I often see wonderful builds with best-of-parts (that is, the builders didn't spare anything) but the signal-wires, mostly twisted pairs, going all the way back-to-front from the RCA-plug to the selector-switch (elma 04, seiden, etc.), over to a luscious volume-control of sorts, and from there all the way back to the board.
Aren't wire-length supposed to be made the shortest possible, or is this more an esthetic problem? My head is often smoking from trying to figure out the really shortest way from input-connector to selector-switch to volume-control to the line-stage... Obsolete?
david
Aren't wire-length supposed to be made the shortest possible, or is this more an esthetic problem? My head is often smoking from trying to figure out the really shortest way from input-connector to selector-switch to volume-control to the line-stage... Obsolete?
david
It's best for all those interconnections and controls to be on a pcb, but regardless,
those connections should all be short and direct.
those connections should all be short and direct.
In general, shorter rather than longer is better. But there can be overriding reasons for not doing so. (clearance from transformers)(access for maintenance)(dressing the wires near the chassis edge). It's way down the list of importance.
* * * * * * * *
In classical music, concert hall recording the mic cables are sometimes 100 meters (300 feet) long. So some inches in a chassis won't make much difference.
* * * * * * * *
In classical music, concert hall recording the mic cables are sometimes 100 meters (300 feet) long. So some inches in a chassis won't make much difference.
Last edited:
I have tried both short and long signal cables in preamps, if you only do it right from the beginning, there is no difference, but if you have short cables, it is preferable.
Thanks, K.A.B
„Doing it Right“ would mean
- twisting
- away from PSU/AC
- shielding if critical by layout or signal-strength
- proper diameters
- along chassis wherever possible
?
Thank you again
Keeping the loop area minimal is way more important then a few centimeters of wire more or less. For that keep in mind that current goes in a loop from the source back to the source. The audiophile idea of "signal path" as a line from source to end is misleading. Also insofar as the power supply is always part of that loop.
So grounding and what's going on along a groundbus with regards to charging pulses from capacitors etc. is worth considering when doing the layout.
Away from PSU/AC is easily done with an external power supply. Placing the power transformer(s), rectifier(s) and first filtering stage away from the signal part can help.
So grounding and what's going on along a groundbus with regards to charging pulses from capacitors etc. is worth considering when doing the layout.
Away from PSU/AC is easily done with an external power supply. Placing the power transformer(s), rectifier(s) and first filtering stage away from the signal part can help.
Aah, this is interesting!
I‘m still feeling a bit overwhelmed by the concepts of grounding, so I may need a while to digest...
🙂
I‘m still feeling a bit overwhelmed by the concepts of grounding, so I may need a while to digest...
🙂
The word 'ground' has many meaning and even more incorrect meaning!
Some of the correct ones:
a] common reference point
b] Safety Ground/Protective Earth
c] chassis
d] Planet Earth
e] DC supply common
f] AC supply secondary common
g] audio circuit common
h] digital circuit common
While there is often continuity from one to another, each has it's own function.
Some of the correct ones:
a] common reference point
b] Safety Ground/Protective Earth
c] chassis
d] Planet Earth
e] DC supply common
f] AC supply secondary common
g] audio circuit common
h] digital circuit common
While there is often continuity from one to another, each has it's own function.
Which is (one of several reasons) why I find it complicated. AFAICS it’s not, but until it’s understood it is... („it“ being, well, „ground“ as a general expression)
Kevin is right...
For me, I distinguish "ground" as two entities. First, the safety conductor used to keep exposed metal surfaces from rising above 50 volts in the event a line conductor accidentally shorts to chassis.
For signals, it is the reference point for the signal.
The NEC has specific rules for "bonding conductors" used for safety. The NEC does not care about analog signals or their integrity. (Well, they might, but it is not possible to cover all possibilities in the codebook.)
For the absolute best understanding of grounding and signal integrity, google Tom Van Doren. He is excellent!!!
Granted, when he gave a class here, he erred in two small things.. But neither were show stoppers, so I bailed on raising my hand and going "ooh mitta kottah, mitta kottah!!!
Told him offline, no problems.
Jn
For me, I distinguish "ground" as two entities. First, the safety conductor used to keep exposed metal surfaces from rising above 50 volts in the event a line conductor accidentally shorts to chassis.
For signals, it is the reference point for the signal.
The NEC has specific rules for "bonding conductors" used for safety. The NEC does not care about analog signals or their integrity. (Well, they might, but it is not possible to cover all possibilities in the codebook.)
For the absolute best understanding of grounding and signal integrity, google Tom Van Doren. He is excellent!!!
Granted, when he gave a class here, he erred in two small things.. But neither were show stoppers, so I bailed on raising my hand and going "ooh mitta kottah, mitta kottah!!!
Told him offline, no problems.
Jn
Cabbage would be on my list, too jneutron! 😱
The thing that has me scratching my head, is when I see low-level signal runs within a chassis implemented in twisted pair. Must just be having seen kilos of old valve gear -- much higher impedance, granted -- using shielded runs for all the sub-line-level stuff.
Cheers
The thing that has me scratching my head, is when I see low-level signal runs within a chassis implemented in twisted pair. Must just be having seen kilos of old valve gear -- much higher impedance, granted -- using shielded runs for all the sub-line-level stuff.
Cheers
Low impedance nodes are of course much less sensitive to noise.
Head scratching: Looking at various gear, old and new at first often seems baffling. Like valve oscilloscopes that run bunches of wires in parralel tied together with line mans' knots and still working up into the MHz region
Erm. I thought that parallel cables were more likely to pick up „noise“.
So what do you mean by „coupling“ if this was to be welcome? Or, do you mean that a pair (core/shield, hot/ground) would/should be coupling or better yet, twisted?
So what do you mean by „coupling“ if this was to be welcome? Or, do you mean that a pair (core/shield, hot/ground) would/should be coupling or better yet, twisted?
Last edited:
Flow and return ideally should be close coupled to reduce the loop area between them which is like an aerial. Sometimes it's useful to forget the term "ground" entirely 😉
A simple question but each different situation, makes for many complicated answers.
Not that a common 4 pair Cat5 cable, can handle 4 different line level signals without problems
One of my many challenges is to find the proper term. Another one is obviously to know what for I should find it.
(Simply put, I „sometimes“ understand the problem, and sometimes get lost in the nomenclature )