[Patrick Turner]

Today I tested the whole Fender Deluxe with its vibrato channel which now has a CF buffer driving standard Fender tone stack but with a linear 10k0 pot for mid adjust. The CF buffer increases the gain by an average of +3dB.

The 1M volume pot following tone stack has 5pF across top of pot and wiper. But standard Deluxe schematic shows a fixed 47pF, so at some volume setting there is a maximum increase of HF because of the combined effect of following 12AX7 Miller input C and the C shunting part of the volume control resistance. Because this particular amp has 5pF, it is perhaps a mod, the HF boost due to 5pF is negligible. I found 47pF at volume pot would boost 5kHz by +3dB,( less at lower F and not much above 7kHz.) But with 47pF, the extra HF varies considerably with volume level because there is an effective CRC filter from tone stack with C drive with a much variable R element of volume pot then Miller C. Some amps have the volume control fitted with a pull switch for some extra HF boost, aka presence. Adding another pot and C to vary presence could be interesting, easier than replacing existing pot. But just remember the speaker may not reproduce much above 3kHz.
There is a danger that with treble turned fully up, and with presence due additional C in path, there can be unwanted HF oscillation via stray C and the high gain of so many 12AX7.Maybe using 47k in series with 47pF across volume pot might avoid such problems

With all controls at 5, mid rotation, the response is not flat, but with
50Hz = 0dB,
100Hz = +1.5dB,
220Hz = 0dB,
500Hz = - 1.5dB,
1kHz = 0dB,
2kHz = +2dB,
3kHz = +5dB,
5kHz = +8dB.
10kHz = +10dB.
This doesn't matter when considering the very variable levels with musical signals from a guitar pick up.

The knob settings for best 1kHz square wave and flattest sine wave response possible with vibrato channel, and **at the amp output** with 8r0 dummy R load is B3, M5, T2, and Volume at 6.

50Hz can be varied +11dB, -8dB.
100Hz can be varied +8dB, - 5dB.
least level adjustment occurs at 320Hz.
2kHz can be varied +14dB, - 5dB
5kHz can be varied +18dB, -2dB.

Maximum mid control is possible with both B and T set at 10.
The mid dip thus created occurs about 320Hz with M at 5.
Dip can be varied +3dB and - 7.5dB at 320Hz, and less than +/- 2dB at 140Hz and 600Hz.
The Q of the maximum dip at M1 = 2.3, higher than 1.0, and slopes of both dip sides is greater than 6dB/octave, and just how well that interacts with perceived sound is anyone's guess. Nobody seems to object to the mid dip.
But to be able to tune the F of the dip and to adjust the Q would at least require an extra triode and 2 pots and some R and C - IMHO.

All F response measurements done without any stage of the amp clipping.

There is no master volume control between preamp 12AX7 just ahead of power amp. Therefore the preamp tubes ahead of the power amp cannot be easily overloaded without power amp clipping first with just 2.5Vrms being made by preceding preamp tubes. To make the preamp tubes clip would need them to produce about 50Vrms.

It would be possible to overload the input tube which drives the tone stack by using an external preamp between guitar and amp input, then set volume at low, so cleanly amplify an already distorted signal, as required by some for venues where low volume is wanted. But other mods are possible to enable devices to work to generate say 7% even numbered H without clipping to thus possibly warm the sound but I won't say just how now because its far too complicated to explain here and now.

Probably, some guitarists would find that a good starting point is setting both B and T to 2.5,
M to 5, and then just boost bass and treble if either of both sound low, and then adjust mid for best sound.

The Deluxe has 2 x 6V6 driven by a 12AT7 in LTP mode. B+ = 400Vdc with fixed grid bias.

OPT has ZR = 6,400r : 8r0.
Maximum Po = 21Watts at clipping onset for 13Vrms to 8r0.
Power amp needs 2.5Vrms input at 12AT7 grid pin 2.

The amount of GNFB = approximately 2dB with 8r0 load.
The NFB network has 820r : 47r so GNFB voltage at 12AT7 grid pin 7 = 0.7Vrms,
and shunting the 47r to remove all GNFB gives Vo increase = +2dB.
I calculated the amp output resistance = 18r, so damping factor with 8r0 0.5 = poor.
But low DF suits such amps where the sound can be allowed to be higher from speaker driver which has peaked Z at bass and around Fs. It also has increasing Z and sensitivity between 300Hz and 2kHz. So the speaker helps to act like a tone control with low bass boost and highish treble boost.

I expect that my customer's complaint about the sound not being as good
as his other old 1965 Fender combo is because the re-issue modern speaker
driver has different ( worse ) sounding characteristics compared to those made 50 years ago.

Having a control to reduce NFB to virtually zero does almost nothing to the
sound or to the measured performance because such a small amount of GNFB
is involved. Much is written about the sound of NFB in guitar amps, but
in the case of Deluxe there is SFA to actually write about.

There is a vast amount typed about guitar amps and by ppl who never ever measure anything or make a graph of responses et all and you may all assume that 90% of what is typed about guitar amps has a very horse **** %.

And nothing the amp can do will make up for the absence of talent in the guitarist.

Patrick Turner.