Makeup Gain Stage for Passive Summing Mixer Using ETI 430 Line Amplifier

[Winston OBoogie]

  I am not quite sure exactly what you mean by force the bus low to ameliorate gain changes but they did not slug the bus to do this. 

For the most part this is so.  But I've worked on quite a few originally built for broadcast small desks where the input to the bus amp was slugged and forced low for this exact purpose.  Of course, lots/most (all?) of the bigger recording desks with comprehensive and buffered router modules didn't and, as you say, it was just brought down to the 300 ohm area for the 10468 xnsformer.
 
Perhaps I should have limited my example to Helios since Dick Swettenham wrote about this technique of forcing the bus low and therefore, it's a more well known example. 

P.S.  Just in case it wasn't obvious -  I'm not advocating using hybrid coils, or parallel plate mixers, or forcing your bus low - I was just chucking in some filler material on the thread.

 

 

[ruffrecords]

Perhaps I should have limited my example to Helios since Dick Swettenham wrote about this technique of forcing the bus low and therefore, it's a more well known example. 

By any chance do you have a copy of or link to Dick's text on this subject?

Cheers

Ian

 

[abbey road d enfer]

For the most part this is so.  But I've worked on quite a few originally built for broadcast small desks where the input to the bus amp was slugged and forced low for this exact purpose. 

I would think the main reason at one moment in time (advent of stereo?) was to limit feedthrough. One has to remember that the output impedance of most active stages was about a few hundred ohms, as opposed to the present near-zero output Z of SS circuits.

 

[abbey road d enfer]

I believe it was Steve who pointed out some decades ago that, all other things being equal, there is a slight noise advantage for voltage summing when channels summed is less than 10. 

I'd be too interested in seeing this, as I'm too lazy to do it myself, but I believe noise is not the only determining factor in summing amps, particularly in broadcast, where feedthrough and EMI/RFI are a big concern.
Anyway, the point is moot, since noise of a 10-stem summing amp is negligible, unless you want to butcher it. 

 

[cpsmusic]

Putting a 10k pot in place of R8 (rheostat wiring), will give you roughly 20dB to 40dB range.

JR

Any way to get the full range from 0dB to 40dB?

 

[JohnRoberts]

Any way to get the full range from 0dB to 40dB?

Of course but for that topology it isn't as simple..the other emitter resistors are important for establishing the DC operating points, so varying them will alter DC bias.

Another possibility is to replace R9 with a 100k pot (wired as a 0-100k rheostat). But you would need to confirm that this does not impact stability.

JR

PS: My original advice stands, there are lots of inexpensive off the shelf mic preamp ICs.

 

[Winston OBoogie]

Gentlemen (Ian & Abbey)
I don't have scanned copies to post here but:

The text by Dick Swettenham is the "Mixing Consoles" chapter of 'Sound Recording Practice' edited by John Borwick. (Dick also wrote another chapter in the book on 'Mobile Control Rooms').
There have been a few editions of this book but I'm not sure the later ones will include the info we're discussing here, it may, but to be safe -  The one to search for is from 1980. 

On Steve Dove's rundown of the noise of voltage and current summing - there is a large chapter written by him in 'Handbook for Sound Engineers, The New Audio Cyclopedia', (edited) by Glen Ballou.
My copy was from around 1987, I'm not sure if there are later editions of this or not?

 

[JohnRoberts]

Gentlemen (Ian & Abbey)
I don't have scanned copies to post here but:

The text by Dick Swettenham is the "Mixing Consoles" chapter of 'Sound Recording Practice' edited by John Borwick. (Dick also wrote another chapter in the book on 'Mobile Control Rooms').
There have been a few editions of this book but I'm not sure the later ones will include the info we're discussing here, it may, but to be safe -  The one to search for is from 1980. 

On Steve Dove's rundown of the noise of voltage and current summing - there is a large chapter written by him in 'Handbook for Sound Engineers, The New Audio Cyclopedia', (edited) by Glen Ballou.
My copy was from around 1987, I'm not sure if there are later editions of this or not?

Here are the Steve Dove console design articles, perhaps this can refresh some memories.

http://www.waynekirkwood.com/images/pdf/Dove_Steve_Console_Design.pdf

JR

PS: The acknowledgements on page 72 was where I got mentioned (among some pretty good company IMO).

 

[abbey road d enfer]

On Steve Dove's rundown of the noise of voltage and current summing - there is a large chapter written by him in 'Handbook for Sound Engineers, The New Audio Cyclopedia', (edited) by Glen Ballou.

Indeed ther's a brief chapter about this at page 96 of the link JR submitted, but I must say, with all due respect, it does not represent good practice in audio design, using resistors so high the noise current becomes dominant.
So, this case cannot be generalized, and I maintain that the less channels the less noise, within the limits of reason.

 

[Winston OBoogie]

Indeed ther's a brief chapter about this at page 96 of the link JR submitted, but I must say, with all due respect, it does not represent good practice in audio design, using resistors so high the noise current becomes dominant.
So, this case cannot be generalized, and I maintain that the less channels the less noise, within the limits of reason.

Those articles pre date the book chapter I mentioned and are not one and the same.
However, I take your point about the 'bad practice' example in the older article and agree.

Further still, if we are mixing fewer channels than the cutoff of 10 stated by Dove, it's not exactly difficult to achieve <20dB of quiet gain. 








 
 

 

[JohnRoberts]

I wold avoid caution about embracing too simple rules of thumb...

I have designed mixers with as few as 6 input channels and with something like 116 stems feeding the L/R mix bus... As I recall there were rather different design constraints to consider. 

Odd that this topic can be so contentious, maybe its just me... :

JR

 

[abbey road d enfer]

Odd that this topic can be so contentious, maybe its just me... :

Blame the mktg guys who claim magic from a bunch of resistors and connectors in a tin box.

 

[ruffrecords]

Blame the mktg guys who claim magic from a bunch of resistors and connectors in a tin box.

One thing passive mixing does have is incredible headroom. You can turn up umpteen stems as high as you like and the bus will not overload - you just turn down the gain of the external gain make up amp. Not something you can do with a VE mixer (actually you can but nobody does as far as I know).

Cheers

Ian

 

[user 37518]

One thing passive mixing does have is incredible headroom. You can turn up umpteen stems as high as you like and the bus will not overload - you just turn down the gain of the external gain make up amp. Not something you can do with a VE mixer (actually you can but nobody does as far as I know).

Cheers

Ian

Dear Ian, does any of your mixer designs use passive mixers?

 

[cpsmusic]

Here are the Steve Dove console design articles, perhaps this can refresh some memories.

http://www.waynekirkwood.com/images/pdf/Dove_Steve_Console_Design.pdf

JR

PS: The acknowledgements on page 72 was where I got mentioned (among some pretty good company IMO).

This should definitely go into the Technical Documents section.

 

[abbey road d enfer]

One thing passive mixing does have is incredible headroom. You can turn up umpteen stems as high as you like and the bus will not overload - you just turn down the gain of the external gain make up amp. Not something you can do with a VE mixer (actually you can but nobody does as far as I know).

I've seen a few examples where the summing stage has a variable resistor that allows accomodating almost any overload situation. That makes sense for an EE, but apparently less for a SE; I believe that's why it diddn't catch.
And many cheap mixers place the fader in the NFB loop of the summer, only not because it increases headroom protection, but because it saves parts (no post-fader buffer needed).

 

[ruffrecords]

I've seen a few examples where the summing stage has a variable resistor that allows accomodating almost any overload situation. That makes sense for an EE, but apparently less for a SE; I believe that's why it diddn't catch.
And many cheap mixers place the fader in the NFB loop of the summer, only not because it increases headroom protection, but because it saves parts (no post-fader buffer needed).

Interesting. Do they couple the fader in the feedback loop via a capacitor or do they just let the bias current flow through it? If it is ac coupled I would expect the fully off attenuation to suffer at sub audio frequencies.

Cheers

Ian

 

[JohnRoberts]

Interesting. Do they couple the fader in the feedback loop via a capacitor or do they just let the bias current flow through it? If it is ac coupled I would expect the fully off attenuation to suffer at sub audio frequencies.

Cheers

Ian

The only console where I used a several dB gain attenuation trim in the L/R master bus amp was my large AMR production series with something like 116 stems in the L/R mix. (I used a dual pot in the feedback path of L & R buses, with additional trim pots to compensate for 20% pot resistance tolerance.)

In a differential bus topology you would need to tweak both legs to preserve the differential math. (I used a hybrid mix of 72 current sources, and the rest conventional resistors feeding the buses, so my differential math was already different.)

For the record a passive mixer with make up gain goes not have infinite headroom. Normalized for unity gain per stem it will deliver pretty similar behavior, to other summing topologies including, similar noise, distortion, and phase shift (unless you intentionally use a dirty make up gain stage for extra flavor). 

Back in the 90s Mackie made a big deal in their advertising about extra headroom from operating their sum buses with -6dB loss. 

JR

 

[Winston OBoogie]

Interesting. Do they couple the fader in the feedback loop via a capacitor or do they just let the bias current flow through it? If it is ac coupled I would expect the fully off attenuation to suffer at sub audio frequencies.

Ian,
here's an example of a summing amplifier with adjustable gain.  One of our esteemed members here - Gareth Connor -  had a hand in the desk which used this scheme.

 

[ruffrecords]

Ian,
here's an example of a summing amplifier with adjustable gain.  One of our esteemed members here - Gareth Connor -  had a hand in the desk which used this scheme.

Thanks for posting that. It is going to take a while to get my head around that.

Cheers

Ian