[sdiy] Any noise avantage to parallel 2164 VCAs?

[rburnett at richieburnett.co.uk]

An interesting thread.  Thanks for the explanation guys.

I'm enjoying this list's return to usual form: discussing real technical 
issues and advancing the art, rather than the recent detour of arguing 
about details of the past.

-Richie,



On 2018-06-18 10:20, Steve wrote:
> Neil:
> Oh, the 2164 has current outputs, I wasn't aware. Thanks for the
> explanations :)
> I have some ols SSM ICs somewhere, funnily years ago when I bought
> them, as well as some old stock SSM VCF chips when they were
> reasonably priced from chinese sellers, I somehow never dared to
> actually use them, and used the easily bought "everywhere" LM13700 for
> "everything", since I, the newbie, didn't want to fry those kinda
> chips in my first lessons involving magic smoke. (like when in the CAD
> prog I flipped the opamp, but not the power block. I still have the
> foto of the magic smoke escape hatch on that opamp.) Now you can buy
> all those clones, and even some remade parts again. Maybe I should
> start using those nicer parts now ;) Who knows for how long this will
> go, though.
> 
> VON: "Neil Johnson" <neil.johnson71 at gmail.com>
> BETREFF: Re: [sdiy] Any noise avantage to parallel 2164 VCAs?
> Steve wrote:
>> 
>> Douglas Self wrote something about using opamps in parallel for
> noise
>> reduction, in his small signal audio book. Don't remembner details,
> and no
>> idea how much it's applicable to that VCA chip. A former colleage of
> mine
>> mentioned it, he tried it and it did what it was supposed to. I
> think the
>> outputs were not "tied together", one of them had a small resistor.
> Not sure
>> though.
> 
> Op-amps have voltage outputs, so you can't just connect them together
> as the individual feedback loops would fight each other. Doug puts
> something like 10R in each output before combining them.
> 
> The point is quite simple: for multiple separate devices assuming
> uncorrelated noise the signal voltage goes up with +6dB when you
> double the number of amplifiers (1 -> 2 -> 4 -> 8 etc), but the noise
> only goes up by +3dB, so overall the S/N increases by +3dB.
> 
> The 2164 is different in that it has current outputs. You can combine
> the outputs directly, for example if making a multi-channel mixer (see
> Figure 25 of the SSM2164 datasheet), or paralleling them for better
> SNR (although you cannot parallel the inputs so you still need the
> input resistor and compensator RC network on each input).
> 
> Neil
> --
> http://www.njohnson.co.uk
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy