This Neve simple example is part of an active circuit. Greetings.
Pan circuit for passive mixer
- Thread starter Fablab
- Start date
Help Support GroupDIY Audio Forum:
ruffrecords
Well-known member
That circuit was commonly found in Neve mixers using passive summing in the 70s. I know, I was there.
Cheers
Ian
That was easy....
JR
JR
Exactly. You can put gain before it, or gain after it, or half and half, but gain is gain and it raises the noise by the exactly the amount of gain...
A passive mixer without help from an active stage is useless.
k brown
Well-known member
Doesn't it just depend on how much gain can can be made up by the device that follows it?
What's this passive mixer going to be connected to?
And this "device that follows it", does it not constitute an active element?
Ó Draighnáin
Well-known member
As Abbey just posted, I would presume that the gain makeup device that follows (e.g. mic pre) would then be considered the active stage for the passive bus. There are bus impedance considerations to take into account, depending on the number of tracks, and if using a mic pre you might want to slug the bus (to about 150 - 200ohm). This slugging increases the amount of makeup gain required. However, as Ian pointed out in another thread, if the track count is relatively low, and depending on the bus feed resistor value, you might not need to use a mic pre.
https://groupdiy.com/threads/how-to-calculate-unbalanced-summing-bus.65160/
This is a very old topic for me, that I have discussed here and about over the years. Indeed a typical passive summer with make-up gain and a virtual earth summing amps will have similar noise floor performance.
Back in the 80s I developed a combining strategy using active current sources that avoided the noise gain (N+1) of virtual earth summers. My practical current sources (based on TL07x op amps) exhibited 10x the source impedance of resistance summers so reduced that noise gain roughly 20+ dB. Today using modern op amps and spending a little more money I could improve upon that, but so what? Digital combining is effectively perfect, and high quality analog summing can already be quieter than mic preamp or ambient room noise.
Back during the peak of the "Passive summing" fashion trend I suggested to a friend still making and selling consoles, to make a DC coupled mic pre amp input to sell to those passive summer fans.
JR
I believe the point here is not so much the noise performance than the (wrong) idea that passive mixing is distortion-free compared to VE.
That's a common selling-point.
That's a common selling-point.
ruffrecords
Well-known member
I've noticed you mention this previously. And I'd still say there's a place for the best quality analogue (yes , that's the correct spelling
) summing. Not disagreeing about the accuracy of current day digital summing - but in some cases it's overkill to go into digital just to sum then DAC back to analogue.True current sources always seemed a good idea to myself even before I ever implemented one eg "Improved Howard".
I think I've seen the detail of your work there previously.
But in UK focus was on balanced VE summing IIRC.
Indeed! I forgot that...That and its phenomenal headroom of course.
Cheers
Ian
Matt Syson
Well-known member
quote JR {Back during the peak of the "Passive summing" fashion trend I suggested to a friend still making and selling consoles, to make a DC coupled mic pre amp input to sell to those passive summer fans.}
You wicked boy!
This is the sort of comment I would come up with. Yes as with the others saying it Virtual earth summing has many advantages and the noise level will (or can be similar to 'Voltage summing. VE also 'solves' the issue of image shift which occurs when other channels (or their pan pots) are moved even with no audio contribution from those channels. With people complaining about fractional dB shifts in level (because they can now see them easily on the DAW too many are conveniently brushing over the reasons why the better desk manufacturers had lots of 'buffer' amplifiers all over the place, To prevent interaction between controls, to get a more predictable control law, reduce crosstalk and the list goes on. Even the headroom question is a bit more complicated because even if you have to consider what signal sources are contributing to the mix and their own (practical) headroom. What kind of weird music has 24 or more tracks of coherent signal that might need very high headroom?
You wicked boy!
This is the sort of comment I would come up with. Yes as with the others saying it Virtual earth summing has many advantages and the noise level will (or can be similar to 'Voltage summing. VE also 'solves' the issue of image shift which occurs when other channels (or their pan pots) are moved even with no audio contribution from those channels. With people complaining about fractional dB shifts in level (because they can now see them easily on the DAW too many are conveniently brushing over the reasons why the better desk manufacturers had lots of 'buffer' amplifiers all over the place, To prevent interaction between controls, to get a more predictable control law, reduce crosstalk and the list goes on. Even the headroom question is a bit more complicated because even if you have to consider what signal sources are contributing to the mix and their own (practical) headroom. What kind of weird music has 24 or more tracks of coherent signal that might need very high headroom?
"Howland" current pump.
It's kind of unfortunate that the same name applies for a supplier of plumbing supplies (Howland pump & supplies).
ruffrecords
Well-known member
It is important to rememeber that, in the worst case, everything you hang on the fader output does two things you don't really want it to do. First it changes the law of the fader and secondly it reduces the input impedance. Ignoring the pan circuit for the moment, if you have a 10K fader with a 10K aux send pot hanging off it which in turn has a 10K bus feed resistor then the worst case input impedance drops from 10K to 3K3. Adding the pan into the equation only lowers it further. The bottom line is passive mixing works best when fed from a known low source impedance. So add a unity gain buffer before this circuit and this problem goes away. Then all you have to worry about is how the fader law is affected by the loading of the pan and aux sends and the way to do that is to make sure the pots used are much larger than 10K which is why I use 50K pots for pan and AUX sends with 47K bus feed resistors.
Cheers
Ian
I have a console that has had some mods done to the panning matrix previously. Basically what was the front/back pan is split off to drive other panning functions. We had to add a buffer before the matrix because the matrix was now loading down the previous stage too much & it just idn't work well.
