DIY Passive Line Mixer?

[FALKEN]

In one of the Lab Meta links, NYDave posted a passive mixer schematic using balanced transformer inputs and then opamps for makeup gain. The major downside is that transformers are expensive.

I can use an external line amp for makeup gain. I realize transformers are expensive; but it might be easier for me to build. Are you saying that my brain fart is do-able?

 

[mshilarious]

I can use an external line amp for makeup gain. I realize transformers are expensive; but it might be easier for me to build. Are you saying that my brain fart is do-able?

http://electronicdave.myhosting.net/pdf/genpurposemxr.PDF

 

[FALKEN]

so no power necessary? I can replace the opamps at the end of this diagram with another pair of transformers (and no volume control) ? what transformers should I be looking at?

 

[frederic]

There is also probably more capacitance than necessary, but I like big capacitors

I do as well, but only if you shunt them with resistors.

 

[FALKEN]

something doesn't make sense though.

does the volume control need to be buffered on BOTH sides to be stable?

is this why all of the input-transformer based schematics I've seen also feed into an opamp?

the link posted a couple posts back....would you run into the same problem with this design of each channel's volume control affecting the mix balance?

 

[frederic]

Why op amps?

Because passive components reduce signal strength, and often by themselves color the sound.

Op amps can fix that, if tuned correctly.

That's why the vast majority of designs, cheap to good to excellent, use op amps.

 

[FALKEN]

Why op amps?

Because passive components reduce signal strength, and often by themselves color the sound.

Op amps can fix that, if tuned correctly.

That's why the vast majority of designs, cheap to good to excellent, use op amps.

I am not dissing op amps, man. I'm saying I have no skills, and I'm wondering if there is a way that's "easier" to build for a moron like myself.

I really appreciate you guys taking the time to help me out.

So, do the volume controls need to be buffered on both sides?

 

[mshilarious]

something doesn't make sense though.

does the volume control need to be buffered on BOTH sides to be stable?

is this why all of the input-transformer based schematics I've seen also feed into an opamp?

No, the opamp is probably there for gain.

the link posted a couple posts back....would you run into the same problem with this design of each channel's volume control affecting the mix balance?

That was minimized with the included resistor formula, I believe.

 

[FALKEN]

that doesn't make any sense. what's the difference if the signal is coming from an internal opamp or transformer or from an external line amp of my tape deck? the same problems must be present when you start messing with volume, without a buffer before the mix bus. if not, why????

sorry if I sound impatient. Its just that I would like to get back to actually recording something!!!! I am [---] this close to building this beast!!!

Sorry if I can't wait for frederic's design! Even if I could wait I prolly couldn't build it!

 

[mshilarious]

that doesn't make any sense. what's the difference if the signal is coming from an internal opamp or transformer or from an external line amp of my tape deck? the same problems must be present when you start messing with volume, without a buffer before the mix bus. if not, why????

Oh I understand the question. Well, in NYDave's design, it was intended to be a passive mixer. If, say in a two-channel version, if both channels are full on, with one channel centered and the other hardpanned to right, then on the left channel there will be 17K (50K * (2-1/2+1) "seen" by the centered channel, which would have output impedance of (somebody check my math) 40K or so. That would be a -10dB loss. With both centered, it's a -6dB loss. A four channel version with three hardpanned channels would drop the center channel -16dB, vs. -12dB with all centered. So yeah, there is a -4dB drop, but as you add more channels, it becomes more of an unusual situation.

If you buffer the volume and pan, then you don't have that issue, but I don't know if that is common in large-scale mixer design, probably too expensive for a small problem.

If you increase the resistor value at RX, the problem is diminished. I note that NYDave recommended larger values than Forssell did in his very similar schematic here:

http://www.forsselltech.com/8chsum_1.pdf

But I imagine the trade-off that keeps one from using 470K summing resistors to eliminate that entirely is noise.

 

[FALKEN]

But I imagine the trade-off that keeps one from using 470K summing resistors to eliminate that entirely is noise.

from what I've read so far I would have to agree with you.

So it seems that simply having input transformers does not solve this problem. Although I could solve the problem you described with a panning *switch* rather than a pot, I am still concerned about the effects of the volume pots on the mix balance. I guess this puts me back at using an opamp buffer between the volume pot and the bus?

it seems you cannot build a mixer that you can ride the gain pots on without taking this step?

 

[FALKEN]

wait a minute....


to be able to adjust volume levels without upsetting the whole balance, would you do this:

tape machine ch. out ---> input buffer --->volume pot--->mix bus

or this:

tape machine ch. out--->volume pot--->buffer--->mix bus

or this:

tape machine ch. out--->input buffer--->volume pot--->buffer--->mix bus



?

my guess is #1 "loads" the mix bus, #2 "loads" the tape deck, and #3 is what you need.

 

[mshilarious]

from what I've read so far I would have to agree with you.

So it seems that simply having input transformers does not solve this problem. Although I could solve the problem you described with a panning *switch* rather than a pot, I am still concerned about the effects of the volume pots on the mix balance. I guess this puts me back at using an opamp buffer between the volume pot and the bus?

it seems you cannot build a mixer that you can ride the gain pots on without taking this step?

No, I don't think it's as big of a problem as you imagine. The transformers are balancing the signal and keeping the input devices happy. That could also be done with an opamp. From there the circuit would look the same. Note that the fader is actually pretty well isolated from causing large volume swings since is it only a 10K pot versus larger resistors downstream. The pan pot is potentially more of a problem, but if you are doing a mix that is at all balanced L-R, it's probably not an audible difference.

I think my comment yesterday about buffering volume pots was brought on by fresh wounds from fighting with a spring reverb tank. Whatever you do, do NOT put one of those in your mixer

 

[FALKEN]

I think my comment yesterday about buffering volume pots was brought on by fresh wounds from fighting with a spring reverb tank. Whatever you do, do NOT put one of those in your mixer

eh heh..

I use outboard spring reverb (Fostex).

That is the point of all of this!!!

modular is coming back!

so.

transformer inputs is all I need to build this beast and have it be useful as a mixing console?

remember I am not summing "subgroups" out of PT. I am working with tape and the "faders" will be a'movin!

 

[frederic]

I am not dissing op amps, man. I'm saying I have no skills, and I'm wondering if there is a way that's "easier" to build for a moron like myself.

I never said you were dissing op-amps. To be honest, it really doesn't matter to me if you build your mixer with passive, actives, or dried spaghetti. Whatever floats your boat, right?

You asked why op-amps are often used, and I simply told you why - to make up for signal losses due to resistors.

Resistors, coincidentally, resist current flow.
Audio signals, are current flow.

Resistors in audio circuits resist audio signals.

 

[FALKEN]

lol.

where do you buy this chassis punch? or should I just use 1/4" drill bits for 1/4" TRS jacks?

 

[frederic]

You can use a punch or a drill bit.

For anything round, I use a drill bit just because of convienence, for things like RS422, RS232 and VGA connectors I have a set of punches.

I used to have a punch for potentiometers, it would punch the main bushing hole as well as the "prevent the pot housing from rotating" hole at the same time, but for the life of me I can't find it.

 

[mshilarious]

Greenlee makes a 15/16" chassis punch which is great for the Neutrik panel mount XLRs. Your basic TRS chassis mount I would drill.

 

[FALKEN]

having trouble finding a 4-U rack case.

I am going to use white fender guitar knobs for the volume pots, but I want to try to find some extra large ones for the busses(a la distressor) preferably that go to 11. I also am looking for those smaller black cylindrical knobs with the white line on the side, for some 5 position pan switches (which I am also having trouble finding)....

can anyone recommend some good parts suppliers? I have been perusing parts xpress but their site is hard to navigate.

oh;

I also need some wiring that has 10 leads in the 1 wire, if u guys know of anything like that.

 

[mshilarious]

having trouble finding a 4-U rack case.

Ooh, can't help you there. Mid Atlantic doesn't sell one that big.

can anyone recommend some good parts suppliers? I have been perusing parts xpress but their site is hard to navigate.

I don't think PartsX have too much in the way of knobs. Try a guitar-specific site like allparts maybe.

For component parts, digikey has nearly everything.

I also need some wiring that has 10 leads in the 1 wire, if u guys know of anything like that.

For inside the box? Use ribbon cable.