SSL 4k Line Input Schematic review

[conleycd]

Hey folks,

As is the flavor of the day (and my ever expanding project)... I was thinking of using the SSL 4000 series line input as the front end of my summing project.

I've grabbed the relevant schematic which looks simple enough.  I follow most of it up to the potentiometers.

Is the 5k pot acting as a pad while the 28k (what??) pot is setting gain?  I think a 25k pot has got to work here?

If so... was this some sort of stacked pot with different values operating at the same time?

Any thoughts are most welcome.

 

[Matt Ward]

You're pretty much there; the 28k pot is the line gain pot on the front panel and the 5k pot is a preset; I think used to set unity gain when the 28 k pot is in its centre (0dB gain) position.  FYI The numbers 55,57,59,60 and 61 refer to pins on the edge connector of the line amp card and so it's clear the 28k pot, 5k pot and 5k1 resistor to ground are off board.  The "on 611" next to the 5k0 pot and 5k1 resistor means that it is on the channel card (not front panel) and the pre. tells us it's a preset of some kind of preset.

Most / all (can't remember exactly) pots on the 4k are stacked / ganged, usually because of recall.  These 2 definitely aren't ganged.

Is there any particular reason you chose this topology?

 

[conleycd]

That makes sense...

Why the front end of an SSL 4k channel? - Well just because it's the front end of an SSL channel.  I need a manner to unbalance and add a little gain if needed before hitting the active summing stage (virtual ground mixer).  How much of the sound does it account for?  Probably not much of anything - but why not?  I also will unlikely use 30.1k resistors with 0.01% margins!

There's lots of options out there but this seems a little more fun than a classic 4 resistor differential amplifier followed by a classic opamp amplifier.

Bunch of great albums were essentially summed through SSL 4000 series boards.  I'm probably not going to use their topology for summing - if I could find it.  I will probably use a very common active-based combiner with a pair discrete set of opamps and some iron on the ends.

CC

 

[Matt Ward]

but why not?

just because it's the front end of an SSL channel.

Bunch of great albums were essentially summed through SSL 4000 series boards.

Well, there are probably many better ways of doing it, but nothing that significantly wrong with this.  A lot of the work is done for you, and you know it's out there working all over the place.  Just don't kid yourself that it counts in any way towards an SSL "sound" or characteristic.

 

[conleycd]

Well, there are probably many better ways of doing it, but nothing that significantly wrong with this.  A lot of the work is done for you, and you know it's out there working all over the place.  Just don't kid yourself that it counts in any way towards an SSL "sound" or characteristic.

There's not many better ways to do this.  A couple but all variations on a similar theme.  I can have an input transformer (nice option - great CMRR), an integrated balanced line receiver (ala THAT and similar stuff), or a 4 resistor differential opamp (which is essentially what this is).

I have already acknowledged that this will unlikely impart any significant "SSL sound" but SSL 4000 stuff was centered around a lot NE5532ANs - so that's a portion.  But really... whatever "sound" is derived from an eq/dynamics flat SSL channel with the line inputs being used is essentially this circuit.

CC

 

[conleycd]

What is that 22pf capacitor noted with each opamp?  Is that over the power rails?  I don't think it is in the feedback to the inverting input (which I believe is the top portion of the opamp) because I think that cap to tame high end is that 33pf capacitor.

Thoughts?

CC

 

[Michael Tibes]

I have already acknowledged that this will unlikely impart any significant "SSL sound" but SSL 4000 stuff was centered around a lot NE5532ANs - so that's a portion.  But really... whatever "sound" is derived from an eq/dynamics flat SSL channel with the line inputs being used is essentially this circuit.

CC

...plus the fet switching, some buffers, the mix bus and the vcas (faders / bus compressor)... My guess is that the fets and the vcas have the biggest effect on the sound. So I wonder, if you're on the right path if you want to recreate the 'SSL-sound' because the line input might be the most transparent part of the whole console. A transparent balanced line receiver might be esier to achieve with a THAT chip nowadays, but anyway, there's nothing wrong with the SSL circuit. I'd probably just knock up 1 channel and see how it sounds - if it 'sounds' at all

The 22pF are the compensation of the 5534 (pin 5 / 8 ), aren't they?

The 33.1k resistors could be hand-matched. The absolute value is not important, but the matching.

Michael

 

[abbey road d enfer]

There's lots of options out there but this seems a little more fun than a classic 4 resistor differential amplifier followed by a classic opamp amplifier.

I don't really understand this line, because the SSL line input is basically a "classic 4 resistor differential amplifier followed by a classic opamp amplifier", to which have been added just the necessary things to make it work.

 

[conleycd]

There's lots of options out there but this seems a little more fun than a classic 4 resistor differential amplifier followed by a classic opamp amplifier.

I don't really understand this line, because the SSL line input is basically a "classic 4 resistor differential amplifier followed by a classic opamp amplifier", to which have been added just the necessary things to make it work.

Yup, you're right.  The values of some components are different but yeah is a 4 resistor balanced line receiver with an opamp following for gain (with some stuff...)  I still think it's a simple enough and neat circuit to make a front end line input for any variety of uses.  I still think it is neat that when I'm done with it - it will be the same as the line input of a SSL 4000 (again acknowledging that the sound it imparts is likely nil from 'that' SSL sound).

CC

 

[abbey road d enfer]

IMO, the trimpot is unnecessary and detrimental to the noise performance. Just take it off and work out the correct value for R41 (anyway, you'll have to work it out for the different pot value and different level diagram).

 

[conleycd]

I'm thinking as well - just noticed - the output of this whole diagram is polarity inverted as well.  The gain stage in the second opamp is utilizing the inverting input.  Right?

 

[Nora]

I'm thinking as well - just noticed - the output of this whole diagram is polarity inverted as well.  The gain stage in the second opamp is utilizing the inverting input.  Right?

Yes, but since the input is balanced, you can compensate for the correct polarity.
Or you could leave it to the summing amp (if it's active summing)