SSL4000-6000 "vs" Neve (V) "vs" Harrison "vs" Mci "vs" Amek L-R Master Summ bus

[r2d2]

Hi to everybody
spent a couple of nights looking "here-there-left-right-up-down".....

for documentation, schematics ,info , diy projects  etc...  about L-R master summ bus
of the mixing consoles SSL4000-6000  ,  Neve (V) , Harrison , Mci 500-600,  Amek Mozart

apart the "Apish " L-R master summ bus project by "magic" Jeff
unfortunately no other found

and then ,
any info , doc, link etc... posted is welcome
thanks in advance

r.

 

[12afael]

what is the reason of your research? you could provably find the schematic of every console on your list but I suppose you are searching something more?

 

[Gareth Connor]

Chapter 22 in the second edition of Douglas Self's Small Signal Audio Design provides much insight into summing - both the theory and the practice. Some circuits he discusses are ready for a simple copy & paste, whilst the full implementation and best performance of others requires a level of existing knowledge to fill in the subtle details that make the end result stand out.

 

[r2d2]

you could provably find the schematic of every console on your list

where ?
thanks
r

 

[abbey road d enfer]

Hi to everybody
spent a couple of nights looking "here-there-left-right-up-down".....

for documentation, schematics ,info , diy projects  etc...  about L-R master summ bus
of the mixing consoles SSL4000-6000  ,  Neve (V) , Harrison , Mci 500-600,  Amek Mozart

apart the "Apish " L-R master summ bus project by "magic" Jeff
unfortunately no other found

and then ,
any info , doc, link etc... posted is welcome
thanks in advance

r.

You must tryu harder... Proper google search will give you most of the schemos.
You will see that designers concentrate mainly on the noise ans bandwidth aspects of Virtual Earth mixing.
"Sound" is not built in the summing amps. Designers have enough on their plate with the basic performance aspects: noise, BW, distortion.
But the schemo is only part of the story; implementation is paramount, especially in big desks. Longitudinal noise is a major concern; it's the noise resulting from the channels being further and further away from the summing amp and their reference ("ground") being more and more polluted by the various circulating currents.
I have never heard a serious mixer designer mention the "sound" of summing amps; sound is built (or not) somewhere else.
The subject is vast enough to be the subject of many books, but I think it would be like writing a book about the history and development of the oil lamp...

 

[abbey road d enfer]

you could provably find the schematic of every console on your list

where ?
thanks
r

For example, the Amek 9098i is here on this website in our Technical Documents
http://groupdiy.com/index.php?topic=44846.0

 

[JohnRoberts]

Hi to everybody
spent a couple of nights looking "here-there-left-right-up-down".....

for documentation, schematics ,info , diy projects  etc...  about L-R master summ bus
of the mixing consoles SSL4000-6000  ,  Neve (V) , Harrison , Mci 500-600,  Amek Mozart

apart the "Apish " L-R master summ bus project by "magic" Jeff
unfortunately no other found

and then ,
any info , doc, link etc... posted is welcome
thanks in advance

r.

You must tryu harder... Proper google search will give you most of the schemos.
You will see that designers concentrate mainly on the noise ans bandwidth aspects of Virtual Earth mixing.
"Sound" is not built in the summing amps. Designers have enough on their plate with the basic performance aspects: noise, BW, distortion.
But the schemo is only part of the story; implementation is paramount, especially in big desks.

Amen... The physical execution is part of the design. Some of the specifics of this physical execution are not obvious from a simple schematic. While all grounds are connected together on paper, how they get connected together matters. 

Longitudinal noise is a major concern; it's the noise resulting from the channels being further and further away from the summing amp and their reference ("ground") being more and more polluted by the various circulating currents.

Ground is a concept not a voltage.

I have never heard a serious mixer designer mention the "sound" of summing amps; sound is built (or not) somewhere else.

Lesser executions of sum buses can have audible consequences but generally negative. A well executed sum bus will just sound like the inputs. 

The subject is vast enough to be the subject of many books, but I think it would be like writing a book about the history and development of the oil lamp...

It this stage in the twilight of big analog consoles it would be re-inventing the wheel. It is paying attention to lots of small details that matter in combination to the end result.

JR

 

[r2d2]

........... A well executed sum bus will just sound like the inputs. 

JR

hi "JR"
and many thanks for post....

i partially agree with this "concept":
........... A well executed sum bus will just sound like the inputs. 

because i think that any guy that want/need to summ out of the box (daw)
for get analog "feel" & co...
better L- R stereo image etc.....
(analog eq and compressors processing apart)

do not want the same sound as at the input

"jumping" between lot of the threads about analog mixdown/summ
here and other forums ,
the common "dream" its make the mixdown on a vintage neve console
(the model with 1073-1081 & co.....  pre-eq)

and i have serious doubts
that the sound from the master L - R bus
be the same like the input
connected to a DA converter (expensive or not...)

What your opinions about ?

thanks
r

 

[abbey road d enfer]

........... A well executed sum bus will just sound like the inputs. 

JR

hi "JR"
and many thanks for post....

i partially agree with this "concept":
........... A well executed sum bus will just sound like the inputs. 

because i think that any guy that want/need to summ out of the box (daw)
for get analog "feel" & co...
better L- R stereo image etc.....
(analog eq and compressors processing apart)

do not want the same sound as at the input

"jumping" between lot of the threads about analog mixdown/summ
here and other forums ,
the common "dream" its make the mixdown on a vintage neve console
(the model with 1073-1081 & co.....  pre-eq)

and i have serious doubts
that the sound from the master L - R bus
be the same like the input
connected to a DA converter (expensive or not...)

What your opinions about ?

thanks
r

Do you think electronic designers have a palette and choose a little bit of 2nd harmonic, a drop of punch, a touch of silk and a pinch of fairydust for their designs? No, they aim for the best performance in objective terms, frequency response, distortion, noise, transient response. Anything that comes on top is the result of happy(?) accident. But the original design brief is transparency, i.e. extended BW, no noise, no distortion, no colour, no flavour.

 

[MHanson]

........... A well executed sum bus will just sound like the inputs. 

JR

the common "dream" its make the mixdown on a vintage neve console
(the model with 1073-1081 & co.....  pre-eq)

and i have serious doubts
that the sound from the master L - R bus
be the same like the input
connected to a DA converter (expensive or not...)

I would say that the sound they are after is actually running through all of those modules ... not the Master Bus.  I bet if you injected a signal straight into the Master (bypassing the modules) it would come out very close to the same ...

If you've ever worked on a big desk ... while tracking you're listening through the monitor returns ... if you switch to a mix real quick and go through channel amps things change quickly ... same Bus ... different gain stages and channels though ...

Michael

 

[r2d2]

I would say that the sound they are after is actually running through all of those modules ... not the Master Bus.  I bet if you injected a signal straight into the Master (bypassing the modules) it would come out very close to the same ...

If you've ever worked on a big desk ... while tracking you're listening through the monitor returns ... if you switch to a mix real quick and go through channel amps things change quickly ... same Bus ... different gain stages and channels though ...

Michael

Hi "Michael"
and thanks for post

unfortunately I have not yet had the pleasure/chance  to make a mixdown
on a vintage neve console (the model with 1073-1081 modules)...may be one day....

but I recently met a sound engineer with about a decade of experience
on most big consoles in very well equipped studios  (tracking,overdubbing,mixdown...)
this guy work also for a my musician  friend that play rock &  blues ,

I was lucky enough to spend a whole afternoon with them
listening to a lot of work done by him and talking about the machines used
for make them
(a very long list.... , consoles and outboards apart ... , also microphones, gtr and bass amps ,
pedal effects , and much other more... cables included )

among the many things we talked about,
of course the fact that depending on the analog console used,
the final sound obtained varies .......... and  much also ,

ssl 4000 - 6000 do not sound as a neve v ,
ssl 9000 it's not same as duality
and more , euphonix cs3000 , amek mozart , etc...
(mackie , soundcraft , allen &heat , etc... also use same 5532-5534 tl072-74 , alps pots etc....
but are part of  another / different "universe" ...obviously...)

ok , at the side of the designer
probably is very tight considering the clean sound .....

but am I wrong or sound, clean  , "cold" , "sterile" ,, etc ...,
generated by daw there is already too much ?

peace
r.

 

[JohnRoberts]

........... A well executed sum bus will just sound like the inputs. 

JR

hi "JR"
and many thanks for post....

i partially agree with this "concept":
........... A well executed sum bus will just sound like the inputs. 

because i think that any guy that want/need to summ out of the box (daw)
for get analog "feel" & co...
better L- R stereo image etc.....
(analog eq and compressors processing apart)

do not want the same sound as at the input

My personal design philosophy is to make all audio paths as linear as I can, then if/when I dabble in effects or processing treat that as a separate function that can be applied when and where needed and bypassed if not wanted,

A sum bus that always colors the path, is like cooking with a dirty pot because you liked the last meal. 

"jumping" between lot of the threads about analog mixdown/summ
here and other forums ,
the common "dream" its make the mixdown on a vintage neve console
(the model with 1073-1081 & co.....  pre-eq)

and i have serious doubts
that the sound from the master L - R bus
be the same like the input
connected to a DA converter (expensive or not...)

If it isn't the same, it probably wasn't on purpose.

What your opinions about ?

thanks
r

Good luck, perhaps look into sound effects and euphonious distortions.

JR

 

[living sounds]

ssl 4000 - 6000 do not sound as a neve v ,
ssl 9000 it's not same as duality
and more , euphonix cs3000 , amek mozart , etc...
(mackie , soundcraft , allen &heat , etc... also use same 5532-5534 tl072-74 , alps pots etc....
but are part of  another / different "universe" ...obviously...)

ok , at the side of the designer
probably is very tight considering the clean sound .....

but am I wrong or sound, clean  , "cold" , "sterile" ,, etc ...,
generated by daw there is already too much ?

peace
r.

A console is a complex thing, affecting the signal at every turn and in various ways. The summing bus is just one location where it happens. An SSL 4000 is build up of 5534 op amps (including the EQs) and lot's of coupling caps in the signal path, some of them (caps) back-to-back, signal travels through VCAs for automation and dynamics, the summing itself is discrete followed by more 5534s for buffers and output etc. It absolutely matters how these op amps are used in the circuit. It even matters how they are decoupled and how (well) the PSU does. It matters how big the coupling caps in the signal path are and what type etc.
And that's just the signal path of one channel to the outputs, all the considerations about noise and hum etc. mentioned above  nonwithstanding.

The SSL 4000 "sound" is due to this entire signal path, and even more the nature of the dynamics and EQ and the use of bus compression. You don't get this by recreating the summing bus in a box .

That said, I've experimented quite a bit on my little console, even built a simple summing bus for it and tried different op amps there. There's a clearly audible difference between a 2520 (API type) op amp doing the summing and driving the following transformer and a BB2521 hybrid op amp using a 5534 (both use the same output transistors).  There's even a clearly audible difference between a current production TI NE5534 and a NOS Phillips 5534 used in that BB2521. Measurements by Douglas Self and Samuel Groner have revealed differences regarding distortion between 5532s/5534s from different manufacturers.

Another area where parts matter: There's a good thread about decoupling, where ricardo explains why using 'bad' (high ESR) decoupling caps might be important for stability in op amps. The designers in the 70s or 80s used those caps, throwing modern caps designed for high frequency applications might change the results in a negative way.

Bottom line is: Using part of some schematics in your summing box or console won't give you the "sound" of that console. Even recreating it completely 1:1 from schematics won't necessarily work, because the parts themselves might be different, let alone many circuit board design considerations, power supply etc.

It's seriously complex even without considering the actual design criteria people above mentioned, of which I cannot tell you anything.

 

[r2d2]

Hi "Living Sounds"
and thanks for post

A console is a complex thing, affecting the signal at every turn and in various ways.

Of course !!! very complex "thing"
it's not my intention to try about.....

and ok ,: "only a master L-R summ bus cannot sound as a full console,"

i'm  so interested only about master L-R summ bus

That said, I've experimented quite a bit on my little console, even built a simple summing bus for it and tried different op amps there. There's a clearly audible difference between a 2520 (API type) op amp doing the summing and driving the following transformer and a BB2521 hybrid op amp using a 5534 (both use the same output transistors).  There's even a clearly audible difference between a current production TI NE5534 and a NOS Phillips 5534 used in that BB2521. Measurements by Douglas Self and Samuel Groner have revealed differences regarding distortion between 5532s/5534s from different manufacturers.

So Fuc..... True !!
About the ic production/manufacturers "topic"
the sound engineer that work also for my musicican friend
told me about it ,
telling me that he had a direct experience ,
working on a console before and after it was "serviced/restored"
He was really very surprised ,
after the "tech treatment "  the console  sound more clean and dynamic , less noise ,
but also more "cold"............. (the same old song/story ?....)
then he asked to the technicians however
they said that the console was recapped , cleaned , re calibrated ,
also many " IC " was replaced (same type but new production , ............ obviously...),
and  where necessary pots and push buttons too ,

but that, as regards the difference in sound was because the recapping ,
the  "ICs" replacing with the  new
and theirs harmonic distortion "parameter/factor" .....

Oh my god........"progress" that generates "regression" ?.....

saying also that the raw materials with which the new electronic components are made
unfortunately cannot more of the same quality as the old ones...... (industry "speculation"...)

Oh my god......(n°2).......

...........Using part of some schematics in your summing box or console won't give you the "sound" of that console

agree with you ,
for all that and other more about wrote above

but i'm interested only in the master L-R summ bus section
all the docs about ("schemes")
info and helps ,

anything about is welcome
thanks so much  in advance
t.

 

[JohnRoberts]

ssl 4000 - 6000 do not sound as a neve v ,
ssl 9000 it's not same as duality
and more , euphonix cs3000 , amek mozart , etc...
(mackie , soundcraft , allen &heat , etc... also use same 5532-5534 tl072-74 , alps pots etc....
but are part of  another / different "universe" ...obviously...)

ok , at the side of the designer
probably is very tight considering the clean sound .....

but am I wrong or sound, clean  , "cold" , "sterile" ,, etc ...,
generated by daw there is already too much ?

peace
r.

A console is a complex thing, affecting the signal at every turn and in various ways. The summing bus is just one location where it happens. An SSL 4000 is build up of 5534 op amps (including the EQs) and lot's of coupling caps in the signal path, some of them (caps) back-to-back, signal travels through VCAs for automation and dynamics, the summing itself is discrete followed by more 5534s for buffers and output etc. It absolutely matters how these op amps are used in the circuit. It even matters how they are decoupled and how (well) the PSU does. It matters how big the coupling caps in the signal path are and what type etc.

We have well established bench measurements to qualify path performance.

And that's just the signal path of one channel to the outputs, all the considerations about noise and hum etc. mentioned above  nonwithstanding.

The SSL 4000 "sound" is due to this entire signal path, and even more the nature of the dynamics and EQ and the use of bus compression. You don't get this by recreating the summing bus in a box .

Dynamics are a huge source of colorations. First older VCAs are far from linear, but once you start modulating the gain that introduces audible artifacts. 

That said, I've experimented quite a bit on my little console, even built a simple summing bus for it and tried different op amps there. There's a clearly audible difference between a 2520 (API type) op amp doing the summing and driving the following transformer and a BB2521 hybrid op amp using a 5534 (both use the same output transistors).  There's even a clearly audible difference between a current production TI NE5534 and a NOS Phillips 5534 used in that BB2521. Measurements by Douglas Self and Samuel Groner have revealed differences regarding distortion between 5532s/5534s from different manufacturers.

I don't like to argue about what people hear but if you can correlate measurements with audible differences that is worth sharing.

Another area where parts matter: There's a good thread about decoupling, where ricardo explains why using 'bad' (high ESR) decoupling caps might be important for stability in op amps. The designers in the 70s or 80s used those caps, throwing modern caps designed for high frequency applications might change the results in a negative way.

Yup, but bench testing should reveal spurious oscillation and other stability issues.

Bottom line is: Using part of some schematics in your summing box or console won't give you the "sound" of that console. Even recreating it completely 1:1 from schematics won't necessarily work, because the parts themselves might be different, let alone many circuit board design considerations, power supply etc.

I would not expect huge differences between otherwise identical circuits, while physical layout, lead dress, ground configurations, etc. could matter for physically large circuits.

It's seriously complex even without considering the actual design criteria people above mentioned, of which I cannot tell you anything.

I do not deny that subtle differences exist, but when we quantify and put numbers to these differences they are easier to hold in perspective.

JR

PS: I am generally unenthusiastic about chasing accidental inaccuracies from old audio paths, but do whatever floats your boat. Bench testing could help reveal these inaccuracies to make them easier to replicate.

 

[living sounds]

Yup, but bench testing should reveal spurious oscillation and other stability issues.

Yes and no. Various sources (Douglas Self among them) talk about osciallations that don't show up on a normal scope and / or are very hard to catch. And most DIYers don't own an AP or other high end measuring equipment.

 

[JohnRoberts]

Yup, but bench testing should reveal spurious oscillation and other stability issues.

Yes and no. Various sources (Douglas Self among them) talk about osciallations that don't show up on a normal scope and / or are very hard to catch. And most DIYers don't own an AP or other high end measuring equipment.

That is the double edged sword from using uber fast modern op amps, and not using good practices like lead capacitance across the feedback resistor, and adequate PS decoupling.

If a scope is not fast enough to easily see a spurious oscillation  you will often see a thicker trace that should normally be a thin line. Sometimes there will be higher distortion (and sometimes a frying noise) when the op amp is oscillating.

Another variable (and I have seen this) attaching the scope probe to a circuit node could actually make it stop oscillating, making it hard to trouble shoot. I have found attaching something like a 1k resistor in series with the scope probe tip  can degenerate the scope probe capacitance enough to allow us to see the rouge circuit oscillating.

Swapping out op amps can be similar to redesigning the circuit and should not be undertaken very casually. If you don't know what you don't know, you may not get a better result than just leaving it stock.

What doesn't kill us makes us stronger (smarter) so have fun... it's not that mysterious. It all makes sense after the fact.  8)

  JR

 

[ruairioflaherty]

Another variable (and I have seen this) attaching the scope probe to a circuit node could actually make it stop oscillating, making it hard to trouble shoot. I have found attaching something like a 1k resistor in series with the scope probe tip  can degenerate the scope probe capacitance enough to allow us to see the rouge circuit oscillating.

Excellent tip John, thanks.

 

[JohnRoberts]

Another variable (and I have seen this) attaching the scope probe to a circuit node could actually make it stop oscillating, making it hard to trouble shoot. I have found attaching something like a 1k resistor in series with the scope probe tip  can degenerate the scope probe capacitance enough to allow us to see the rouge circuit oscillating.

Excellent tip John, thanks.

Not my first rodeo, but I have heard horror stories of big dollar government systems shipped with a scope probe left inside,  attached to a circuit node so it would work properly. 

JR

 

[living sounds]

That is the double edged sword from using uber fast modern op amps, and not using good practices like lead capacitance across the feedback resistor, and adequate PS decoupling.

They were all talking about the NE5532 I think. Self definitiely writes this in "Small Signal Audio Design":

The 5532 and 5534 type op-amps require adequate supply decoupling if they are to remain stable, otherwise they appear to be subject to some sort of internal oscillation that degrades linearity without being visible on a normal oscilloscope.


I have given up on high speed op amps, no need. Well, except for the LM6172. ;-)