Reducing gain in mic input stage to match line level

[quote author="gnd"]

I see. Like you explained, mix bus would work cooler, at lower gain, meaning lower noise and distortion.

If you see schematics of my console, after master fader, there is another 3x gain. Would it make sense to make this additional 3x already at channel, with channel output post-pan gain buffer? Thus instead of gain 1.68x, it could have gain 5x. Thus sum amp would still combine with unity gain, but at higher voltage output.[/quote]

That would indeed optimize bus S/N for the specific case of 10 dB post bus gain, but that same 10dB would come out of per channel headroom, not a very good trade off. Perhaps of academic interest, I did the exact opposite strategy in a series of small SR mixers where I moved the 10dB per channel fader gain to after the sum amp. This was useful in the live SR mixing environment where by the end of the night you'd find yourself mixing hot on the channel faders without the luxury of backing up and resetting all your mix levels.

For recording the 10dB gain post sum bus is mostly there to make all the faders the same (you don't want to confuse mix engineers) :wink: . If you need any gain there you're not hitting the bus hard enough.

[quote author="gnd"]

How about summing resistors values? If I do usual summing bus (not devolved), there will be 24 channels connected to bus. Using NE5532 for channel buffers, I could go for lower summing resistor value, like 1k? But what about current then? 24 channels at high voltage output (possibly 10V with NE5532 at 5x gain) will create substantial current for sum amp. If I feed 10V on 1k sum resistors, thats 10mA from each channel, together 250mA. I would have to go discrete then on sum amp? What is max practical source/sink current for existing NE5532? Any suggestion on this?

thnx
gnd

PS: Regarding panning resistors, now they are normal carbon type. Would it be better to use metal film type? Would it matter?[/quote]

There are small noise improvements available from using reduced resistor values but in the context of other noise sources in the audio chain these will be pretty minor. To better match the R to a bus combining amp like 5534 (decompensated to like 10db or so), you consider the summing resistors all in parallel so even a 20k resistors will parallel down to less than 1k, so I wouldn't worry about it. Note: your analysis of needing to match the combined current from every input ignores that your bus would long ago have clipped, so that's a non-issue.

If you are trying to stay in dual opamp package to simplify mod perhaps try to find some newer dual that is decompensated. While a 5532 won't hurt you much running at a noise gain of 25x, a decomp part would be better.

Regarding resistor type I have never heard a difference but have seen distortion due to voltage coefficient in some cheap resistors used at high voltage (like in a power amp feedback network). If you already need to replace the parts to make a value change, you might as well use metal films which seem to be well liked, as your labor is worth more the the few pennies for resistor types (and you can then brag that you used the finest metal film resistors.) :grin:

JR

I'm redrawing channel schematics, to include your suggested changes on pan resistors and post pan buffers.

But I'm wondering about that post channel-fader gain stage with 10dB gain. (schematics http://84.255.203.119/smsm16-2schem.pdf , page 2)
It seems to be there, to make up for gain loss that will happen on pan pots. And clipping LED takes refference from that point too, besides pre-fader point.

But this gain stage seems a weak point of channel. Probably clipping will happen on this stage (unless channel faders are pulled down below -10dB), because it is 10dB above all other levels. Would it be good to do something with this gain stage? Maybe changing it to unity gain, and then recover gain with pan pot buffers?

...
Another thing is that I tried to do balanced outputs for channels with inverter diff outputs. But first stage of inverter driver is inverting, and needs resistor, which destroys 3dB pan law. Those pan buffers need to be non-inverting, connecting pan directly to + of buffer opamp, is it?
Like this, if I add balanced drivers after pan buffer, I get 6 additional opams per channel (or 3 dual)
1. Is there a way to make balanced channel outputs, which would be at the same time pan buffers, and thus go by with 4 opamps?
2. Or maybe it would be best to do balanced bus in a way I planned before, using devolved method, submixing channels in groups of four, splitting 24 channels into 6 groups with balanced outputs?
3. Or maybe best just forget fully balanced bus, and rather go for quasi balanced?
What would you suggest?

thnx
gnd

[quote author="gnd"]I'm redrawing channel schematics, to include your suggested changes on pan resistors and post pan buffers.

But I'm wondering about that post channel-fader gain stage with 10dB gain. (schematics http://84.255.203.119/smsm16-2schem.pdf , page 2)
It seems to be there, to make up for gain loss that will happen on pan pots. And clipping LED takes refference from that point too, besides pre-fader point.

But this gain stage seems a weak point of channel. Probably clipping will happen on this stage (unless channel faders are pulled down below -10dB), because it is 10dB above all other levels. Would it be good to do something with this gain stage? Maybe changing it to unity gain, and then recover gain with pan pot buffers?

[/quote]

No the 10db gain at the channel fader is for convenience while mixing and doesn't make significant noise contribution or cause headroom issues. Clipping here before earlier signal path only occurs for gains above unity as is true for any gain stage.

I mentioned the alternate gain structure I used for a series of SR mixers where I moved this 10dB of gain to after the sum bus for information but not as a suggestion for this application. While mixing for live SR, you don't have time to readjust all of your other faders down when one channel is clipping and you still want to push it hotter. In recording you are more interested in optimizing S/N than getting a speedy mix, so you prefer to take the time to rebalance all the channels.

[quote author="gnd"]
Another thing is that I tried to do balanced outputs for channels with inverter diff outputs. But first stage of inverter driver is inverting, and needs resistor, which destroys 3dB pan law. Those pan buffers need to be non-inverting, connecting pan directly to + of buffer opamp, is it?
Like this, if I add balanced drivers after pan buffer, I get 6 additional opams per channel (or 3 dual)
1. Is there a way to make balanced channel outputs, which would be at the same time pan buffers, and thus go by with 4 opamps?
2. Or maybe it would be best to do balanced bus in a way I planned before, using devolved method, submixing channels in groups of four, splitting 24 channels into 6 groups with balanced outputs?
3. Or maybe best just forget fully balanced bus, and rather go for quasi balanced?
What would you suggest?

thnx
gnd[/quote]

I'm don't see the merit in using balanced signal transmission for such a short distance in a protected environment. Even differential bus topologies use unbalanced impedances. Using low ohms value resistors to combine/reference the grounds forward keeps the noise down while still delivering an accurate sum of the ground potentials.

It looks like you might use 0A line in the ribbon cable which looks like a bus ground to carry a differential bus ground reference but I'm not sure it's worth the trouble.

JR