|
 |
Design Notes
Blue Jeans Cable LC-1 Low-Capacitance Audio Cable
The funny thing about unbalanced audio, as we discovered when we entered the cable business, is that there's very little cable on the market which is expressly designed for use as unbalanced audio interconnect. Unbalanced audio doesn't see a lot of professional use, so many broadcast professionals often simply use whatever balanced audio cable they have on hand, and leave one of the conductors unconnected; meanwhile, though some manufacturers do make cable specific to unbalanced audio, it's often not really well-built for the application and is intended more for building "throw-in" style economy cables--poorly shielded, and high in capacitance.
These two attributes, shielding and capacitance, are the most important factors in unbalanced audio cable quality. Shielding is important, of course, because it keeps out externally-induced noise, and because unbalanced audio, unlike balanced audio, can't take advantage of common-mode noise rejection. Capacitance contributes to high-frequency rolloff, so the lower the capacitance of the cable, the flatter the frequency response in any given application (how flat will depend on the device impedances as well as the capacitance, so it's not possible to generate a one-size-fits-all frequency response chart; but in every case, the lower the capacitance, the flatter that curve will be). Capacitance, like many cable attributes, is a per-foot characteristic; while high capacitance won't ordinarily make a significant difference in short runs, it becomes an increasing problem with longer runs.
When we entered the cable business, the best cables we could find for unbalanced audio use were cables which had actually been designed and built for video. Why? Well, that's because video cables are coaxial, which is the right geometry for an unbalanced audio cable; because video cables are typically well shielded; and because video cables, being designed to a 75 ohm characteristic impedance, are relatively low in capacitance, ranging from about 16 pF/ft for an HDPE-foamed dielectric precision video cable (e.g. Belden 1694A) up to about 21 pF/ft for a solid PE dielectric cable (e.g. Belden 8281 or Canare LV-77S).
In testing various available cables against one another, we found that double-braid shield designs, where one braid shield is laid directly on top of another, provided the best low-frequency noise rejection. As it happens, conveniently, double-braid shielding is usually found on high-flexibility video cables, so the best noise rejection and the best flexibility for cable size were typically combined in the same products. As a result, we recommended Belden 1505F for general analog audio use, and Canare LV-77S (higher capacitance, but slightly better low-frequency noise rejection) for subwoofer use. These were, and remain, excellent products for this application; but we felt we could build something a bit better.
Video cables are constrained in their design by the need to maintain a characteristic impedance of 75 ohms. As we've mentioned, that generally locks capacitance in at 16 to 21 pF/ft. But analog audio cables don't need to maintain any particular characteristic impedance. Unbalanced audio cable is usually run from a low-impedance output into a high-impedance input; it's not an impedance-matched system like video, and it doesn't need to be, because the wavelengths of analog audio signals are so long that--barring audio cables that are miles long--impedance just doesn't matter. Because analog audio isn't an impedance-matched system, we can make the characteristic impedance of an analog audio cable come out to any value at all, without any adverse consequence.
By the way: we often hear from people on internet discussion boards, or in e-mails inquiring about our products, that "audio cable is supposed to be 50 ohms." There is indeed a lot of 50 ohm coaxial cable in the world, and no doubt some of it has been used for analog audio; but there is not now, and has never been, any standard impedance spec for unbalanced analog audio cable. 50 ohm cable isn't a good choice for analog audio. That's not because of the impedance, which doesn't matter at all, but because of the capacitance, which is quite high in 50 ohm cables (typically 25 to 31 pF/ft). We're not sure what the origin of the "50 ohm audio cable" myth is, but it doesn't seem to want to die.
Since impedance isn't a factor, and capacitance is, we aren't bound by the practical limits of capacitance for a 75 ohm cable. We are, however, bound by various other practical limits. If a cable gets too thick, it becomes inflexible, and incompatible with reasonably-sized connectors. If the center conductor is too thin, it becomes fragile, and becomes difficult to reliably attach to connectors. Capacitance, in a video cable, is a function of three things: outer diameter of the center conductor, inner diameter of the shield, and the dielectric constant of the material between them. Within reasonable limits, we wanted to optimize all of these factors to reduce capacitance: small center conductor, large shield diameter, and low-density foamed dielectric.
Meanwhile, we wanted to maintain a very high shield effectiveness, particularly at low frequencies. To do that, we chose the shield design used on two large high-flex cables: Belden 8281F and Canare LV-77S. Both of these cables use a heavy double-braid shield, and the Canare LV-77S tested best for hum rejection in our comparison of a number of well-shielded cables. The 8281F has a tinned shield, while the LV-77S has a bare copper shield. There's no significant difference in performance between the two; however, we decided to go with the bare copper braid because it's marginally more conductive than tinned copper.
The dimensions of the cable were, ultimately, dictated by practical connectorization; we went with a 25 AWG solid copper center conductor because that was as small as we felt we could make it while maintaining (1) the ability to firmly crimp a center pin to it and (2) adequate strength for durability under pulling force. The shield dimensions of the LV-77S/8281F configuration, meanwhile, not only were as large as we wanted the cable to get from a flexibility standpoint, but also were compatible with crimping to the largest Canare RCA connector bodies, which are made for LV-77S.
For the dielectric, we wanted a material as foamy and soft as possible; solid polyethylene, like that in LV-77S and 8281F, was not good because its dielectric constant is relatively high and would keep capacitance up. Air, of course, is ideal as a dielectric, but one can't build a coax using only air; foamed polyethylene is the answer, because it combines the mechanical stability of PE with the best dielectric: air. We went with Belden's low-density PE foam.
The result is a cable which combines, to the extent practical, the best possible attributes of an analog audio line-level cable. Its capacitance is extremely low, at 12.2 pF/ft, while its shielding is extremely effective at rejecting audio-frequency interference. Meanwhile, though it has the outer dimensions and appearance of 8281F, it is more flexible due to the extremely soft dielectric and smaller center conductor. Theoretically, one could improve further on both shielding effectiveness (by inserting more layers of shielding) and capacitance (by increasing the size of the dielectric, shields and jacket)--but either would require making the cable impractically large. We feel that LC-1 represents the best combination of electrical characteristics and usability of any analog audio cable on the market today.
| LC-1 Audio Cable Specs: |
| Physical Characteristics: |
| Center Conductor - Solid Bare Copper, 25 AWG |
| Dielectric - Nitrogen-Injected Low-Density Polyethylene |
| Shield - Braid/Braid, 98% coverage, bare copper |
| Outer diameter - .305 inch |
| Certifications: |
| UL Listing: Yes |
| NEC Rating: CM (Communications rated; suitable for residential and commercial in-wall installation) |
| Electrical Characteristics: |
| Capacitance, conductor to shield: 12.2 pF/ft |
| Resistance, center conductor: 34 ohms/1000 feet |
| Resistance, shield: 1.7 ohms/1000 feet |