Its a technique used a lot for other communication systems and has been for a number of years. The connection also has to be made to the chassis where the connector comes in, long pig tails to some distant star ground is useless for RF screening. It just good EMC practice that the likes of H. Ott have been preaching for years. The same applies to RCA connections, if you have EMC problems you need to connect the screen in a similar manor where the cable enters the equipment.
If this was done then cable manufacturers could stop putting arrows on their cables and confusing some poor audiophiles into believing cables are directional carrying analogue audio signals.
Then all we have to do is convince them that low level analogue layout is better with ground planes than the spiders leg style atrocious ground connections some like, all going to a central point and looking like a lovely wide band antenna to that nasty air borne RFI.
If this was done then cable manufacturers could stop putting arrows on their cables and confusing some poor audiophiles into believing cables are directional carrying analogue audio signals.
Then all we have to do is convince them that low level analogue layout is better with ground planes than the spiders leg style atrocious ground connections some like, all going to a central point and looking like a lovely wide band antenna to that nasty air borne RFI.
Then all we have to do is convince them that low level analogue layout is better with ground planes than the spiders leg style atrocious ground connections some like, all going to a central point and looking like a lovely wide band antenna to that nasty air borne RFI.
I'm building a new vacuum valve amplifier, and have been thinking about shielding the "ground returns". Point-to-point wiring is inherently EMI prone, and "ground returns" don't usually lend themselves to twisted balanced pairs.
Thanks,
Chris
Not much info there and I cant find the Audio express article...
But as with solder joints if these things were real phenomena to be concerned with other fields of electronic design would have picked up on them and there would be a whole range of articles regarding directivity in metal based electrical signal paths
I would be interested in the solder joint stuff, I wont say they are not problematic, I have worked on a big project regarding lead free component finishes, solder joints and reliability for military applications and don't remember any issues with distortions with a well made joint with the right finishes and intermetallic layers present in the joint (confirmed by micro sectioning etc), bad joints being a different ball game obviously.
But with cable directivity I think its definitely Myth Busted....
The AX article was not posted online. You'd actually have to buy a copy!
The rest of the JJ thread is here:A very interesting article in Audio Xpress | Stereophile.com
Note how it dissolves into the usual folks with the usual lines.
Scott last I looked Sam used AC. It is the DC across the capacitor that contributes to the V dc/dt noise. The higher the voltage obviously the greater the noise. Orientation to the axis of motion also has impact!
You can read the figures any way you want to. I do not plan to explain anymore about the directional issue here. There are some more experiments I may eventually run as I now have a working theory of what makes cable have different distortion depending on the direction of power flow.
If all you want to apply is circuit theory then that does not include the issues at play.
Sam was only measuring distortion. The equivalent SPL of blunt force trauma on a cap is rather high, in practice I have used pretty badly microphonic inputs with a little care.
Once DC is blocked filters and frequency shaping networks don't see DC anyway. That is a state variable oscillator can be designed with no DC on any caps. There are sidebands on the fundamental but again in practice there are many remarkable results out there.
I can't wait. BTW be careful with "direction of power flow" in this context (AC audio signals) you will soon add confusion.
Last edited:
A general comment on this wire directionality issue. A lot of folks have a knee-jerk reaction that somehow this is always a claim that the symmetry principle in physics is violated. A cable with one bad end or obviously the shield connected at only one end can be "directional" in the context of a specific application.
Capacitors used as frequency determining elements are of course under different stresses than filter or DC blocking caps.
The actual cable issue under test is how very low currents vary in propagation than ones in which the current is high enough that the signal charges interact in addition to the ones in the conductor. There are some interesting papers on this but I will not cite them as those who can't find them will misread them one way or the other. Those who can find them probably are too smart to post here.
Odd. Nobody I know has experienced anything like that..(edit: in a wire, of course. In a vacuum beampipe, the whole name of the game is using alternating gradient multipoles to keep the beam focussed, otherwise it will disperse into the vacuum chamber walls.)
Of course, in wires we only run current densities out to 2000 amperes per mm2.
I guess your talking audio, where the numbers get really high...
Gee thanks.
jn
Last edited:
Odd. Nobody I know has experienced anything like that..(edit: in a wire, of course. In a vacuum beampipe, the whole name of the game is using alternating gradient multipoles to keep the beam focussed, otherwise it will disperse into the vacuum chamber walls.)
Of course, in wires we only run current densities out to 2000 amperes per mm2.
I guess your talking audio, where the numbers get really high...
Gee thanks.
jn
I take it your native language is not English!
The issue is at LOW currents not at high currents. High current defined in this case as where the charges do interact. That might just happen at densities of 2000a/mm^2. (?)Try 1 nanoamp/10cm^3 what happens then?
I take it your native language is not English!
If you look carefully, you will see I quoted you...
There you go again with this "high current where charged do interact". As I said, and I repeat, nobody I know has experienced that in a metallic conductor. Of course, who would I know??
Interesting use of units there Ed.. Current per volume doesn't mean too much to me..
Perhaps you meant current per unit area?
jn
If we can't find them how do we misread them? Probably like the folks that read papers on nano-scale cryogentic experiments and immediately extrapolate the results to macroscopic devices that are "just around the corner".
I never knew that copper would "Molecular in level, quite bit like the same way that copper, over time, under heavy electrical loading, will crystallize and loose up to 2% of it's weight. Purely from the orbital knock out effect of the electrical loading."
I wonder if that is due to mass loss or changes in gravity? ;}
I wonder if that is due to mass loss or changes in gravity? ;}
I believe it may be a misappropriation of the effect known as electro-migration.
It started to raise it's ugly head on the aluminum interconnects of VLSI chips as the cell density kept going up but the current didn't go down proportionally. Since the aluminum traces on a silicon chip are quite well heatsinked to the silicon, very high current densities are achieved.
The problem is, because of the high current densities, the electrons would accelerate so much in between collisions, that the momentum lost at impact would actually be high enough to dislodge the atoms and drag them in the direction of electron flow. It created hillocks of aluminum, and eventually caused failure of the chip. Googling the term will get to electron microscope photos of this.
Copper on silicon was a solution to that problem (no doubt short term as circuits shrink even more.
I came across the same issue with the LHC splices, as the current densities they forced across a tin/silver solder joint exceeded the current densities which can push electromigration, leading to joint failure.
jn
ps..course, I could be wrong and it's really gravity..
Last edited:
If you look carefully, you will see I quoted you...
There you go again with this "high current where charged do interact". As I said, and I repeat, nobody I know has experienced that in a metallic conductor. Of course, who would I know??
Interesting use of units there Ed.. Current per volume doesn't mean too much to me..
Perhaps you meant current per unit area?
jn
Compared to nano amps, micro amps are high current!
If I have 1 nano amp current flow how many electrons per second is that? (10^-9 x 6.24^18 q/S) What is the speed of propagation in a copper wire? (2.4^8 M/S ish) If we calculated q/M (6.24^9/2.4^8 q/S x S/M) for a given wire length. We can then add cross section and get charge per volume.
OK a complex problem that can be solved very simply if you look at it sideways.
On an analog clock at what time do the hour, minute and second hand divide the face into three equal parts? Note that at 12:20:40 the hour and minute hands have moved so that it is not the correct answer.
Now what is interesting is that this problem can be solved without the use of trigonometry.
George I am posing theories and willing to accept comments based on understanding the issues. I think the issue is actually the use of language.
On an analog clock at what time do the hour, minute and second hand divide the face into three equal parts? Note that at 12:20:40 the hour and minute hands have moved so that it is not the correct answer.
Now what is interesting is that this problem can be solved without the use of trigonometry.
George I am posing theories and willing to accept comments based on understanding the issues. I think the issue is actually the use of language.
Last edited:
Wrong answer. Here these guys have nice review material I've used it before, http://hyperphysics.phy-astr.gsu.edu/hbase/electric/ohmmic.html
EDIT - I just realized how confused you are on this, best to just let you carry on.
Last edited:
That's a nice problem, I solved a similar one using primes in high shool. Prove that the three hands are coincident at only noon (12). Of course these problems are different on a military analog clock.
- Status
- Not open for further replies.