A sim of a treble boost you might want to look at

[Gus]

A sim of a TB schematic you can find on the web
It looks to be forgiving of the NPN Si transistor used, due to the lower value voltage divider bias string along with the 2.2k emitter resistor
If you make R2 a fixed 680ohm in series with a 5K potentiometer you can bias the transistor with 6.8k and 10K collector resistors AND you might like it try it at different collector voltage

A Si power transistor build?

Any questions?

EDIT if you use a variable bias control make C5 at least 100uf

[Kipper4]

Yer but I always get this wrong. Is C6 providing a positive feedback loop?
I wouldn't know about si power transistors but I'm willing to learn.

[Gus]

Note C5.
C5 bypasses the emitter resistor so the node can't be "wiggled" by the signal above a few times the RC time constant so no feedback.
C6 looks to be a HF reject from base to emitter to ground via the emitter bypass cap
Or you could look at is as adding 1000pf to the base to emitter BJT capacitance.
I wonder why it was not base to ground?

PRR, R.G. anyone else what do you think?

[Groovenut]

It's very possible C6 is used to negate some of the high frequency loading caused by the low input impedance of the BJT, much like the similar cap in the Cornish buffer design.

[PRR]

It's a simple thing.

First thought: R3 22K is a very heavy load on the treble of a guitar pickup. An odd thing to do in a "treble boost"-- load-away the treble before you start. (Such a trick is used in some distortions, because distorted highs on top of distorted middles are annoying.)

C5 would appear to be a plain simple Emitter Bypass. At some high frequency this impedance becomes similar to Q1's internal emitter impedance. At lower frequency the gain falls off, ultimately to near R1/R2 6.8K/2.2K or about 3.

Q1 is run at 0.95V/3.3K or 0.29mA. re is 26r at 1mA, so at 0.29mA re is about 90r. Any reactance chart (I hear there are web-cheats) will tell you that 47uFd is 90 Ohms around 38Hz. Above that the gain is more like R1/re or about 75(!!). And 38hz is below guitar band, so this gain happens to "all guitar".

So part of any"treble boost" is that any signal above 40mV (which is most guitar) is overloading the amplifier and throwing harmonics.

But now look at C1 R3 (and R5 and Q1, but let's keep it simple). 0.0068uFd into 22K is a 1,064Hz high-pass. "Most" of the guitar is sloped-off. So 100Hz bass may not be hitting Q1 at the 40mV level which overloads it. 880Hz hi-A will be near full strength and overloading. This input bass-cut is a good start when distorting speech/music without destroying intelligibility and articulation with mudd.

If you throw in R5 and Q1 you may find the C1 low-cut starts well above 1KHz. So "most guitar" is heavily sloped-off, bass more than treble.

The output impedance is very nearly R1 6.8K. C4 against R1 forms a high-cut, but way up at 5KHz. Considering all the mid-high hash coming out, that's not a significant cut, but does reduce nearly-supersonic dizzle-haszh.

The alternate values for R1 R2 just do a similar thing at 2/3 the battery drain and a half-octave lower on the filters.

C6 spans a unity-gain path. Alone it can't do any real boost. Similar things are often used just to cut *radio* interference-- 1,000pFd (0.001u) across the first B-E is found in most low-cost transistor phono preamps.