WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
Building, troubleshooting and testing of these amplifiers should only be
performed by someone who is thoroughly familiar with
the safety precautions around high voltages.
I am a relative novice but have successfully built two amps. I did not use star grounding but my understanding is that you are correct in your statement. Having said this (and I don't really want to open the pandora's box of star vs BUS grounding) when I first considered building, star grounding seemed a little more complicated and so I went for a bus earth wire. Overall there is something very satisfying about bending and soldering thick copper wire and I must admit that the results spoke fr themselves with absolutely no hum or buzz whatsoever.
That will usually work, but sometimes your unit will not play well with others and a nasty ground loop can result. The best trick I've seen is to isolate the signal and power supply star from the chassis (safety ground still needs to go to chassis), then connect the two grounds with a groundbreaker consisting of two diodes in parallel, each facing the opposite way, then bypassed with a small (10nF or so) capacitor.
Here's how I do. Bring all the audio circuit grounds to a single (star) point. Short,thick bus will do. Bring all the input and output signal grounds to another star point (I usually use a 12ga. bus across the back of the chassis). Power supply grounds may be daisy-chained any old way, least amount of wire.
Bring a single wire from the biggest PS sink ground point (the output stage cap ground terminal, in a standard power amp supply) to the audio circuit star. Bring another single wire from the input-output ground point to the audio circuit star.
The earth ground green wire from your AC in goes to the chassis. If you have several otherwise unconnected chassis plates like I do, connect them together. Drop your amp ground to the earth (chassis) ground from the input-output ground, not the audio or PS stars, and do this with a low-value, big wattage resistor, not directly (to prevent this ground loop). I use a 1Ω 10W cement block.
Using this topology, I get hum+noise in a PP power amp at between 20 and 40µV (0.02-0.04 millivolts), depending on the amount of tube rush in the particular tubes. This figure will, in fact, be dominated by the tube noise; if I put my frequency counter on it, it can't pick up a freek.
For all projects, I biuld each subsystem on its own circuit board, using the modified "dead bug" construction. Make certain that these boards aren't connected to the rest of the chassis, this is important, otherwise, you can get power supply "hash" in the signal paths. The ground planes of each circuit board connect to the DC neutral of the PS. Cathodes of the finals also connect there as well, each by its own lead.
Connect the AC safety ground (barely useable as an AF ground, utterly useless as an RF ground, due to impedance) directly to the metal chassis to keep any accidental AC or DC off the parts the user comes into contact with. Make a ground loop breaker from a high current, low voltage integrated bridge so that the internal diodes are connected "69 style", a 10R / 10W resistor bypassed by a 0u1 capacitor.
Almost dead quiet with no detectable trace of PS hash, and under 5mV (barely audible) 60Hz hum on the output.
Z1 is a MOV to protect against voltage spikes riding in on the AC mains. Excess voltage turns MOVs on, however, they have a nasty habit of staying on once the transient has triggered the MOV. That puts a near dead short across the AC mains, and would likely cause the MOV to ignite or explode. Not good, especially since these release toxic fumes. So you add a fuse ahead of the MOV to break the circuit should a MOV turn on, but not off. It's a "must have" if you use MOVs.