Prototype Passive Mixer

[dintymoore]

This is a continuation of this thread:
https://homerecording.com/bbs/showthread.php?t=294802

I put this 8 in 2 out out passive sum box together, what makes it a prototype is a) I imagine it will be tweaked a lot and b) in order to get it done, I used stuff on hand, from Radio Shack and I didn't want to be concerned about cosmetics. It's butt-ugly!!!

The box was a God-awful ADC graphic eq (puke-barf-hurl) and it's in terrible shape. I put some foil tape on the front. Actually, painted and with a new aluminum front it could be ok. I like the size.
The back is brutal:

All the guts parts came from R.S., I know the resistors are crap. The pots aren't too horrible and the jacks are so-so. Everything there is airport - candy priced.

It got pretty busy real fast. I would make it differently if I was to do it again.

Eventually I'd like to make it, my tube amp, my speaker selector and my passive volume box all into one console, and that's when I'd buy some koa.

It's begging for something in that space on the right on the front... maybe a '63 VW speedometer or something...
Here's the schematic, I call it iGor2000:

here's how it's set up:

ch 1 and 2 come from my Summit tube eq and those are my important lead vox and lead solo channels

ch 3 and 4 are separate channels for stuff like bass or another mono track. I put kill switches (S1 & S2) on the reverb sends on those channels

ch 5 and 6 are a stereo pair and the panning is set on the computer. This is where most all the tracks are

ch 7 and 8 are from my Roland SC-8850, which is the synth module I use for pretty much everything.

It's got a stereo effects send and receive so that I can use my SPX1000, which is the best reverb I have at the moment. It's a dinosaur but a good one. I imagine I will replace it sometime.

What I do is put the output into ch 7&8 of my Delta 1010LT card and that's re-routed to the SPDIF out which goes into my T.C. Finalizer (that gets you 24 bit as my Finalizer is only 20bit but 24bit if you use digital in). Then the Finalizer outputs it into the Delta 1010LT's SPDIF in and that's what I record as my final LR stereo mix. It's sort of based on the system I grew into with my Otari MX5050 8 track reel to reel.

I know it's odd having 100KΩ pots, but that's all the ganged ones they had at R.S. and I wanted it to be done with one pot.

I used the 10KΩ pots for pan. It doesn't allow full panning but for what I use those channels for it's fine. I would think that higher values would make the panning wider. Now it's just the middle 25˚.

I didn't attach one contact of each effects send pot to anything. Those are the 100KΩ pots. Does that matter?

Should I be using resistors in some other places, like between some of the bus bars?

 

[mshilarious]

10K value has nothing to do with width, I think you have the panpot wrong. See this schematic:

http://electronicdave.myhosting.net/miscimages/panpot.gif

Resistor quality makes nearly no difference. I won't say "any", but it's tiny tiny tiny small. Don't worry about it.

The problem with 100K pots there is it makes the output impedance of that node higher than it needs to be, and each node feeding the network has highly variable impedance depending on the pot setting. You can get dual pots at Mouser for about $3 each, you need them for the panpots so might as well pony up for the proper value.

 

[dintymoore]

I knew the way I did the pan pots was bonkers. Thanks for that schematic.

It does work tho... it only gives a narrow sweep in the middle which actually for lead vocals is fine with me.

What I wonder, because I did some tests with this, is if I had used lower resistor values, like in the schematic you posted they used 2.2k and 4.7k, if I would need substantially less makeup gain.

If I could get the makeup gain to be only 6dB total I wouldn't need a mic pre after it, I would be able to do that with my Finalizer's 6dB input boost control.

I'm not sure how much boost I end up needing now, it might be something like 16dB. It's beyond the Finalizer's boost capabilities with 10k resistors throughout.

Now I have to use my John Hardy mic pre for makeup gain, and that requires patching 'cause I use that pre for my mics, which is ok but I'd have something hardwired to eliminate patching.

I'd love to hear this mixer with a Summit tube mic pre on the end of it, that would, to me, be ideal.

 

[mshilarious]

I knew the way I did the pan pots was bonkers. Thanks for that schematic.

It does work tho... it only gives a narrow sweep in the middle which actually for lead vocals is fine with me.

What I wonder, because I did some tests with this, is if I had used lower resistor values, like in the schematic you posted they used 2.2k and 4.7k, if I would need substantially less makeup gain.

If I could get the makeup gain to be only 6dB total I wouldn't need a mic pre after it

That's impossible, an eight node network will have 18dB loss plus you will lose about 10dB to the panpots.

Right now your panpot has 0dB attenuation at hard left to left channel and 6dB at hard right to left channel (and vice versa). At the center position you lose 3.5dB to each channel. That ain't the way to make a panpot

 

[dintymoore]

That's impossible, an eight node network will have 18dB loss plus you will lose about 10dB to the panpots.

That's what we're talking about.

Right now your panpot has 0dB attenuation at hard left to left channel and 6dB at hard right to left channel (and vice versa). At the center position you lose 3.5dB to each channel. That ain't the way to make a panpot

I'll have to get some ganged pots.

Overall, this experiment has worked tho. I can see that I will end up making what I want.

It reminds me of when I was about 12, and couldn't imagine why you couldn't just put a garden hose on a snorkel and go down deep. So I tried it and was surprised that you needed to suck like crazy at 2' and at 4' you couldn't even breath in - felt like a metal band around your chest. This was before I took scuba lessons.

 

[mshilarious]

Get a wider hose! That's an analogy to a smaller resistor, but instead you just you breathing, there are 7 other people and you're all connected to each other but not the surface. So the size of the hose doesn't matter too much . . .

 

[dintymoore]

... So the size of the hose doesn't matter too much . . .

I am trying to be mature and hold back my 5th grade attempts at humor.

 

[dintymoore]

Update: new channel wiring

I was having trouble finding double ganged 10k linear pots for the pan and found another schematic. Here's what a channel is now wired like, the 10k pot is the panpot and the 100k is the FX send:

The differences from before are a new panpot wiring and I ran a ground to the pot contacts that were unconnected before.
The new panpot wiring appears to work (!) - it can be panned hard left or right and is the same volume either way. The only bizarreness is that the center pan position isn't in the middle of the pot's sweep - it's about 60% of the way instead of 50%. That doesn't really matter though because you just put the knob on in the right position.
Running the ground wire to the pot connections that were open before makes it so that the reverb will shut off fully in the full counterclockwise position.

I went through a bunch of tests and to my ears it sounds very, very pure to mix this way. Not fucked up. I want a Summit tube pre to go with it.

Q: If I had used 5k resistors and pots, would I need less makeup gain and would there be other problems?

 

[designer.zang]

hi i love you very neat stuff!
if you use 5k resistors and pots,the input impedance is a bit low.it is still works...but you'd better add an active input stage.

zang

 

[dintymoore]

hi i love you very neat stuff!
if you use 5k resistors and pots,the input impedance is a bit low.it is still works...but you'd better add an active input stage.

zang

Thanks zang.

I would hardly call my stuff "neat"! More it's the work of a guy who's skills are in other areas trying to do his best.

When you say "if you use 5k resistors and pots,the input impedance is a bit low"... what kind of problems might that bring up? Crosstalk? Increased interaction between the inputs?

 

[mshilarious]

hi i love you very neat stuff!
if you use 5k resistors and pots,the input impedance is a bit low.it is still works...but you'd better add an active input stage.

zang

What? It's a crap output that can't drive 5K.

 

[dintymoore]

The outputs going to this mixer are all +4 balanced, it's healthy.

I found some 10k ganged pots on eBay.

One thing I'd like to do is see what I could do to make the makeup gain required as low as possible. Other than that, my gut still tells me that my current panpot is hokey but it seems to work other than it's not exactly on center at 50%. So I can see myself changing over to the NewYorkDave panpot design that you posted mshilarious, when the 10k pots arrive.

Thanks for all the input!

 

[dintymoore]

Last night I used a pair of Summit TLA-100A's for makeup gain with iGor and I'm quite sure I saw Jesus.

 

[dintymoore]

enter Boris2000

Well here was the problem: when I put my tracks through iGor2000, my passive mixer, I needed something to boost the level back up. No surprise there, there's no such thing as a passive preamp! The only thing I had lying around was a Presonus Bluetube which is good for what it is but not good enough for this. When I tried my Summit TLA-100A tube compressors for makeup gain I was blown away and knew that that was what I'd been looking for for about 20 years!

The trouble is that I need the Summit comp's for recording. I like to use them after my mic pre whenever I record. I like the sound.

There's a lot of cable swapping that needs to occur to exchange the tube compressors for the BlueTube, which is what I ended up doing. In other words, I use the Summit comp's for both recording and passive mixer makeup gain, and when I'm recording I use the BlueTube as a "fill-in" makeup gain just for monitoring. I made the following box. When the switches are up I'm recording and using the BlueTube for makeup gain and the Summits for my mics. When all the switches are down the Summits are doing the makeup gain. It also switches the outputs to the correct ones on the computer ins.

It's ugly. Again a prototype and this time a temporary one until I can buy a pair of those 1/2 rack Summit compressors, but it works great. It was made out of an old 360 Systems Pro MidiBass, dated 1988 on the inside.

It swaps 8 ins and 8 outs, and lets me swap my best gear to use in two places in my audio chain without patchbay'ing it. Way faster.