Maker Pro
Maker Pro

passive matrix mixer - need help

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Hi there,

I found this project a while ago and decided to build it myself as it would perfectly serve as a monitoring console for my band. The idea is to connect each input with each output so I can decide about multiple monitoring mixes.

1) http://sound.westhost.com/project129.htm
2) http://sound.westhost.com/p129-f1.gif

In attached pictures you will find my version of it. I used the same pot(100k)/res(15k) values as in the source project:

As I tried to test it there was no signal coming at the output. After double checking the description I found this:
Although each input and output is shown directly connected, this isn't the way it's done in practice. All inputs must be buffered with an opamp that has enough output capability to drive all the pots for that row (or column, depending on how you arrange it), and the outputs need to be fed into a proper mixing stage.

What I have done is basically an passive aproach to it. I found also many similar web project that didn't require the opamp/mixing stage. Just to mention a few of them:

3) https://youngrites.wordpress.com/2012/11/24/4-channel-passive-matrix-mixer/
4) http://forum.noiseguide.com/viewtopic.php?t=13005
5) http://www.talkbass.com/threads/the-4x4-matrix-mixer-fx-junkies-you-need-to-see-this.990222/

Can I somehow make it working without building the additional stages? The input will be fed with a line-level signal comming from my main mixing console direct outputs. Other name of this device would be a pre-fader aux expander. The output of the matrix/expander will be routed to a four channel headphone mixer, so that everybody will get her own headphone monitoring mix. Hope this makes sense to you. btw my first post here :D

Cheers,
Marek
 

Attachments

  • IMG_1111.JPG
    IMG_1111.JPG
    344.1 KB · Views: 787
  • IMG_1112.JPG
    IMG_1112.JPG
    215.2 KB · Views: 1,889

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
I don't see how you could get no sound out unless something is shorted somewhere. You should get something like 1/6 of the input at the output for any channel that is turned all the way up.

Bob
 

Gryd3

Jun 25, 2014
4,098
Joined
Jun 25, 2014
Messages
4,098
So even if you turn the potentiometer ALL the way up to listen to input 1 on the Mix 1 output there is not signal? Or is the signal simply too small to hear?
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
Your wiring looks okay. There should have been something at the output, unless you connected the output to a really low impedance load, like a headphone set with 8 or 16 ohms impedance. That will kill any signal trying to come through the 15 kΩ resistor connected to the wiper arm of the potentiometer. The outputs need to connect to an amplifier/mixer with high impedance inputs, say 10 kΩ input impedance or thereabouts. The line-level signals coming from your main mixing console direct outputs should be fine driving the 100 kΩ potentiometers.
 

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
First of all thank you guys for such an extremely quick response!
Now, I was sending the output signal to my home stereo in order to test it. The unit is somehow old ...and yes, since a while it has only one working channel (therefore mono) and the working channel was hooked up to the ring of the stereo jack (not tip). Since the matrix mixer has mono socket it obviously couldn't work!!!
Sorry for all the fuss right at the begining of my carrier here ;)
Nevertheless, now, as I can confirm an uninterrupted signal flow there is one more issue:

Input 1 is fed with a signal,
Output 1 is connected to the amp,
all knobs down to 0.

When I turn up the first knob sending in_1 to out_1 I can hear the sound.
When I turn up any of the knobs in the range 2,2 - 5,5, I can hear a weak signal. every other knob adds some more signal to it.
If I do this:
v
00000
01111
01111
01111
01111
(where 1 is max up, 0 max down)
I end up having signal as strong as if Iturned the 1,1 knob ( or 0,0, depending how you preffer index an array ;) )
Logically, only turning the very first knob(1,1) should cause signal flow to the output bus 1 and therefore to the amplifier ?!
Am I missing some important point?
Greetz, Marek
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Yes you are, and I was going to point this out to you. Each potentiometer in that circuit will affect all of the other inputs because they all form one big voltage divider.

Bob
 

Gryd3

Jun 25, 2014
4,098
Joined
Jun 25, 2014
Messages
4,098
Yes you are, and I was going to point this out to you. Each potentiometer in that circuit will affect all of the other inputs because they all form one big voltage divider.

Bob
Hence why the active buffered and OpAmp solutions are preferred ;)
Hopefully the changes are subtle enough to correct for this behavior on the fly.
 

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Yes you are, and I was going to point this out to you. Each potentiometer in that circuit will affect all of the other inputs because they all form one big voltage divider.

Bob
I am affraid I am not quite familiar with the "voltage divider" term:(
Are the passive matrix mixer project examples (see my first post's links 3,4 &5) also be leeking signal between different busses, I think that is not how they supose to function. I cannot see any design difference between mine and the others apart of the pots/resistors values.
Marek
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
This is why the author said you cannot connect it up without additional amplifiers.

Any passive mixer will have this problem to a certain extent, but the values of the pots and mixing resistors you have used accentuates it. For less cross dependence you would want the pots to be lower in resistance than the mixing resistors. The problem with that is that you will overload the inputs and the output impedance will be too high. Which is why you really need an active mixer to do this right.

Bob
 

Gryd3

Jun 25, 2014
4,098
Joined
Jun 25, 2014
Messages
4,098
I am affraid I am not quite familiar with the "voltage divider" term:(
Are the passive matrix mixer project examples (see my first post's links 3,4 &5) also be leeking signal between different busses, I think that is not how they supose to function. I cannot see any design difference between mine and the others apart of the pots/resistors values.
Marek
Image taken from http://en.wikipedia.org/wiki/Voltage_divider

Z1 is the combination of the Potentiometer, and the resistor that is connected between the Input1, and Mix Ouput1. Z2 is a more complex combination of Potentiometers and resistors between MixOutput1, and the rest of the Inputs.
The circuit you have built does not 'direct' audio signals... it simply uses resistances to divide the audio into many small pieces. The biggest piece typically ends up at the Mix Output, and the smaller pieces end up feeding back through the circuit and wind up at the other inputs and outputs. These small pieces are much too small to hear or do anything with, but as you have found out, if you make all of the smaller pieces a little larger or smaller, it affects the value of the big piece.
Unfortunately, passive circuits like this do not care what direction the signal is going... that is why active filters are usually desired. They can compensate for volume loss, and prevent the signal from back feeding.
 

Attachments

  • Impedance_voltage_divider.svg.png
    Impedance_voltage_divider.svg.png
    2.2 KB · Views: 147

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Thank you Bob, thank you Gryd,

looks like I'm halfway there. I will follow the instruction from the original project.
That would be the input stage: http://sound.westhost.com/p129-f2.gif
he says the inverter stage is optional. Would it be true in my application?

and the output stage http://sound.westhost.com/p129-f3.gif
since I am using a chassis of an old console I will use the faders as the master output bus volume controller (must check their values though).
Looks like plenty of work for an amateur :D
Wish me luck!
and Hevans1944, thank you too.
all the best,
Marek
 

Gryd3

Jun 25, 2014
4,098
Joined
Jun 25, 2014
Messages
4,098
he says the inverter stage is optional. Would it be true in my application?
This is personal preference.
Without it, the output the speakers play will be 180 degrees out of phase with the input.
If the only playback device is after the mixer then there is no problem, but if you have any other playback device that bypasses the mixer you will find that the inverted signal will end up cancelling sounds out. (This will result in quieter sound, or certain sounds being quieter. Often times stage performances will intentionally face speakers toward the stage that play out of phase with the crowd facing speakers to help cancel out the volume on stage.)
 

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Often times stage performances will intentionally face speakers toward the stage that play out of phase with the crowd facing speakers to help cancel out the volume on stage
Well I can only see two applications for the mixer. Either in a rehearsal situation using headphones which eliminates the interference of the external sound sources or on stage where the signal would be routed to the on-stage monitor wedges/boxes facing the band and not the crowd. The only interference/cancelling there would be between backline guitar/bass amp and the monitoring system. If I got you right that would be desired condition. I can always introduce the inverter at any later point if there will be such a need.
Marek
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
The only reason for the inverter is to make sure all signals retain their original phase, which would be important if later mixing occurs between inverted and non-inverted signals derived from the same source, such as a single musical instrument. It will not be apparent to the ear if this subsequent mixing does not occur, but in a stage environment there will always be people who love to move plugs around, just "to see what happens". Keeping all the signals in their original phase helps prevent inadvertent cancellation of two out-of-phase signals derived from the same source.

It is unfortunate that your passive mixer board didn't perform as expected. Ideally, if the inputs are derived from line-level low impedance sources and the outputs (from the 15 kΩ resistors attached to the potentiometers) are driving high input impedance amplifiers or mixers or whatever, there should be little cross-channel interference, although the potentiometers on a given channel will interact.

An input buffer for each input channel, and a separate summing buffer for each output channel, will make it so. Good layout and construction is necessary, preferably on a PCB instead of point-to-point wiring. Make sure the power supply leads (both positive and negative) to each op-amp are by-passed as close as possible to the op-amp terminals to analog common with two capacitors: about 1 μF electrolytic in parallel with 0.1 μF film or ceramic.
 

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Make sure the power supply leads (both positive and negative) to each op-amp are by-passed as close as possible to the op-amp terminals to analog common with two capacitors: about 1 μF electrolytic in parallel with 0.1 μF film or ceramic.
wow, one sentence and so many question marks on my side ;)

I would like to stick to the the input buffer circuit listed here http://sound.westhost.com/p129-f2.gif as a reference since I have minimal skills in designing/calculating my own.
I will use a perforated board to connect the components of the circuit.


"Make sure the power supply leads (both positive and negative) to each op-amp are by-passed as close as possible to the op-amp terminals to analog common"

What is the "analog common" thing?
Would that mean the power supply circuit needs to be wired directly to the +VE and -VE terminals of the inverter stage without running through the entire board?

"with two capacitors: about 1 μF electrolytic in parallel with 0.1 μF film or ceramic."

There are two circuits present on the picture. I assume your statement reffers to the lower one (the inverter?).
Are the positions for those two caps already mentioned in the reference circuit (C3? C4? C5?)? if so, three of them already there, two with 100 nF(which is indeed 0.1 μF). Which one of them should the 1μF electrolytic and which the film/ceramic one? You are mentioning only two caps but in the circuit there are three. should I get rid of any of them?
Can multilple input buffers share the same single inverter? (Have feeling I will be banned for this question :S )

The upper circuit(input buffer) has two op-amps. Which of them should be connect with the inverter circuit?


As I mentioned before I am using an old, non-functioning Carlsbro power mixer PMX 8-2 as the 8x8 potentiometer layout perfectly suits my 8x8 matrix mixer idea (I would design a new layout using some graphical program like PS or any others). Now, my idea is to reuse the part of each channel strip that is responsible for buffering/gaining the signal. I could spot one op-amp ne5534n near the inputs as well as one rc4558p in the middle of the pcb. The output stage(near the fader) has two transistors. I can provide way more detailed pictures.
I just wanted to know if this idea would work at all? Unfortunately I wasn't able to find any schematics in the internet :(
Marek
 

Attachments

  • IMG_1114.JPG
    IMG_1114.JPG
    136.5 KB · Views: 257
  • IMG_1115.JPG
    IMG_1115.JPG
    353.9 KB · Views: 284
  • IMG_1116.JPG
    IMG_1116.JPG
    306 KB · Views: 259

mooras

Mar 11, 2015
7
Joined
Mar 11, 2015
Messages
7
Anyone there to give me a helping hand?
Would really appreciate it.
Marek
 
Top