Class A vs AB impact on bus compressors and use of neve cards in output buffer

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I am coming to you as a person who has no knowledge, asking people who claim to have some, or even a lot of knowledge... whether A vs A\B has any impact on the DYNAMICs of the neve card, if it were wired to the output buffer of a compressor, or if it is static. Whether it would behave just the same run after the compressor (inline).

This, in a nutshell, is the issue you're having getting your question answered. The question as you've framed it, as someone who claims no knowledge, isn't making sense to those *with* knowledge - and believe me, you have had some serious electronic engineering heavyweights already respond to this thread. So let's try to unpack things a little.

The BA283 can't operate in Class A/B, so that part of your question is moot and we can leave that aside. It is a Class A output stage. I think you're then trying to ask whether the BA283 will affect the action of the compressor before it - no, I can't see that it would, particularly if there's a buffer between the sections.

I think you're also asking whether the BA283 stage will saturate in a level-dependent way, or independently of the output level of the compressor section. I'd expect it to be level-dependent. It may therefore smooth out some loud peaks slightly, depending on how hard it's driven.

If there's something else you're getting at, perhaps have a go at rephrasing it and we'll see what we can do. In particular, I don't understand your distinction between "wired to the output buffer of a compressor" and "run after the compressor (inline)", so any schematics you can point to in order to clarify that may help.

*Edit*: on rereading, I understand you're also asking whether there's any real point in having the BA283 in the same box as the compressor. Not a great deal, as far as I can see - you maybe avoid balancing and debalancing the signal if everything's in the same box, but that may not be particularly audible and is unlikely to change the action of the compressor or the tone of the BA283 (as long as all the impedances are handled properly). Convenience? Rack space?
 
whether A vs A\B has any impact on the DYNAMICs of the neve card, if it were wired to the output buffer of a compressor, or if it is static. Whether it would behave just the same run after the compressor (inline).
...if your definition of dynamic is how signal behave traveling an active (amplifier) chain then maybe...
...if your definition of dynamic is how signal behave in a voltage controlled amplifier (like 4k comp) then NO... topology in a 4k type compressor don't feed side-chain from output signal (whatever output buffer you use...)

Best
Zam
 
This, in a nutshell, is the issue you're having getting your question answered. The question as you've framed it, as someone who claims no knowledge, isn't making sense to those *with* knowledge - and believe me, you have had some serious electronic engineering heavyweights already respond to this thread. So let's try to unpack things a little.

The BA283 can't operate in Class A/B, so that part of your question is moot and we can leave that aside. It is a Class A output stage. I think you're then trying to ask whether the BA283 will affect the action of the compressor before it - no, I can't see that it would, particularly if there's a buffer between the sections.

I think you're also asking whether the BA283 stage will saturate in a level-dependent way, or independently of the output level of the compressor section. I'd expect it to be level-dependent. It may therefore smooth out some loud peaks slightly, depending on how hard it's driven.

If there's something else you're getting at, perhaps have a go at rephrasing it and we'll see what we can do. In particular, I don't understand your distinction between "wired to the output buffer of a compressor" and "run after the compressor (inline)", so any schematics you can point to in order to clarify that may help.

*Edit*: on rereading, I understand you're also asking whether there's any real point in having the BA283 in the same box as the compressor. Not a great deal, as far as I can see - you maybe avoid balancing and debalancing the signal if everything's in the same box, but that may not be particularly audible and is unlikely to change the action of the compressor or the tone of the BA283 (as long as all the impedances are handled properly). Convenience? Rack space?

Yeah that's what I thought. Was just sort of confirming it. The design is something like an SSL clone with a built in culture vulture then. Kinda cool honestly.
 
I am coming to you as a person who has no knowledge, asking people who claim to have some, or even a lot of knowledge... whether A vs A\B has any impact on the DYNAMICs of the neve card, if it were wired to the output buffer of a compressor, or if it is static. Whether it would behave just the same run after the compressor (inline).
The answer is it depends.
In the absence of a proper schemo, I can only speculate.
The class of operation of any stage that is in the signal path may have an impact on sound, albeit quite minor.
As I mentioned earlier, it probably matters for audiophools.
But the presence of an additional stage in the signal path may have a significant impact on sound (or not), because it may change the operating level and alter (in a negative or positive way) the response. Again, only a proper schemo (or a detailed level diagram) can determine that.
Finally, the presence of an additional stage will not impact the side-chain, thus cannot introduce dynamic response changes.
I am asking if there is any benefit to wire this into a compressor at all, whether class makes a difference in how it behaves in this instance. Because I would like to purchase a regular off the rack SSL clone and then wire these after if they would behave the same.
You mean you want to take an SSL4000 clone and graft a BA28/3 at the output?
A typical SSL4000 clone already has an output buffer, so I suspect the only reason to do so is to inject some of the "Neve magic" to it, whatever it is.
Apart from adding some floobydust, it won't change the dynamics.*
Remember that floobydust is proportional to the amount of work and money you put into it.
The major effect comes from lightening your wallet and increasing your level of expectation.

*The BA283 is typically used with some significant gain (about 12-20dB). The SSL4000 does not need this extra gain, so I don't know how they manage that. There are several ways of doing it.
 
OK, to answer your question, Stam does not know what he is talking about. The Neve BA283 is class A, end of story. It cannot be modified for class A/B. So his statement it is working is class A/B is incorrect. The only way he could be correct is if his BV238 board is actually a copy of a BA440 which has a class A/B push pull output

To answer the other part of your question, how an amplifier reacts to a load depends on its design and i particular its output impedance. Class A and class B amplifiers can both be designed with low enough output impedances not to be affected by their load. The class of amplifier does not specifically affect its output impedance. All pro audio output stages should be unaffected by the load they drive.

"Does anyone know the answer to my question or does everyone here just copy schematics and know exactly the same amount as Stam?"

I hope that answers your question and no I do not copy schematics. I design tube mixers and I designed consoles at Neve back in the 70s.

Cheers

Ian
 
I found this on Stam Facebook from 2017, maybe it will help clear up some doubts

View attachment 93593

Mmmm...freq response - fair enough ; THD - fair enough. "Fattens the audio" - what does that actually mean and how - extra harmonics aka distortion so not so good for the reduced THD ? Freq response deviation ie 'EQ' ? Intentional phase shifting /delay of selected frequencies ?
 
Yeah that's what I thought. Was just sort of confirming it. The design is something like an SSL clone with a built in culture vulture then. Kinda cool honestly.
Not really. It doesn't have anything even remotely close to the culture culture circuitry wise. It's an ss bus comp with some "neveish" output line amps....they do make a sonic different but it's pretty subtle. Way less than a 1272 line amp after the comp 😁
 
There is no mention of Neve or the BA283 in that article. Just a couple of Carnhill transformers and an unspecified class A/B buffer. I have blown up the picture and I can see no sign of chunky 2N3055 transistors so whatever is in there it ain't BA283 circuit.

Cheers

Ian
 
Not really. It doesn't have anything even remotely close to the culture culture circuitry wise. It's an ss bus comp with some "neveish" output line amps....they do make a sonic different but it's pretty subtle. Way less than a 1272 line amp after the comp 😁

Line amp at unity? What do you think? Just want to cop the sound. The Neve MBC does the same thing, so technically I think you need to compress the signal if only a touch. But it just distorts in the right way just sitting in the lower mids in a way that makes me just immediately think "polished."
 
The answer is it depends.
In the absence of a proper schemo, I can only speculate.
The class of operation of any stage that is in the signal path may have an impact on sound, albeit quite minor.
As I mentioned earlier, it probably matters for audiophools.
But the presence of an additional stage in the signal path may have a significant impact on sound (or not), because it may change the operating level and alter (in a negative or positive way) the response. Again, only a proper schemo (or a detailed level diagram) can determine that.
Finally, the presence of an additional stage will not impact the side-chain, thus cannot introduce dynamic response changes.

You mean you want to take an SSL4000 clone and graft a BA28/3 at the output?
A typical SSL4000 clone already has an output buffer, so I suspect the only reason to do so is to inject some of the "Neve magic" to it, whatever it is.
Apart from adding some floobydust, it won't change the dynamics.*
Remember that floobydust is proportional to the amount of work and money you put into it.
The major effect comes from lightening your wallet and increasing your level of expectation.

*The BA283 is typically used with some significant gain (about 12-20dB). The SSL4000 does not need this extra gain, so I don't know how they manage that. There are several ways of doing it.

Definately tought to seperate whether the impact is "subtle" or "dramatic." But it's there. Put another way, do you think a neve line amp at unity would add any Neve mojo? I know it could be attenuated down to line level and straight up 1073'd back to line level but I would want to avoid that if possible for obvious reasons (noise floor). I don't even care about noise that much and love tape hiss but I won't want to take a nice signal I've spent so much time taking care of through the signal and recording path and then hammering it down and back up right at the end.

Suppose it circles right back around to demanding a saturation stage and actually delving into what the RND MBC and SSL Fusion are doing.
 
Is that the SA4000 MKI +? Or is that the full boat "neve mod"
As far as I know the Mk1 has no output transformers and a switch to turn them on at the front panel. So, this is obviously some MK (higher number) version. I read somewhere that there is an API mode as well.
 
As far as I know the Mk1 has no output transformers and a switch to turn them on at the front panel. So, this is obviously some MK (higher number) version. I read somewhere that there is an API mode as well.

The + has transformers ;p I think this is the MK I + and pictured below is the MK II with "british" (neve) mod.

stam-audio-engineering-sa4000-mk2-british-mod-3638852.jpeg
 
Fine, I don’t follow Stam innovations so I might be wrong. Btw, how do you know it's the MKII in the picture?
 
Here are a few more pictures from Stam Facebook that seem to confirm that the pictures I have already posted show the MKII version. Maybe that blue empty PCB is doing that great improvement. Or is something hidden underneath?

1651871754867.png

1651871819988.png
 

Latest posts

Back
Top