Summing mixer

[Jbliss]

Hi
Im looking a building a Inverting Summing Amplifier like this http://masteringelectronicsdesign.com/wp-content/uploads/2010/01/3_input-summing-amplifier.png 

Pretty Simple. However just wondering if it is possible to sum 20 inputs using this design or would it be better to say make 5 of the summing circuits and then sum the outputs of them together. The signals are Line level Audio signals 
Thanks you for any help

[dmills]

What is on the back of that link is not an inverting stage, but assuming you are in fact building a standard virtual earth stage, you can indeed do 20 inputs, but the noise tends to come up because the noise gain increases with each input you add.

This may or may not be a serious problem in your application, and there is some good discussion of the subtle points in "Small Signal Audio Design" by Douglas Self.

Regards, Dan.

[Jbliss]

Hi Dan
thanks for the reply. is there a way I can combat the noise issue

Thanks 

[Jbliss]

Would it be better to do it in stages

What is on the back of that link is not an inverting stage, but assuming you are in fact building a standard virtual earth stage, you can indeed do 20 inputs, but the noise tends to come up because the noise gain increases with each input you add.

This may or may not be a serious problem in your application, and there is some good discussion of the subtle points in "Small Signal Audio Design" by Douglas Self.

Regards, Dan.

[danadak]

Short answer is depends on G thru signal path.

Take a look at page 47 -


http://www.pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Motchenbacher_Connelly/Low-noise_Electronic_Design.pdf


Regards, Dana.

[Jbliss]

Short answer is depends on G thru signal path.

Take a look at page 47 -


http://www.pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Motchenbacher_Connelly/Low-noise_Electronic_Design.pdf


Regards, Dana.

Thanks for the link Interesting.
would isolating each input reduce noise. Eg using a buffer

Thanks

[Zero999]

Hi
Im looking a building a Inverting Summing Amplifier like this http://masteringelectronicsdesign.com/wp-content/uploads/2010/01/3_input-summing-amplifier.png 

Pretty Simple. However just wondering if it is possible to sum 20 inputs using this design or would it be better to say make 5 of the summing circuits and then sum the outputs of them together. The signals are Line level Audio signals 
Thanks you for any help

What's the intended gain and bandwidth?

The op-amp sees all the input resistors in parallel so with 20 of them, negative feedback will be 1/20th of what it would be with one input. For example, if the design gain is unity, then expect to see the same, bandwidth, transient response, output impedance etc. as you would if the same op-amp were configured as an inverting amplifier with a gain of 20.

[Jbliss]

Hi
Im looking a building a Inverting Summing Amplifier like this http://masteringelectronicsdesign.com/wp-content/uploads/2010/01/3_input-summing-amplifier.png 

Pretty Simple. However just wondering if it is possible to sum 20 inputs using this design or would it be better to say make 5 of the summing circuits and then sum the outputs of them together. The signals are Line level Audio signals 
Thanks you for any help

What's the intended gain and bandwidth?

The op-amp sees all the input resistors in parallel so with 20 of them, negative feedback will be 1/20th of what it would be with one input. For example, if the design gain is unity, then expect to see the same, bandwidth, transient response, output impedance etc. as you would if the same op-amp were configured as an inverting amplifier with a gain of 20.


Its intended of monitoring vocal. would what I'm planing on doing not work for this application ?
thanks 

[Jbliss]

Would it be better to do it passively.
http://wiki.diyrecordingequipment.com/wp-content/uploads/2011/12/schematic.png
however i would lose gain due to the resistors ?
Thanks

[Zero999]

Would it be better to do it passively.
http://wiki.diyrecordingequipment.com/wp-content/uploads/2011/12/schematic.png
however i would lose gain due to the resistors ?
Thanks

The passive option would work and yes you're right, there will be some attenuation, by a factor of 20 minimum.

[Jbliss]

Ok thanks every one
what would be the best option for summing 20 Channels of audio each with there on volume pot and no attenuation due to summing resistors (use of preamp) also the least Noisy??
thanks you all very much. 

[Jbliss]

Hi
Im looking a building a Inverting Summing Amplifier like this http://masteringelectronicsdesign.com/wp-content/uploads/2010/01/3_input-summing-amplifier.png 

Pretty Simple. However just wondering if it is possible to sum 20 inputs using this design or would it be better to say make 5 of the summing circuits and then sum the outputs of them together. The signals are Line level Audio signals 
Thanks you for any help

What's the intended gain and bandwidth?

The op-amp sees all the input resistors in parallel so with 20 of them, negative feedback will be 1/20th of what it would be with one input. For example, if the design gain is unity, then expect to see the same, bandwidth, transient response, output impedance etc. as you would if the same op-amp were configured as an inverting amplifier with a gain of 20.

I assume this would this Cause distortion as well ??

[ruairi]

Hey J,

Mastering engineer here, 22 years into the game.  A question - why are you building a summing mixer?  (I know there are several possible answers, curious to hear yours specifically).

Cheers,
Ruairi

[Jbliss]

Hey J,

Mastering engineer here, 22 years into the game.  A question - why are you building a summing mixer?  (I know there are several possible answers, curious to hear yours specifically).

Cheers,
Ruairi

Well its actually for monitoring a rack of radio Mics for a music theatre show. So the quality does not have to be amazing as it not or studio work.

[Jbliss]

Hey J,

Mastering engineer here, 22 years into the game.  A question - why are you building a summing mixer?  (I know there are several possible answers, curious to hear yours specifically).

Cheers,
Ruairi

What would you recommend thanks.

[Zero999]

Why not just use an ordinary, inverting summing amplifer?

http://www.electronics-tutorials.ws/opamp/opamp_4.html

[Jbliss]

Why not just use an ordinary, inverting summing amplifer?
[img]http://www.electronics-tutorials.ws/opamp/opamp11.gif?81223b[/imh]
http://www.electronics-tutorials.ws/opamp/opamp_4.html

Hi Thanks your reply. Thats a great idea just wondering how much noise there would be.
Thanks

[Jbliss]

Why not just use an ordinary, inverting summing amplifer?
[img]http://www.electronics-tutorials.ws/opamp/opamp11.gif?81223b[/imh]
http://www.electronics-tutorials.ws/opamp/opamp_4.html

Also what is the benefit of using a Inverting summing amplifier instead of a non inverting
Thanks

[Zero999]

The amount of noise depends on the type of op-amp used and the source impedance.

What sort of microphones are they?

[Jbliss]

The amount of noise depends on the type of op-amp used and the source impedance.

What sort of microphones are they?

I would be a line level output from a RF Mic receiver.

[Hideki]

The typical way to mix audio signals (in an audio mixer) is the inverting circuit shown above, and NOT the non-inverting one from the first post. It will sum the currents at the virtual earth node, and you avoid each input interfering with the others.

Noise from the mixer itself isn't going to be a problem for your application as long as you keep the resistor values reasonably low. Less than 10k would be good.

[ciccio]

When I read the first post I though the OP wanted to buid a "standard mixer".
Now I understand that his needs are simpler.
He has 20 line level, low impedance outputs.
If he does not need individual volume controls, a passive mixer (it does not needs to be balanced) will be enough.
Using 10K resistors everywhere, it will have an attenuation of 20 times, eg - 26 dB.
I think he will need some monitoring amplifier, that can easily supply the gain needed to compensate the attenuation.
The amplifier will have a volume control, to adjust monitoring level.
If he wants to go into experimenting, all this can be built with a "chip amplifier" (LM3886 or equivalent, boards can be found on ebay) summing the channels' resistor on the volume pot.

[Zero999]

The amount of noise depends on the type of op-amp used and the source impedance.

What sort of microphones are they?

I would be a line level output from a RF Mic receiver.

I'd just use the circuit linked to in my previous post with 1k resistors all round and an NE5532 op-amp.

[FlyingHacker]

Might want to stick a small cap on each input too to block any DC.

Also perhaps a pot for each input (along with a smaller fixed resistor) to set relative levels for each source?

[Jbliss]

The typical way to mix audio signals (in an audio mixer) is the inverting circuit shown above, and NOT the non-inverting one from the first post. It will sum the currents at the virtual earth node, and you avoid each input interfering with the others.

Noise from the mixer itself isn't going to be a problem for your application as long as you keep the resistor values reasonably low. Less than 10k would be good.

Thanks everyone !!! great help all of you. So would say a 10 k Pot then in to a 1uf cap then a fixed 1k resistor??
thanks again 

[Jbliss]

The amount of noise depends on the type of op-amp used and the source impedance.

What sort of microphones are they?

I would be a line level output from a RF Mic receiver.

I'd just use the circuit linked to in my previous post with 1k resistors all round and an NE5532 op-amp.

So you would say 10k potentiometer then a 1k fixed

[Audioguru]

An inverting opamp with 1k input resistors cannot use 10k volume controls because the 1k will load down the 10k control so much that it will be difficult to adjust the level. Maybe use 20k input resistors fed from 10k volume controls.

[Jbliss]

An inverting opamp with 1k input resistors cannot use 10k volume controls because the 1k will load down the 10k control so much that it will be difficult to adjust the level. Maybe use 20k input resistors fed from 10k volume controls.

Thanks for your input will definitely implement.    So would you Have the pot first then the fixed or other way. 
thanks

[Audioguru]

So would you Have the pot first then the fixed or other way?

If the input goes to the fixed series resistor which feeds the pot to ground then when the pot is turned down the low frequency gain of the opamp will be a few hundred thousand and its output noise will be very very high. Actually the output of the opamp will be as high as it can go and it will not amplify because it will be amplifying its DC bias voltage or its input offset voltage instead.

[Jbliss]

So would you Have the pot first then the fixed or other way?

If the input goes to the fixed series resistor which feeds the pot to ground then when the pot is turned down the low frequency gain of the opamp will be a few hundred thousand and its output noise will be very very high. Actually the output of the opamp will be as high as it can go and it will not amplify because it will be amplifying its DC bias voltage or its input offset voltage instead.

so pot first then fixed. thats what most of the circuits I have seen implement
thanks 

[Jbliss]

Might want to stick a small cap on each input too to block any DC.

Also perhaps a pot for each input (along with a smaller fixed resistor) to set relative levels for each source?

so say a 10k Pot then a 20k Fixed resistor.


Thanks