Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

12 inputs mixer

Status
Not open for further replies.

wip

Member
Hi all,

I am trying to build a "simple virtual earth active (op-amp) 12 channels mixer". The problem is the attenuation is around 20db (log(12) * 20). I don't know much about audio & electronics but using ltspice I can see that if only 1 channel is feeding the mixer, I will get a very low level output (1). But if all the channels are feeding the mixer now I have the line out level (2) that i want. Is there a simple solution to have line out level no matter how many channels are feeding the mixer? Are we talking about compressor / limiter?

(1)
12mixer.png

(2)
12mixerall.png
 
The input is 40mV peak. The LM386 power amplifier has a gain of 200 so its output is clipping like crazy. With its 9V supply and 6 ohm load its output clips at about 2.3V peak then its maximum allowed input is 2.3V/200= 11.5mV peak.
U2 has a gain of 11 so its output is 40mV x 11= 440mV peak as shown on your first graph.
U5 has a loss of 3.4k/10k= 0.34 so its output is 440mV x 0.34= 149.6mV peak.
U6 has a gain of one so its output (line out) is also 149.6mV peak but is shown as 160mV peak on your first graph.
If two inputs feed the same signal to the mixer then of course the output level is doubled.
If ten inputs feed the same signal to the mixer then of course the output level is ten times but is shown as 1.7V peak on your second graph.
A compressor/limiter has an attack time so there will be a loud blast before the level drops to what you want, then a moment of very low sound for the level to return for a single input.
What on Earth (or on another planet) produces ten IDENTICAL sounds? Not musical instruments and not people singing.
 
  • Like
Reactions: wip
1st question... why the LM386 amps. they don't do anything useful.
2nd question why are you isolating the input preamps with resistances larger than the feedback resistor in U5's circuit (hint, make r11 larger)
 
  • Like
Reactions: wip
It's for a guitar with 12 strings / 12 pickups, if only 1 string is playing the attenuation of the mixer is too large (20db), I am calculating the output stage to have a line out level (1.76V). I removed the LM386 (i wanted to maybe have x12 of them with speakers, but won't go in this direction). I use 3.4k in U5 so that if all strings are playing the level will be 1.76V. Here's the corrected simulation showing only the valid circuit (2 / 12):

12mixercorrected.png

Maybe there's no simple solution when using that many channels?
 
The simulation assumes that all 12 strings of the guitar are playing exactly the same frequency and phase which is completely WRONG.
The frequencies will all be different (their harmonics will even beat against each other) and the phases will also all be different. Then they do not simply add.
Briefly and occasionally they all might add but you might not notice.
 
I think you're best to work to a "worst case" scinareo - assuming that you are getting the maximum expected level, in phase, into all your inputs at once. That way you can be sure that your mixer amp won't clip under any real-signal condition. You're right that just a single input will give a very low output (relative to all inputs driven), but that's just the same as the difference between playing one string and all of them at once. Don't worry about it!
Also remember that the guitar gives a very wide dynamic range output anyway, depending upon how its played, so your circuit will have that to deal with in any case. You may find that you wan to add a compressor at some stage, but they can be difficult to design.
Forgive me as I can't seem to see your attachements, but if the other guys think your basic design is sound then build it and just be prepared to adjust some resistor values to get an output level that works well in practice.
 
I just heard some demos on You-Tube of 12 string guitars. Most were acoustic but one was electric. It had "sustain" which is a form of automatic volume control.
The melody played on one or two strings sounded at the same loudness as 12-string chords.
 
I made a new simulation using real soundfile from a guitar, here's the results (mp3):
**broken link removed** <- original sound (1 string)
**broken link removed** <- preamplified / mixed (1 string playing only)
**broken link removed** <- preamplifier / mixed (6 strings playing at the same time)

So I will use 7k for R11 (gain of the final stage) assuming that a maximum of 6 strings will be played at the same time.

Now my question about this circuit:
Do I need to "AC Coupling" (V4 / V2). My signal is already AC (from a pickup coil). What about the final stage do I need to do something else before sending my signal to line out (i think i need to provide a ground for the next stage)?

Thanks for all the help so far!
 
The guitar in the MP3's sounds like the high frequencies are reduced which makes the lower frequencies (and lower frequency strings) sound louder than the high frequencies (and the higher frequency strings).
 
Do I need to "AC Coupling" (V4 / V2)? My signal is already AC (from a pickup coil).
It doesn't matter since your opamps do not have a power supply.

What about the final stage do I need to do something else before sending my signal to line out (i think i need to provide a ground for the next stage)?
If the last opamp has a dual-polarity power supply then its DC output reference voltage is already at ground (0V).
 
Do individual pickups for the 12 strings have any advantage over a single 12-pole pickup?
 
Why not wire the pickups in series. They will "add up" just like the mixer?

I can see if they are wired in parallel there will be interaction. In series there should be very little current and little interaction.
 
Why not wire the pickups in series. They will "add up" just like the mixer?
That's effectively what a 12-pole pickup does. It would be a lot simpler. A single coil could be wound round 12 poles in such a way as to provide hum-bucking too. Not being a musician I can't see 12 individual pickups being advantageous in any way.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top