Shielded cable in Rocket builds?

[Mark]

I've noticed all of the rockets builds I've seen don't use shielded cable on the input of the amp. The 33K grid stopper could be mount on the tube itself and shielded cable could be used. I'm sure no one would notice an extra six inches of shielded cable, the other point that gets missed is the wiper of the volume pot.

Has anyone experimented with shielded cable in these amps?
Of course this is about reducing the noise floor of the amp which might be handy for studio use and situations where higher gain might be desirable.

[Reeltarded]

The Rocket doesn't have enough gain up there to care much. When you use even 3 inches you setup a situation with reduced clarity of that air above the notes that is best dealt with at the cut control on the other end of the amp.

That said, almost every other amp I love has a shielded feed to V1.

Best to put the input as close to the socket as you can get it on any of the Wrecka and get away with the least shield possible. The inaudible frequencies make the amp sound alive in image.

I think.

[Jackie Treehorn]

I used shielded cable on the input. Seems like good practice to me, not much capacitance in those short runs and noise degrades the sound in ways that can't be fixed by a slightly different treble or cut setting.

[M Fowler]

I looked at build pictures of the 12 different Rocket amps I've built and none have shielded wire on the input and have the 33k grid blocker on the board.

Mark

[kwijabo]

i've tried both out. it initially had the 33k on the board with unshielded cable. then i replaced it with a shielded cable and the 33k got attached directly to V1's socket. i could not hear any difference between the two.

also, i recently got done with a 6V6, diode rectified version of this schematic with a slightly different layout. on that one i used an unshielded cable to the 33k right on the socket. no issues, even at full volume.

[strato17]

I didn't use shielded cable on mine, and my to let build is probably the quietist amp I own. Even in the studio with the volume dimed, it's very quiet.

[Blackburn]

I didn't use any in mine. Like Reel said, these aren't gain monsters and I don't think they will suffer without it. I don't think using any will hurt the amp either. Mine's done fine this long, but then again, I don't get any further than noon on the volume.

[romberg]

I can't image (does not mean there isn't any) a reason why adding a shielded wire from the jack to the first grid could hurt anything. There is already 18 feet of shielded cable on the input already. A few more inches can't make that big of a difference.

Mike

[Jackie Treehorn]

The problem with RFI is that if you take your amp out somewhere, it may be exposed to more trouble than you get at home. There's a rehearsal studio in Alameda, Ca which is located near an air force base. Every so often this incredibly loud blast of white noise would burst out of my amp; very spinal tap. Everyone would look at me like it was my fault. It actually was the cable I was using. It was a Mogami mic cable I wired with a telescoping shield. When I switched to a coaxial, everything was fine. I played a gig at a brewery hat had horrible noise problems. I had to switch to my spare strat which had a rwrp pickup.

But, that's just the noise that gets so bad you notice. It very well may be that different RFI conditions just lead to one of those situations where your amp just doesn't sound as good or as consistent as you remember.

This type of minutia does seem to fall under a builder's style moreso than any hard rules, but I like to use the shielded runs to a turret right next to the grid and put the grid resistor right between the socket and turret.

[Mark]

I can hear the difference between a 10 and a 20 foot cable, but I'm sure I won't notice an extra three inches of co-ax. The other area I've had issues is the wiper of the volume pot. As we know most people are using volume pots with a 10% taper so with the volume at noon there is only 10% of the output of the first stage present.

[LeftyStrat]

I agree with everyone that shielding from the input jack to the V1 is a good idea.

However, I have gone crazy in shielding an amp and put a wet blanket over the tone. I'm not sure I understand the physics of it. But it seems coaxial cable behaves differently when subjected to a large AC voltage with a big DC offset.

I hate subjective observations without the science to back it up.

Tiny signals need shielding, but remember some old modular synthesizers used unshielded patch cables, with just 10 V p-p signals.

[Tillydog]

However, I have gone crazy in shielding an amp and put a wet blanket over the tone. I'm not sure I understand the physics of it. But it seems coaxial cable behaves differently when subjected to a large AC voltage with a big DC offset.

I hate subjective observations without the science to back it up.

One difference is the output impedance of whatever is sending signal into the co-ax:

A guitar pickup might have an *impedance* of ~20k ohm (DC resistance + inductive reactance)

If it's connected to ~20' of 40pF/foot guitar cable, then it sees ~~1000pF of capacitance.

=>The signal will be reduced to 70% at ~8 kHz (rough maths, I know).

A foot of cable either way won't make much difference, and the cable is pretty much unavoidable (unless you go wireless).

If there is (say) 1' of co-ax from the wiper of a 1Meg gain pot, then the output impedance could be as high as 250k ohm when the pot is set at mid value.

Miniature co-ax often used for internal wiring might have a capacitance of 100pF/foot.

=> The signal will be reduced to 70% at ~6.4 kHz

i.e. 1' of small coax inside the amp can have a greater effect than 20' of guitar cable!

Now suppose we add another, similar screened cable from (say) a volume pot. Because we've already lost 30% of the signal at 6.4kHz, putting the signal through this new cable will lose another 30%, so it will be down to one half of it's original value at 6.4 kHz, and down to 70% at ~ 4 kHz

Each time a length of co-ax is added, its effect *mulitplies* with what is already there. The graph shows the effect of 1 to 4 sections as described above.

Even if you used that lower capacitance guitar cable inside the amp, you can't escape the multiplying effect 4 sections using 40pF/ft cable loses 70% the signal at 7kHz

The wiper of high value pots is probably the worst case to consider, but every section will have a similar calculation, and they all multiply together.

For anyone wondering about what frequencies come out of a guitar (including the *longitudinal* string vibrations!), I think this is an excellent and thorough description (a bit of a long read, but well worth it IMHO):

http://www.alcarruthluthier.com/Downloa ... Theory.pdf

HTH

EDITED to correct my translation of -3dB from 50% to 70% voltage, and to correct the formula used for multiple sections (oops!)

[martin manning]

Good stuff, Tillydog! I'll point out that 1' might be a bit long for the typical coax run inside an amp, and cable of 25-30 pF/ft is not uncommon, so the capacitance to ground may be as low as 1/4 of what you assumed in your calculations.

[Tillydog]

Thanks!

It's just round numbers to show that there is some science that underpins observations.

( Even 4 x 8" sections of 30pF/ft co-ax on a 1 meg pot have a -3dB point at ~6kHz )[Edit] Ignore that - it would be about 14 kHz [/edit]

The moral of the tale is to be very careful with what shielded cable you use, and to use it sparingly (but you know that!).

[Mark]

Hi Tillydog

I think I need a few equations in your explanation. I think that would walk me through you explanation easier.

If it's connected to ~20' of 40pF/foot guitar cable, then it sees ~~1000pF of capacitance.

BTW shouldn't that be roughly 800pF?

Thanks