[Patrick Turner]

hugeshows
Feb 16
Hi Patrick,

You're totally right about the non-linear response of guitar amps in general, they're notoriously curvy when you plot the response.

When you examined the response in your experiments, did you use a speaker or a dummy load? The reason I ask is because the 'full range' single driver type setup most common in guitar amps leaves a lot of room for the impedance curve of the speaker to effect overall frequency response, often showing reduced or exaggerated efficiency at certain spots.

I understand that, and most 12" drivers used in an open backed cabinet make a lot more SPL between say 500Hz and 2kHz than at F outside these F.
Impedance isn't a huge concern because most guitar amps do in fact have maybe 10dB global NFB which is enough to reduce the pentode output stage Rout to
sensible levels.

The problem is to get meaningful amounts of boost and cut, and a lack of interaction between the bass and treble and mid control if there is one.
The r s control NFB applied, fct Ha mts

If you haven't already tried the speaker variable in your listening, I'd perhaps start with at least ruling out the possibility that modern drivers aren't more to blame than modern electronics.

Possibly old drivers are more sensitive than more recent mades, so all the more reason why a greater variability of bass, treble and mids is needed.


Maybe A/B the two amps with a signal sweep and across a resistive dummy load?

I done all that, and the older one had much more bass than newer re-issue, and had more treble boost.

Pink noise can also work but often loading at all frequencies changes response across the spectrum in some areas, which can be misleading since guitar signal (hopefully) rarely approaches pink noise in terms of spectrum use.

Tone controls offer a varying load to the tube driving them. It seems best sound and tone variability occurs with 1/2 12AX7 gain tube anode direct coupled to 1/2 12AX7 as a cathode follower. The Fender Deluxe has tone stack driven from anode circuit, while the Early Fender Twin Reverb has the CF - and the better sound by far.

Response curves with pink noise bight be quite different than with a single sine or sqaure sweep. And as much as I hate to suggest tube rolling, there's always that variable to consider as well.

Different brands of 12AX7 make very little difference with what I measure, although sound may change a little.
Its the engineering and numbers that make the most difference and the curves of response with sine waves and variability. 99% of BS artists online who have a lot to say about guitar amp mods often have no clue what they are talking about.
The owner of Fender Deluxe I have here asked me to reverse the R&C parts change he did himself after reading a long winded article online by a wannabe guitar amp "specialist" He said the changes made no difference or made it worse.
When I analyzed, I could see why.


Regards,

-forkinthesocket

I found that first need for guitar amp tone stack is to have low Rout tube driving the "stack". While it could be a 12AX7 CF, when there isn't one, one has to think of putting in an extra socket and tube OR using a BJT direct coupled as emitter follower. Maybe MJE340 will do.........
Collector to B+,
Base to gain tube anode,
Emitter to 56k to 0V so that Ic = 3.5mA approx.
This can be done easily, and the sound is determined by the gain triode,
as followers are very neutral sounding ( very low THD)

The Baxandal tone control gives the MOST tone variability.
The Deluxe has 260k log bass pot and when set at No5 level, midway rotation, there is 21k below wiper and 240k above. The treble is also a 260k "log" pot but with 50k below wiper and 210k above at 1/2 rotation.

The Baxandal passive tone control is the easiest to build to get a flat
response with B and T settings at 5. Unlike most other tone stacks designed by penny pinching accountants, the Baxandal gives the most cut and boost and steepest curves closest to 6dB/octave maximum possible.
Just what is possible as a mid control boost and cut remains to be seen
when I send another day on Wednesday making experiments.
But the Baxandal must be made carefully. I have 47k in series with bass pot,
then no usual shunt between top and wiper. Wiper goes to 200k to treble RC array.
But when set at 5, bass should be flat from 1k down to 50Hz. To ensure this,
the ratio of 47k : 240k above pot wiper must be the same as the grounding R and the 21k below wiper, so I found R = 3k9.
This means that for say 10V input, you get 1V output at 5, and max boost and cut
is over +/- 12dB, enough. The 2caps used to shunt wiper to each pot end must have the same XC ratio of 240k : 20k, and if R or C values are more than 10% inaccurate, then you get much poorer max boost or cut, or vice versa, and the best performance is given with the correct R&C values and correct ratios.
The feed to treble boost/cut pot wiper probably will have 220k that connects to 850k log volume pot. Some small C will be put from top volume pot and wiper to at least neutralize the Miller of following 12AX7, and perhaps give slight presence.

I have yet to figure out a mid adjustment which puts a dip up to -10dB in response at 400Hz with Q a bit over 1.

The simplest things can trap me in a shed for days, but then I like to give my best input before asking the owner to tray it out to see how it sounds.
Patrick Turner.