Bedridden with back pain, I'm afraid. Had to cancel my trip to Burning Amp this week.
Should be recovered enough in a day or two to try the test. I'll be sure to post.
I assumed that there was a way to do that but I couldn't figure out what to multiply by! <grin>
It looks almost the same, but apparently using the Laplace source to multiply by s adds about 16 dB, and 90 degrees of phase angle, everywhere. See attached plot comparing them.
Attachments
Last edited:
Bedridden with back pain, I'm afraid. Had to cancel my trip to Burning Amp this week.
Should be recovered enough in a day or two to try the test. I'll be sure to post.
Wow, so sorry to hear that. I hope that you feel better soon!
I was unsure yesterday because I did calculations that indicated it was pretty significant, but I just discovered that F in the equations was in MHz, so I got results much larger than they should have been.
I'm curious though, is there any way other than using a really long wire that you could maximize skin effect so that it would dominate over inductance and resistance? That continuous 3db/oct curve could be useful.
Probably not.
Ha. Maybe use only the skin of the wires, for the wiring, such as a thin layer of metal on plastic. But the frequencies are so low that the effect would probably still be insignificant (if I had to guess), even for long lengths.
By the way, "significant" does not mean "greater than", or "dominating". It only means "not insignificant".
Ha. Maybe use only the skin of the wires, for the wiring, such as a thin layer of metal on plastic. But the frequencies are so low that the effect would probably still be insignificant (if I had to guess), even for long lengths.
By the way, "significant" does not mean "greater than", or "dominating". It only means "not insignificant".
So, like a sputtered ceramic cylinder, like some resistors.
But that would actually make the DC resistance closer to the skin resistance, so the dynamic effect of transitioning from DC resistance to skin resistance with F would not create a wide 3db/oct region. Very wide conductors seem to work best to exhibit skin effect, but then you get something so conductive that you can't measure the resistance or voltage drop.
I guess the solution to that would be to use a very resistive, very wide conductor. Maybe like a large graphite pencil stick? Spark plug cables? Perhaps Nichrome would be more convenient barring expense? Resistance needs to be large enough that it dominates inductance, so it would be easy to compare resistance to inductive reactance to figure out how much resistance is needed for a given frequency range.
But that would actually make the DC resistance closer to the skin resistance, so the dynamic effect of transitioning from DC resistance to skin resistance with F would not create a wide 3db/oct region. Very wide conductors seem to work best to exhibit skin effect, but then you get something so conductive that you can't measure the resistance or voltage drop.
I guess the solution to that would be to use a very resistive, very wide conductor. Maybe like a large graphite pencil stick? Spark plug cables? Perhaps Nichrome would be more convenient barring expense? Resistance needs to be large enough that it dominates inductance, so it would be easy to compare resistance to inductive reactance to figure out how much resistance is needed for a given frequency range.
Thanks Gootee, I'm getting better. Will be up and about today. Human backs are fragile things.
I'm curious about the skin effect. Is it supposed to be better in stranded cable? If it is, I can't figure out why the stranded CAT5 cables I have are so bandwidth limited compared to solid core. Ditto coax.
I'm curious about the skin effect. Is it supposed to be better in stranded cable? If it is, I can't figure out why the stranded CAT5 cables I have are so bandwidth limited compared to solid core. Ditto coax.
Pretty much. No matter what I did, the sims did not like the first few cycles because of the transient nature of the signals there and the math. That's when I switched over to hardware.
Remember, the mere thought of a difference between the two wiring methods encroaches on the theory of superposition. And the sim packages were designed using that. The primary question becomes, can a simulation package which is based on a mathematical model properly simulate an effect which does not? Is there a class of zero net energy signals which FFT's cannot properly see, such as the 2AB component of the double sine monowire dissipation? That is why I cautioned earlier about this, using the AC vs DC voltmeter as an example. Getting a zero result can mean there is nothing there, or the viewer you are using is insensitive to the effect. I caution against assuming the opposite, that it's there but the tools cannot see it.. I do realize that proponents of biwiring will jump on that conclusion, but this is the web after all, discussion must be open and inclusive.
I think a step function may be able to show a distinction at the tweeters, but again, I do not know how the math in the package is setup.
Consider what the effect truly is. It is the Lenz effect, where the conductivity of the current carrying conductor is attempting to exclude time varying magnetic field. Making the conductor a higher permeability will do this. The problem is the magnetic non linearity of nickel and iron.
A ferrite over the conductor, designed to increase the magnetic field. Placed judiciously such that the increased permeability enhances eddy currents in the conductor..perhaps ferrite over a flat conductor, as when they skin, the current goes to the edges.
jn
Bedridden with back pain, I'm afraid. Had to cancel my trip to Burning Amp this week.
Should be recovered enough in a day or two to try the test. I'll be sure to post.
Feel better. I look forward to further discussion.
jn
Good to hear you're improving, Pano ...
Skin effect and stranded work when the conductors are individually insulated - Litz wire - Wikipedia, the free encyclopedia.
Skin effect and stranded work when the conductors are individually insulated - Litz wire - Wikipedia, the free encyclopedia.
Oh yeah, duh! Where's that head slap icon?
Stranded non-isolated is awful, at least up in the >20MHz range. Under 100KHz I can't see it making much difference.
Actually, it can also impact inductors.
Here is measurement of 3 inductors. Note that the resistance increase of the air readings is self inflicted proximity losses. Note also it's 18 guage and 20 guage wire I'm talking about.
The cu readings is eddy losses due to a pcb plane.
On the second, the inductance as a function of frequency. Lenz effect drives the hf inductance down due to field exclusion.
jn
Attachments
Last edited:
I would assume so. The ones I tested constricted the current in the audio band, at 100K they are 10 to 20 ohms..at rf they'd be glowing orange.
jn
Yes, but I was talking about wire, primarily speaker wire, not an inductor or a capacitor or CPU.
Everything becomes a factor if you take an extreme enough case or some completely different kind of device/element. Speaker wire does matter if you decide to wire something with 28 guage wire. It's why you don't use that for it.
My car makes for a lousy turntable.
Part of the story is determining at what point the impact of something becomes audible, or disappears. If one's attention is not drawn to some aspect of the sound reproduction changing when something is altered you may never notice it happening - but once you do become aware it strikes you every time - the small stain in the carpet syndrome.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.