shileld of input coax to b+

[VacuumVoodoo]

I'm pretty sure capacitance of the shielded cable doesn't vary much with voltage. Perhaps an amount measurable with a capacitance bridge but not audible. Question is how much voltage the isolation between shield and signal wire can take before braking down. Risking to sound like a broken record: don't do it. Unless you want to impress the ignorant and fashion yourself an electronic mystic.

[Super_Reverb]

If the preamp sagged, the plate voltage would change on the shield. Would this then in return effect the capacitance of the shielded cable?

The shield voltage would change, that is, the voltage across this parasitic capacitance would change, but the dielectric (inner insulation) is so thick, that I can't see any measureable effect. Ceramic and other caps have very thin dielectrics, so some of the have measureable voltco of capacitance.

I agree with VV's earlier post: place a small cap grid to plate of 1st pre stage (start with 10s of pF) if'n you want to roll off treble. Remember, Miller effect makes this capacitance effect larger due to largish (~60) inverting gain.

cheers,

rob

[tweedeluxe]

Now, if you can live with losing 4db gain in the first gain stage....remove its cathode decoupling capacitor, use a plastic input jack NOT grounded to the chassie (have ground at PSU) and connect one end of input cable shield to the jack and the other to 1st gain stage cathode.....the treble clarity will surprise you.

Aleksander, why must one remove the cathode cap for this to work?

I was just inside a stock 1982 JCM800 that had the shield tied to the plate on the input triode. Doesn't seem safe to me either.

[Cliff Schecht]

I tied the coax shield to the B+ on the first Trainwreck I built (the ugly one with all of the band-aid fixes). It did a squash the oscillation problems I was having but it also knocked out some high-end in the process (as well as the noise). It's obviously important to make sure that the shield will never short to the input jack but a bit of heatshrink did the trick for me.

[VacuumVoodoo]

Now, if you can live with losing 4db gain in the first gain stage....remove its cathode decoupling capacitor, use a plastic input jack NOT grounded to the chassie (have ground at PSU) and connect one end of input cable shield to the jack and the other to 1st gain stage cathode.....the treble clarity will surprise you.

Aleksander, why must one remove the cathode cap for this to work?

I was just inside a stock 1982 JCM800 that had the shield tied to the plate on the input triode. Doesn't seem safe to me either.

The trick works by bootstrapping the shield with signal voltage equal to what is on the inner signal wire. Removing cathode capacitor makes the gain stage look like cathode follower when observed from the outside. That's where we take the bootstrap voltage for the shield from. Amplified signal is still present at the plate. The physics behind is that when both signal wire and its shield swing in unison with the signal source the cable capacitance is virtually reduced. Look at it as opposite to Miller effect which makes a small capacitance look big, bootstrap makes a capacitance look smaller.
I believe there's a company selling guitar cables with bootstrap buffer built in into a box attached to the cable. It needs external power supply though.
What I described puts guitar cable shield at about 1.5V above ground DC wise, it's still grounded AC wise through cathode internal impedance of 2-300 Ohm.
Bootstrapping shield of a signal wire to reduce cable capacitance and thus increase bandwidth is an old technique initially developed in 1950s for use on long cables in measurement and data acquisition systems. Nihil novi.

Page 18 in this collection describes 3 implementation variants.
http://www.scribd.com/doc/23975182/400-Ideas-for-Design