Chassis ground in a star grounding scheme

[Paul P]

In the amp I'm designing I don't use the chassis for any circuit ground.
All my jacks are insulated and the pots have ground wires back to their
circuits. I bring all the circuit grounds to a single point at the chassis
ground.

Now I'm wondering, why bother connecting circuit ground to chassis ground ?
Why not let the system float with respect to the electrical system ? Wouldn't
this be safer in that it would be like the amp were run off an isolation transformer ?
It seems to me that the only way you could get a shock would be to touch two
points inside the amp simultaneously, one ground and one high voltage.

The chassis ground to the electrical system would then only be for that part of
the amp that comes before the power transformer secondary.

Paul P

[cbarrow7625]

Here's the problem with that:

In order for you to get good RF shielding for the amp, you have to connect the shield wire from your guitar cable directly to the chassis. If the guitar cable is attached to the chassis but not the circuit signal ground, then you will not have any input signal flowing!

Why do you need to attach the cable to chassis ground, you say? Well, if you don't you will take any hum, noise, RF interference, etc that is picked up on the guitar cable shiled wire (or inside the guitar) and just inject it into your amplifeir signal through the "floating" signal ground. You aren't letting the cable shield do its job of protecting your pristine signal from outside interference.

Can you do it? Yes, absolutely. I have done it plenty of times to see what the difference is in the sound & feel of the amp (there is some difference). As long as the chassis is grounded for safety, you do not have to connect signal ground to chassis ground. However, that is always in my shop in a controlled situation and not out in the real world where I have no control over RF interference & wireless mic signals etc. (Remember that sceen from Spinal Tap when the airplane radio transmits through his amp??)

Now, a way around this is to modify your guitar and input jack to use a 3-conductor TRS connector & use a shielded, twisted pair cable (also with TRS plugs) and connect the twisted pair as your signal wires & use then attach the shield directly to the chassis on the amplifier end only (do not connect the shield to the - signal ground fromthe guitar or at the amp end). That way your shield wire goes to chassis ground and protects your signal but your signal hot & ground can be left "floating".

Seems like a big hassle to me & it would only work properly with your guitar.

By the way, I prefer the sound of the signal ground going to chassis. It ends up sounding tighter & cleaner, probably due to the fact that the cable shield is actually doing its job.

Chris

[Paul P]

Now, a way around this is to modify your guitar and input jack to use a 3-conductor TRS connector & use a shielded, twisted pair cable

Yeah, I was thinking the exact same thing

I've been wondering why all guitars aren't hooked up this way. Microphones are.

My system is only going to be used in my home so it isn't any more work
for me to do it one way or another.

Once my amp is built I'll lift the signal ground and see what happens.
Thanks for the input.

Paul P

[dai h.]

The chassis ground to the electrical system

by "chassis ground" I gather you mean the safety ground from the outlet? I think this needs to be connected to the circuit ground for safety. You have a direct connection to the circuit ground via your hands on the gtr. and the string ground, so if there is a fault current and there is no safety ground I think this can end up flowing through your body through your feet, butt, lips, etc. Also I think there would be a potential difference without the safety ground of the amp and other equipment tied together via the safety ground, so shock and noise could be the result when interconnected and used together or when physically contacting another electrical device. Also, my understanding is that the chassis should be tied to circuit and should not be left floating since tying to chassis reduces noise by keeping the potential the same. Finally the 3-conductor pseudo-balanced wiring is not the same thing as the balanced wiring seen on mics, etc. but apparently some favor this wiring scheme like Pete Cornish who IIRC sells cable configured this way.

[cbarrow7625]

Dai is right on all of his points. It is a safety issue, you cannot connect to external gear without inducing massive hum, and it can be noisier (not that the signal to noise ratio off aguitar amp is that great to begin with).

The 3 conductor wiring is still un unbalanced connection. Just becasue you have 2 wires with an independent shield it doesn't automatically turn it into a balanced signal.....unless......each conductor is driving it's own input of opposite polarity and then summed back to an unbalanced signal within the amp. Once you connect one end of the signal wire to signal ground, you are unbalanced.

Or you could build an entirely "balanced" guitar amplifier straight through to the push-pull outputs. That would entirely eliminate the need for a phase splitter. That is an idea I have had floating around in my head for a while but never tried building. Getting the signals matched perfectly in frequency response and amplitude to properly drive the power tubes would be a big pain (unless you make it a very short, simple signal path).

Anyone want to build one and let the ret of us know if it works?

Chris

[Paul P]

Yes, by chassis ground I mean the connection between the chassis and the
safety ground of the electrical supply.

If the entire circuit between the power transformer secondary and the guitar
is not grounded to the electrical supply how can electrons flow from one to
the other, or to another piece of equipment, or to earth ground ? I thought
this was the whole point of using an isolation transformer.

As far as noise goes, wouldn't a complete shield act as a faraday cage
even if the shield were not grounded ? I don't think you'd want to ground
(ie. tie to the power transformer's secondary center tap) your strings,
bridge and so on, so that you don't act as an antenna into the system.

BTW, I'm discussing strictly out of curiosity and fully intend to ground my
circuit ground to the chassis, unless I come across irrefutable evidence that
I shouldn't, which is doubtful.

Paul P

[dai h.]

think about this--if electronics needed a physical connection to ground to work, then iPods, airplane electronics, satellites, etc. wouldn't work. Electronics are made of circuits (send and returns) which don't necessarily need to be tied to physical earth.

As far as noise goes, wouldn't a complete shield act as a faraday cage even if the shield were not grounded?

not sure but I'd guess it would and that tying to earth makes it more effective (the better the lower impedance the it is?)

I don't think you'd want to ground (ie. tie to the power transformer's secondary center tap) your strings,
bridge and so on, so that you don't act as an antenna into the system.

I think basically yes the loop out to the gtr. and back is an antenna (guessing), but as long as the signal is more powerful than the noise you are okay(?). Maybe freq. response of the amp has something to do with it also (cap on the input, cap from grid to plate on triode)--that is why it seems okay. The bridge is a big hunk of metal, so it seems like it should be (in general). (Not real sure about the answer.)