Jump to content

Talk:Coaxial cable

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia


RG-142

[edit]

RG-142 often used in aircraft, not test configuration only.

[1]

References

  1. ^ I am in an A & P electronics class. ~~~~

Attenuation vs. frequency

[edit]

Searching the Internet for various graphs of attenuation vs. frequency, I can see that there is a simple relationship. To the first approximation, presumably when resistive and radiation losses are low, the square of attenuation (measured in decibels) is proportional to the frequency. This is important, because companies selling cable usually display attenuation at only one frequency, and it would be helpful to figure out what it is (roughly) at the frequency of interest.

Perhaps someone familiar with academic literature on the subject could post the correct formula, quoted from some relevant manual. I don't have access to one, unfortunately. The formula seems to be: where is the frequency, and is the attenuation in decibels. Morycm (talk) 21:06, 10 February 2023 (UTC)[reply]

Here is a brief reply that is slightly WP:OR. There are nine possible frequency regimes (yes, nine) although typically we only concern ourselves with three of them which I will call low frequency (LF), intermediate frequency (IF) and high frequency (HF). LF includes the voice frequency range up until the point when skin effect is felt (perhaps 30 kHz). HF regime is where G/ωC << 1 and R/ωL << 1 or perhaps greater than 10 MHz. The IF regime is in-between. In the LF regime, attenuation (in dB) is approximately proportional to the square root of frequency. In the HF regime, attenuation is also approximately proportional to the square root of frequency, but with a higher proportionality constant than the LF regime. For the attenuation changes smoothly between the LF and HF regimes, it sometimes increases more rapidly than the square root of frequency. Finally, at a high enough frequency, dielectric loss may become dominant (not in an air core coax, of course) and you will see the attenuation increase proportionally to frequency.
Here is a representative plot based on published data and what I consider reasonable extrapolations at very low and very high frequencies. Note: I will collapse this figure to a thumb after the discussion is finished. At 1GHz, you can see the curve start to curve upward as dielectric loss starts to become significant. The transition points between the regimes depend on a number of things, so it is generally not reliable to extrapolate between low frequency attenuation and high frequency attenuation. However, the low frequency attenuation has a simple formula which is where is the loss in nepers. In this regime are roughly constant.
Typical Good Transmission Line Loss

Constant314 (talk) 22:06, 10 February 2023 (UTC)[reply]

On inductance of coax

[edit]

@Constant314

My edit had the purpose of clarifying the specific assumption, used for formula derivation

In this formula it's assumed that no magnetic field exists inside central conductor; which is the case for high frequency due to skin effect. However; if current distribution in the center conductor cross-section assumed even (low frequency range); the value of inductance changes to Brsbrs (talk) 17:04, 8 October 2023 (UTC)[reply]

Yes, I agree with you. I thought that I reverted my reversion. In general, at low frequency, L includes the internal inductance of the wire. However, that formula is the formula that applies at frequencies high enough for skin effect to be well developed. I was too fast on the revert button. So, my apologies for reverting your entirely reasonable edit. Constant314 (talk) 17:27, 8 October 2023 (UTC)[reply]
Actually, I think the formula for low frequency is Constant314 (talk) 17:50, 8 October 2023 (UTC)[reply]
Thanks! Yes, you are right, there was a typo in my formula. 1/4 should be outside of the logarithm Brsbrs (talk) 18:36, 8 October 2023 (UTC)[reply]
The inductance and capacitance per unit length should come out to give the right propagation velocity. But propagation velocity is only useful at higher frequencies. (Or extremely long cables.) Nice to get the low frequency limit right, but it doesn't come up all that often. Note also, that there are delay cables with a helical center conductor for increased inductance and lower velocity. Or you can figure how long it takes the signal to follow around the helix. Comes out pretty close that way, too. Gah4 (talk) 19:34, 8 October 2023 (UTC)[reply]

L and C

[edit]

In electronics, L and C are commonly inductance and capacitance. This article, confusingly, wants to use them as inductance/length and capacitance/length. The article could, near the top, explain this unusual usage. I believe some use l and c, but don't know of a WP:RS for that. Gah4 (talk) 08:30, 26 May 2024 (UTC)[reply]

I will check a few, but I believe upper case is the clear majority. Constant314 (talk) 10:00, 26 May 2024 (UTC)[reply]
The following use upper case letters: Brian C. Wadell, Johnson & Graham, Edward F. Vance, Gupta, Metzger, Miano & Maffucc1, Karakash, Albert A. Smith, Harrington, Hayt, Kraus, Jordan & Balman, Ramo & Whinnery & Van Duzer, Marshall & Skitek
The following use upper case script letters: Walter Weeks
The following use lower case letters: Clayton R. Paul, Skilling, Magnusson Constant314 (talk) 10:31, 26 May 2024 (UTC)[reply]
Thanks much. I am remembering from some reference from some years ago. It mostly matters if you need both near each other.
I do still believe that the article should state early that they are the per unit length variables, to avoid confusion. Gah4 (talk) 06:15, 27 May 2024 (UTC)[reply]