Star grounding problem

[Dave H]

Thanks! Now I can try to decipher what all the fuss was about...

Same here, it looks like an interesting 'discussion'. How did I miss it first time around? And where did SGM go? I miss him.

[g1]

That's borderline infraction material, mister!

[big_teee]

Same here, it looks like an interesting 'discussion'. How did I miss it first time around? And where did SGM go? I miss him.

Be Careful what you wish for!
T

[Chuck H]

I hope SGM is doing well wherever he is that isn't here.

[Jazz P Bass]

He is still at it.
soundguruman - User Profile | Harmony Central

[jazbo8]

I read the thread with the accompaning figures supplied by dai h., so I guess Alan had an issue with the star grounding scheme and made a big fuss about why one should always use the chassis as local ground points. Yup...

[Malcolm Irving]

I'm just wondering if one reason that 'chassis as a ground plane' is not usually favoured for guitar amps, is that there is a big power transformer bolted on to it. The leakage flux from the transformer will result in 50 or 60Hz eddy currents in the chassis. I think the eddy current paths could cover a fairly wide area and could impinge on signal return paths. The whole issue is quite complex - so I am happy to be corrected if I am wrong.

[jazbo8]

I'm just wondering if one reason that 'chassis as a ground plane' is not usually favoured for guitar amps, is that there is a big power transformer bolted on to it. The leakage flux from the transformer will result in 50 or 60Hz eddy currents in the chassis. I think the eddy current paths could cover a fairly wide area and could impinge on signal return paths. The whole issue is quite complex - so I am happy to be corrected if I am wrong.

Chassis as a ground plane has been used by many manufucturers in the past, and could indeed work very well, e.g., the old Fenders. But for a noisey PT, that is always going to an issue whichever grounding method you choose to use.

R.G., Merlin and others suggested that the star grounding scheme could be used as a simple, albeit brute force way to get good results, since it takes away the guesswork on the ground currents. While Alan insisted that "his way" was better, and went on to great length to discount the star grounding scheme. So different strokes for different folks...

[dai h.]

(the way I see it) if you use the chassis, it's more likely to be a problem when connecting other 3-prong gear (the ground scheme is messed up). Do techs change the grounding scheme typically when converting an old 2-prong plug amp to 3?

[Tooboob]

Found this thread in the index and started reading. Didn't realize it was from 2012. It was good to see SGM and Stevie C. back at it.

Anyway, meandering through the discussion, it seems like a big point is always missed about grounding, and I'm going to state it again.

When the FCC said that our computers couldn't radiate like refrigerators, blenders and vacuum cleaners, designers like myself studied up with the folks who appeared "in the know". I was talking to an expensive guy, and I said "Wait a minute. There a big loop for everything. the +5V comes out of the supply, runs way over to a logic IC, runs from the power pin to the ground pin, then way back to the supply!"

He said something like "No." The HF AC current runs from the closest supply cap, through the IC, and back to the closest supply cap. Of course, with SOCs running in GHz range, he's wrong, since at those frequencies, caps, resistors, etc. take on multiple personalities, so we have to take extraordinary measures, but it's as true for tube amps as it was in 1987.

The supply rectifier bridge conducts short spikes when the AC out of the PT secondary actually gets higher than the cap voltage. The cap needs to be right next to the rectifier, which needs to be right next to the PT secondary. This makes a small loop. If you're implementing star grounding, you should wire the grounds tight, and run one wire to your star ground. There will be transformer leads of some length, so twist them together. If you want to mess it up, run long separate wires to your star ground.

If you have a load that injects supply noise, put a cap next to it. Again, one wire to the star ground from the circuit. Long wires carrying AC create AC offset voltages, and ground is just a concept. Caps reduce the AC content by providing the AC energy locally.

So don't think that the length and number of connections on the ground wiring are the whole problem, solved by faithful star grounding. The cap locations and grounding are a big deal too, and even a broken SGM is right twice a day. R.G. was right that there are capacitive and magnetic sources of hum too, and tight AC loops also help with these problems. Star grounding, religiously applied without being the electrons, can yield large coupled ground loops.

You may have heard this approach called galactic grounding, which is obviously much cooler than star grounding.