls, like in R.G's mixer.

[soggybag]

If I wire to the pots as in R.G.s example, the signal feeding into the x2 and /2 sections will be attenuated.

The pots I have them should control the current. Is it possible to create a current mixer?

Or do I need to create a high impedance buffer for each signal before it goes into the mixer?

[Minion]

If I wire to the pots as in R.G.s example, the signal feeding into the x2 and /2 sections will be attenuated.

The pots I have them should control the current. Is it possible to create a current mixer?

Or do I need to create a high impedance buffer for each signal before it goes into the mixer?

Haveing the pots the way you have them wired will do practicly nothing when you turn them , all they do is add series resistance and not a lot because all the pots are in paralell with each other, ...

Maybe this is more what you need...

http://sound.westhost.com/project94a.htm

[Projectile]

Using pots in series with the opamp input won't do anything. You need pots wired like volume controls, like in R.G's mixer.

Haveing the pots the way you have them wired will do practicly nothing when you turn them , all they do is add series resistance and not a lot because all the pots are in paralell with each other, ...

Umm... guys, are you sure you've got that right? I agree that RG's design is the way to go, but I'm pretty sure that his idea will sort of work like a mixer... sort of. At least I'm pretty sure that it won't do "practically nothing" as you say. I don't think it would be just adding series resistance. The individual inputs are coming from separate sources, likely with low impedances, so each input would probably load the others down enough that turning a pot would probably attenuate the signal quite a bit, although having multiple pots maxed (no impedance) at the same time would cause problems, and it would not behave like a normal mixer.  It's not the most elegant solution, and it would be a strange balancing act to operate, but I think it might actually work. I'm no expert though.

[Projectile]

If I wire to the pots as in R.G.s example, the signal feeding into the x2 and /2 sections will be attenuated.

If you are worried about too much attenuation from the 10K pots loading down the inputs, then just increase the value of the pots to 100K. Unless you are feeding the mixer with really high impedance sources, there shouldn't be any noticeable loading from 100K pots.  Honestly, RG's mixer is  probably going to be a lot better solution than what you are proposing. What are you trying to mix anyway?

[juse]

Hey soggybag -

I think you're on the right track with your original pic. Just ground your pots, add some fixed series resistance coming off of them & a resistor/cap coming out:

[Projectile]

Hey soggybag -

I think you're on the right track with your original pic. Just ground your pots, add some fixed series resistance coming off of them & a resistor/cap coming out:

How is this any different from what everyone else has been suggesting?

[juse]

Affirmation, Joe - hard copy verification.. Never said it was different.   

Carry on.........

[MohiZ]

The individual inputs are coming from separate sources, likely with low impedances, so each input would probably load the others down enough that turning a pot would probably attenuate the signal quite a bit, although having multiple pots maxed (no impedance) at the same time would cause problems, and it would not behave like a normal mixer.

But you have to remember that the op-amp input is very high impedance, so there's almost zero current flowing through the pots.

How is this any different from what everyone else has been suggesting?

I think the problem is that the pots need to act as voltage dividers. This can be done by having a resistor from the output of the op-amp to the same input that the mixing pots are connected.

Check this out:
http://upload.wikimedia.org/wikipedia/commons/3/3e/Op-Amp_Summing_Amplifier.svg

R1...Rn form a voltage divider with Rf setting the gain of each input. Of course, this is inverting.

EDIT: Pardon me, I may have missed the point here ... don't mind me There are so many different schematics here I'm not sure which you are talking about in different posts.

[Projectile]

The individual inputs are coming from separate sources, likely with low impedances, so each input would probably load the others down enough that turning a pot would probably attenuate the signal quite a bit, although having multiple pots maxed (no impedance) at the same time would cause problems, and it would not behave like a normal mixer.

But you have to remember that the op-amp input is very high impedance, so there's almost zero current flowing through the pots.

The inputs will still load each other significantly. I'm pretty sure it will sort of work like a mixer.  I'm not saying it's a good design, or that he should use it, just that you will be able to adjust the levels relative to each other.

How is this any different from what everyone else has been suggesting?

I think the problem is that the pots need to act as voltage dividers. This can be done by having a resistor from the output of the op-amp to the same input that the mixing pots are connected.


Check this out:
http://upload.wikimedia.org/wikipedia/commons/3/3e/Op-Amp_Summing_Amplifier.svg

R1...Rn form a voltage divider with Rf setting the gain of each input. Of course, this is inverting.

EDIT: Pardon me, I may have missed the point here ... don't mind me There are so many different schematics here I'm not sure which you are talking about in different posts.

Yes, this is getting confusing, especially since we have both inverting and non-inverting versions floating around. I think at this point we are all recommending that the OP just simply stick to RG's design, which is the same basic design that Jose also recommended, although some of the parts values are different and his version runs off a bipolar power supply.

[soggybag]

Hey soggybag -

I think you're on the right track with your original pic. Just ground your pots, add some fixed series resistance coming off of them & a resistor/cap coming out:

Thanks for all of the suggestions. This schematic looks very interesting. The cap before the pot at the input is the arrangement that I think I need. I want to feed the signal to another circuit block and to the mixer. My electronic ineptitude was making me think that this arrangement would not work.

[Projectile]

Thanks for all of the suggestions. This schematic looks very interesting. The cap before the pot at the input is the arrangement that I think I need. I want to feed the signal to another circuit block and to the mixer. My electronic ineptitude was making me think that this arrangement would not work.

It should not matter in most cases whether the cap is before the pot, like above, or after the pot, like in RG's mixer.  It's just a blocking cap. Note, that if you do choose to put the cap before the pot, the pot MUST have a value of at least about 100K or higher, or the cap will act as a highpass filter on the input. Also, beware that this circuit is designed to work on +18v DC, and those resistor values also seem way to high for dealing with stompbox signals. I'd say stick with RG's design which will work for what you are doing, and just increase the values of the pots as needed.