Subject: [linux-audio-dev] Fwd: RE: [music-dsp] Transformer emulation revisited
From: Steve Harris (S.W.Harris_AT_ecs.soton.ac.uk)
Date: Thu Nov 07 2002 - 17:30:18 EET
This is from the music-dsp list, it seems relevant to what people were
discussing with SPICE emualtion of tube circuits.
- Steve
----- Forwarded message from "Sergio R. Caprile" <yogurthu_AT_arnet.com.ar> -----
> Date: Thu, 07 Nov 2002 11:05:32 -0300 (ART)
> From: "Sergio R. Caprile" <yogurthu_AT_arnet.com.ar>
> Subject: RE: [music-dsp] Transformer emulation revisited
> To: music-dsp_AT_shoko.calarts.edu
> X-Mailer: XFMail 1.5.2 on Linux
>
>
> On 06-Nov-2002 music-dsp-digest wrote:
>
> > I've been too lazy to sit down and calculate the typical Preamp hi-freq
> > boosting from schematic component values, need to do that sometime.
>
> I did some Spice simulations some time ago, as far as I remember it was more a
> sort of low-freq roll-off: single pole hi-pass between stages and poor resistor
> decoupling.
>
> > Even if the preamp has hi-freq boost that emphasizes higher-frequency
> > distortion, a downstream saturating transformer might "sound good" by
> > selectively adding extra harmonics to the low notes? From one view, the
> > transformer might be "undoing" the preamp's hi-freq emphasized distortion.
> > But the two might work synergystically to make a more complex sound than a
> > simple broadband fuzz?
> >
>
> Since there is non-linear processing involved, the transformer is unable to
> undo the preamp hi-freq distortion, but it adds a different color. The main
> reason to hi-pass before distorting stages was to avoid that cutting rumbling
> sound or excessive bass, well, that was before grunge and Metallica and...
>
> > There was no tone control setting that would completely straighten out the
> > lines in the tri wave, even with the amp running clean. There were
> > frequency-dependent effects in distorted tones which made the waves even
> > more difficult to interpret.
>
> The Fender tone network (later adopted and modified by Marshall and almost
> everyone, including Roland) is designed to color the signal, there is no "clean"
> position as in a regular Baxandall network. There is an intentional notch at
> 250/500 Hz that can be softened by the mid control and two bumps at 100Hz and
> 5KHz.
> I am not much into tubes, but most of the preamp stages where intentionally
> biased so the tubes could easily be overdriven. Some Marshall have "feature"
> capacitors running across different stages, I mean, the schems say: for model A
> this capacitor, model B no capacitor, model C, change to this value. Pretty much
> of the sound seems to have come out of sorcery ;^), or trial and error btw.
>
> James:
> You said you simulated a simplified tone control
> Did you simulate the Fender/Marshall network ? I tried to get the transfer
> function but it gets a bit complicated and I always give up before bilinear
> transform, and don't want to split the circuit as that might remove all the
> interactions along the controls.
>
> Regards
>
>
>
> --
> --------------------------------------------------------------
> Sergio R. Caprile, Electronics Engineer, Bs.As., Argentina
> http://www.geocities.com/scaprile
> --------------------------------------------------------------
>
> dupswapdrop -- the music-dsp mailing list and website: subscription info,
> FAQ, source code archive, list archive, book reviews, dsp links
> http://shoko.calarts.edu/musicdsp/
----- End forwarded message -----
This archive was generated by hypermail 2b28 : Thu Nov 07 2002 - 17:37:32 EET