I agree with you that different components will require different techniques in ground, and that circuit board layout is different than point to point wiring. But we still use both and must deal with both. Even when circuit boards are used, wire is often also used to connect boards together in some fashion, deliver power, transfer signal, etc. How you ground these must always be considered as a primary aspect of the layout.
I think one of the main differences between tubes and silicon devices is simply the lower potential difference you see with more modern components, and the way we use them in circuit, current amplifier as opposed to voltage amplifier for example. That difference asside, Ohms law still applies. The more resistance your ground plane has, the more effect it will have on your circuit noise figure. Star grounding is not the answer in all cases, certainly not in all circuit board layouts, but some permutation of it will work in a lot of cases. For example, grounds from three circuit boards connected in series, i.e. daisy chained together, will almost always cause noise. Take the same circuit boards an connect the grounds in parallel, i.e. star ground, and they will be less noisey.
The layout of the circuit board itself usually does not lend itself easily to star, or single point grounding, and is usually better done with larger grounding traces, well placed, or a ground plane. On this I think we agree. My main point is that grounding technique seems to be the single biggest reason for trouble in most of the DIY projects.
If we want repassed tube theory to the transistors, transistor
amps will have worse parameters. Because there is not theory for
transistors, but transistors with tube theory reincarnated.
To make this theory usable many passive stuff in along the transistors must be used and this stuff take parameters of transistors down.
Transistors evolved from tubes so naturally some of the same things apply to each. They are both amplifiers They both must have passive components to function in a given circuit. These passive components do not necessarily make the transistor parameters worse. In some cases they make them better, but in every case, they are necessary to make a functioning circuit.
The one common thread in all of these circuits though, is they must all have some way to return current to a common point, usually ground. I think it makes much more difference in the noise generated in a good circuit, what your ground scheme is, than what components you use. The people here who design circuit boards for the projects seem to have an excellent grasp of this. The few problems that do crop up though are usually related to some error in grounding.
I just do not believe that any grounding technique is antiquated and should be discounted. I may have misunderstood you, but I think you suggest that this is the case. My assertion is that many of the oldest techniques still apply to modern situations. Ohms law still applies in every case for ground technique and impropper grounding is the number one downfall of our group.
from the ssl project page)
