Ground Loops and interconnecting pieces of equipment

[rubberlips]

G'day guys,

I have a question about ground loops and wiring. How do people wire up their project amps, preamps, effects etc from the power supplies. In respect to the 0V or ground, do you

connect the 0V transformer to the regulated power supply 0V, then loop to the rest of the boards and interconnect the shields of the input signal to each board and then ground the last connection to the chassis

OR
connect the 0V of the transformer and power supply board to the chassis and then either
* lift one shield wire off  to the proceeding board or
* lift one shield wire off  to the proceeding board and connect a small resistor in series with a small cap to ground as a low pass filter??

I know it's a big topic, but which methods are most effective in stopping ground loops and avoiding RF problems??

Pete

[zachary vex]

for filament transformers, take a look at older fender schematics.  in some cases, they tie the two hot leads of the 6.3 V transformer to the ends of a pot, and then tie the wiper to chassis.  take a look at the "hum" pot in those examples.

[R.G.]

I have a question about ground loops and wiring. How do people wire up their project amps, preamps, effects etc from the power supplies.

First, go read "The Spyder Power Supply" at GEO. The Spyder was intended specifically to defeat ground loops in effects setups.

There are two problems with "ground":
1) any complete loops of grounding lets the loop pick up radiated interference at power line frequency and at RF
2) If there are shared paths to your chosen reference voltage location, the shared path generates a voltage for each of the currents because it's not a perfect conductor, and this voltage appears at each input on the way as a signal voltage.

I guess the biggest problem with grounds is that you have to be concerned with where the currents are flowing, in which wire, and how much, and you have to treat the wires as resistors, not as 0-voltage-drop wires.

In respect to the 0V or ground, do you

connect the 0V transformer to the regulated power supply 0V, then loop to the rest of the boards and interconnect the shields of the input signal to each board and then ground the last connection to the chassis

OR
connect the 0V of the transformer and power supply board to the chassis and then either
* lift one shield wire off to the proceeding board or
* lift one shield wire off to the proceeding board and connect a small resistor in series with a small cap to ground as a low pass filter??

The real answers are
a) use balanced signal wires with single-end shields if you can. Unfortunately you can't unless you build all your own equipment because the single conductor coax is too deeply ingrained to be tossed.
b) Isolate safety ground (wall socket green wire) from signal ground at every place except ONE AND ONLY ONE. Ideally safety ground connects to the chassis at the AC power entry and then one and only one wire goes to the single point where the signal circuits are referenced. No other connections of signal ground to chassis, including jacks. Only works on single box equipment.
c) For multiple boards inside a single box, you isolate chassis/safety ground on one wire, then take a wire from your power supply ground to each board separately so that no board trickles its waste current down any other board's ground wire.
d) If you build your own power supplies, the first filter capacitor is super special. The wires from the rectifiers and transformer center tap (if that CT is used) carry BIG pulses of current to it at the power line frequency. You do not want those pulses causing voltage drops in the signal ground of your circuits. So the transformer/rectifier connections to the first filter cap go the cap, and signal ground wires come off the filter cap NOT OFF THE RECTIFIERS OR TRANSFORMER. Any other connection gives you some degree of rectifier buzzzzzzzzz.

Within a box, shield only signal runs that are high impedance, like the inputs to FETS and tubes that run to controls on the panel. Ground this shield at one end only.

Between multiple boxes, either isolate box safety ground from signal ground with a 10 ohm resistor shunted by a 0.01uF 1kv cap and use shielded cable between boxes, or hard ground all the chassis to safety ground and use cables with shields connected at one end only.

I know it's a big topic, but which methods are most effective in stopping ground loops and avoiding RF problems??

Also see "Star Grounding a Tube Amp" at GEO - http://geofex.com/Article_Folders/stargnd/stargnd.htm

There are tons more of considerations. I guess I have to write a "Technology of Grounds" article.

[guitarhacknoise]

There are tons more of considerations. I guess I have to write a "Technology of Grounds" article.

 There is alot of great info in these few paragraphs!
A whole paper on it would rule!

................Of course, a little time for yourself would probably be more fun.
Thanks.
-matthias

[Rob Strand]

For bipolar supply arrangements, the *tranformer* 0V should *never* connect anywhere but to the filter caps first, and nothing should connect to the track that runs *between* the transformer 0V and the two caps.

The example ofa tube amp heater system is not a good example to follow at all - it is only valid for tube amps and even then only when the heaters are isolated (as they usually are).

For a good example of how to do grounding on solid state amplifier see:

http://sound.westhost.com/project27.htm

That articles shows all the elements of good grounding.

Note:

- Don't worry about, and don't add, the 0.1 ohm resistor in the speaker 0V.  This is part of the design.  In a normal amplifier just run that wire through.
- The article recommends isolated jacks.
- Some power amps have 10ohm ground loop isolation on the input.  It is still valid to connect these as that article shows.

[rubberlips]

Thanks for the info guys,

Now that i know what to look out for I can see ground loops gallore. What I thought of ground wiring was connect everything to one point (basically), but in reality it's more following the path of ground from the power supply back to the chassis.

One other question while I was looking at it:
On this http://sound.westhost.com/project27.htm
there's a 10ohm resistor to the preamp from the ground point. Would this really be needed if the amps is connected as in diagram 1B, as the ground would be going through the power amp ground through the input of the power amp, through the pre in/out and then to the preamp board?

Pete

[Rob Strand]

Would this really be needed if the amps is connected

It's largely precautionary: It prevents the preamp totally losing power when you disconnect the ground to the power amp (perhaps when debugging).  You could waste hours debugging the preamp only to find the power amp input connection you removed was supplying power.....  Also it provides a path for some current i fthe power amp has a resistance in the ground.

[rubberlips]

thanks everyone for the tips,
Finally I've seperated the power amp and preamps, no ground loops now on the amp. Closed that unit up and added the +/- 15V to an insulated stereo socket to power any preamp I connect.
Now onto fine tuning other preamps and cranking some motorhead out :twisted:

Pete