[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/diy/ - Do It Yourself

View post   
File: 3 KB, 640x300, audmix-f2.gif [View same] [iqdb] [saucenao] [google]
542735 No.542735 [Reply] [Original]

I'm trying to mix a mono signal with a separate stereo signal without losing the stereo separation. How do I do so electrically? I found pic related, but in order to get mono in both channels, I would have to bridge them, which would ruin the stereo separation I believe.

>> No.542740

use an active buffer (no gain amp) for the mono input and then mix the output with the stereo inputs (with a passive resistor-based mix, you'll get crosstalk). probably best to buffer all the inputs.

>> No.542742
File: 17 KB, 500x750, 500px-NPN_emitter_follower.svg.png [View same] [iqdb] [saucenao] [google]
542742

>>542740
From what I [think] I know about opamps is that they require feedback from the output to be streamed back into the input for unity gain, which would end up bridging the two mono outputs back to the single input.

Or are you talking of a different type of buffer like this unity gain voltage buffer?

>> No.542743
File: 8 KB, 404x175, 1382646508807.jpg [View same] [iqdb] [saucenao] [google]
542743

>>542742
talking about this kind of unity buffer configuration. that would probably work also.

>> No.542744
File: 71 KB, 740x483, 1382646723811.jpg [View same] [iqdb] [saucenao] [google]
542744

example

>> No.542747

>>542743
I was thinking the circuit looked a little differently and got confused.

Does that adequately isolate the inputs from the outputs?

>>542744
Would omitting most of that circuit and going with >>542735 except with unity gain buffers (NJM4556) on all 4 inputs work fine as well?

>> No.542750

>>542747
yes and yes.

>> No.542751

>>542750
Thank you for your time.

>> No.542757
File: 5 KB, 561x404, stereo + mono mixer.gif [View same] [iqdb] [saucenao] [google]
542757

this will work

>> No.542762
File: 14 KB, 1839x477, Mixer v01 sch render.png [View same] [iqdb] [saucenao] [google]
542762

This is what I have so far.

>> No.542764

>>542757

I entered this thread to suggest this. Buffer the inputs if you really want to. Google "summing opamp".

>> No.542765

>>542764
>>542757
What resistors should I use? It seems like the stereo channel would get quite a bit of crosstalk with that design, which is why I started the thread in the first place.

>> No.542766

>>542762
yeah, that should do it. siince its buffered, you could use pots (10k-20k) for the fixed resistors so you can change the volume.

>> No.542768

>>542766
I already plan to have an attenuation pot coming from the mono output (which happens to be a mic output stage, pic related) so I think it's all set. This is for mixing live audio from the inputs with input from a DAC into a headphone amplifier (gutted O2 board)

on that topic, is there any harm in pulling power from the voltage rails of the O2 (after the mosfets and power section >>542211) to supply the mic preamps? It shouldn't be more than a few tens of milliamps, and doing so would save me from having to make a dedicated supply (see other thread) with exactly the same circuit as the O2, minus the startup protection circuit and running at +/-12v instead of +/-15v.

>> No.542769

>>542768
pic >>542218

>> No.542770

> It seems like the stereo channel would get quite a bit of crosstalk with that design

no, it wouldnt. coz the + input to the op-amp in >>542757 is at ground level, by definition. the signals cant generate cross-talk, if they're both at ground potential when they meet.

>> No.542771

> What resistors should I use?

10K for everything if you want to just sum the signals. or change the feedback ones to 100K to get a 10x gain.

>> No.542772

> coz the + input to the op-amp

sorrt, i meant the - input. in fact, they're both at ground, by definition of op-amp action.

>> No.542775

>>542768
should be fine.

btw that section you have in red, thats to keep the rail voltages balanced when either rail is loaded. it might lead to distortion during loud signals (chopping of the top/bottom of the signals).

>> No.542776

>>542775
"changing/omitting it might lead to distortion during loud signals (chopping of the top/bottom of the signals)."

oops

>> No.542778
File: 4 KB, 640x300, active mixing circuit.gif [View same] [iqdb] [saucenao] [google]
542778

>>542775
>>542776
Sounds Good. That'll save me about $10 in components and $13 in boards, plus make the circuit smaller to boot!

>>542770
So pretty much this circuit?

>> No.542786

> So pretty much this circuit?

no. of course not. you need two such circuits, with 2 inputs each. if you're gonna mix L + R + MONO, as it appears, then how the hell do you retain stereo separation? that's nuts.

>> No.542792

>>542786
test

>> No.542856

>>542740
>active buffer (no gain amp)
Actually a op amp used as a buffer has a gain of 1.