Posted August 24 2003 - 10:17 AM
The purpose of this post is to take a look at cables that can be used as component cables from the point of view of bandwidth. A discussion of this will also include some additional bandwidth comments regarding associated devices.
Often the topic of what component cable should I buy is asked. The likely responses will be brands or particular vendors. I find this information generally useless and while everyone may indeed be correct, it doesn't address the fundamental question of what are the requirements needed to successfully pass a signal from one area to another such that it can be effectively used. Note that this does not mean which cable has the least amount of attenuation. While buying such a cable in not a bad thing, often in cases where the lengths are very nominal, a few meters, it's completely unnecessary and results in a gross overspecification that doesn't result in a visible difference. In the situations where there may be a visible difference such as in long lengths, often a slight adjustment on the display device restores everything to the original level of acceptability. When that's done, no visible difference is seen. We've all seen TV's in stores and it's nothing short of amazing how different many brands can look. Too much red, not sharp enough, contrast is off, etc. Most of us go to the trouble of adjusting our sets using calibration disks to provide references and targets. Some even go so far as to enlist the services of a calibration professional. It's quite reasonable to expect that given a long length of cable one may have to adjust their TV sets. If you find this disturbing then just buy the widest bandwidth, heaviest shielded cable you can. Personally, periodic calibration after making changes I consider just the way things are. In the scientific fields it's Standard Operating Practice and given that the differences are generally small, no one makes a big thing about it. However, audio, if nothing, is about raising trivialities to the level capital offences.
Bandwidth is generally defined as the point at which a signal of a particular frequency is down 3 dB. Being down 3 dB means that we've lost 1/2 of our signal. Generally, but not always, this is taken at a distance of 100 feet. Hence, when a vendor specifies this, it'll generally read something like "bandwidth of 750 MHz @100 feet"
. When a vendor doesn't specifiy bandwidth but instead specifies just the number you're at a loss as to what they really mean. You should contact the vendor to get clarification.
So what are the problems associated with insufficient bandwidth? Well if the signal you're pushing through has a bandwidth of 5 MHz and some device in the middle has a bandwidth of 2 MHz, then most certainly that device will severely attenuate or chop out the signal above 3 MHz. You'll be losing information that may well not be compensated for by adjusting controls on the TV. Fine detail will be severely compromised. Kind of like clipping if you will. Receivers or other processing devices that allow video switching should have a bandwidth specification. Depending upon what's plugged into the device, the bandwidth may be adequate, marginal, or simply deficient. I take the manufacturer's specification to mean that it will attenuate the value by 3 dB. Various values that I've seen here are 50, 100, 200 MHz and there are no doubt others.
So what bandwidth is necessary in order to pass these signals? Our eminently capable Bob McElfresh has posted the following information before.
| But if you have a HD source (Sat or Cable box) and a HDTV - this needs coax/connectors designed for the higher bandwidth signals: |
Standard Component Video: 4 Mhz max
Progressive Scan Video: 13 Mhz max
1080 HD Video: 35 Mhz max
This is sufficiently accurate for my purposes and if you wish to take issue and toss in some fractions of a dB corrections then by all means you've been splitting hairs for far too long. The rule of thumb for video bandwidth is that it should be three times the source bandwidth. So, in the case of an XGA computer that has a video bandwidth of about 100 MHz you'd need a cable capable of supporting 300 MHz in order to maintain signal integrity. THIS DOES NOT MEAN AT 100 FEET. IT MEANS THAT THE BANDWIDTH FOR THE GIVEN LENGTH YOU'RE USING NEEDS TO BE 300 MHz!!!
It is extraordinarily difficult at times to obtain comprehensive bandwidth specifications from vendors or manufacturers. Signal, BC, Rhino, Heartland, take your pick generally have inadequate numbers or don't publish anything at all. Getting comprehensive information I found was difficult. Hence I searched around the web to find a manufacturer of coax that would have some kind of meaningful information from which we can draw some inferences and reasonable conclusions. I found that Madison cable to have fairly good information from which some data mining could be done. They also meet my personal requirement of a major manufacturer (they're very large) that not only draws wire but makes cable, sells to the industry as a whole, and has a Quality Control and Research laboratory with scientific instruments. The information below lists various cables, their gauge and whether it's solid or stranded, and bandwidth at various frequencies and lengths. This was calculated from information gleaned from their website. I believe you can use the information presented in this post as a guide for estimating the bandwidth capabilities of a cable that you may be interested in when the only thing you know is it's gauge, solid or stranded.
Component cables are generally run from one's DVD or similar unit to the display device. It can also be run from the DVD to the receiver and from the receiver to the display device. When the latter occurs, one will also find a bandwidth specification for the receiver. Again, as mentioned above, a receiver that has a bandwidth of 50 MHz means that at 50 MHz the signal will be down 3 dB. Using the 3x rule we can see that this receiver should be capable of passing a progressive scan signal with no degradation. However consider the following. If your cable is too long and marginally meets the spec, then the cumulative degradation may result in an unacceptable video signal. In those cases you've got a few choices.
1) buy a receiver with a wider bandwidth
2) use a cable with higher bandwidth. this need not be expensive unless you choose to make it so.
3) just plug your DVD directly into the TV
If we triple the values that Bob gave and do a little bit of rounding to facilitate the requirements with Madison's published numbers we roughly say that our bandwidth requirements in the three cases will be ~10 MHz, ~50 MHz, and ~100 MHz respectively. Moving to Madison's site we have the following:
Part #:017GD00004 RG-59 solid copper 22 AWG FEP insulation
Capcitance:17.3 pF/ft @ 1 MHz
10 MHz:288 feet
50 MHz:125 feet
100 MHz:83 feet
Part #:01LGD00001 RG-59 solid copper 22 AWG LDPE insulation
Capcitance:21 pF/ft @ 1 MHz
10 MHz:273 feet
50 MHz:125 feet
100 MHz:83 feet
Part #:017GD00004 RG-6U solid copper 18 AWG Foam PE insulation
Capcitance:16 pF/ft @ 1 MHz
10 MHz:429 feet
50 MHz:200 feet
100 MHz:150 feet
Video Cable26 AWG MultiStrand
10 MHz176 feet
50 MHz81 feet
100 MHz58 feet
Video Cable28 AWG MultiStrand
10 MHz:150 feet
50 MHz:67 feet
100 MHz:47 feet
Video Cable30 AWG MultiStrand
10 MHz:120 feet
50 MHz:53 feet
100 MHz:38 feet
Perhaps it's more comfortable to look at a well-regarded coax: Belden 1694a
| Part #1694a solid copper 18 AWG Foam HDPE insulation |
Capcitance:16.2 pF/ft @ 1 MHz
10 MHz:417 feet
67.5 MHz:191 feet (no value for 50 MHz from Belden's website)
100 MHz:163 feet
So how good is Radio Shack's 15-1563 cable? Well we know (I know because they sent me the information regarding this cable) that they use a stranded copper with an effective gauge of 20 AWG. That's somewhere between the RG59 and RG6. If we shade it towards the lesser performing cable we see that the RS product is a very good performer that vastly exceeds our requirements.
So what does all this information suggest? Well an examination certainly indicates that the thicker the center conductor, the longer the run can be before you get 3 dB down. It also suggests that within a particular gauge there's not a whole lot of difference in the length before one get's 3 dB down. It also suggests that even quite thin cables are emminently usable if the lengths aren't too long. For example, let's look at the 30 gauge and contrast it to the RG-6 for a 2 meter or a 6 foot run. That's pretty typical. Well if we're running progressive out of our DVD and using the 3-times rule, we'll be considering the 50 MHz values.
RG-6 down approximately 0.09 dB
30 AWG stranded down approximately 0.34 dB
So is the RG-6 better? Would we expect to possibly see a major or a minor difference? Well consider that many people arbitrarily use splitters if not multiple ones for their cable TV's and report entirely satisfactory performance. Aside from a trivial insertion loss a 1:2 splitter is going to cut your signal in half. In other words a 3 dB loss. How much of a loss is there with 0.09 dB or 0.34 dB? Percentage wise it's 2.09% and 7.25% respectively. Don't forget, that's at 3 times the expected signal you're going to be sending. Given that seeing the difference between SVideo and component can be pretty tough for people, I think a 5% change is going to be damned tough. And if you adjust your TV I dare say it'll be next to impossible to pick one from another with any degree of reliability.
So are there reasons to choose one cable over another for a relatively short length? Well sure. If you've got an interference problem you can't figure out where it's coming from, you just might want to go to a quad shielded RG6 or RG59. Maybe you don't like molded connectors. Maybe you don't want to futz with your TV (surprising from audio and videophiles though!). Maybe you just want to overspecify (you could be the sort of person for whom distilled water isn't good enough for your steam iron and you get 18 megaohm water from the lab). Flexibility may or may not be important to you. Maybe you're a solid center conductor person. Maybe you like stranded. Maybe you've got it in for Radio Shack. Maybe you think that cable looks cheesy or this cable's RCA don't cut it for you.
From the point of view of perceivable performance though, maybe it just doesn't matter. As careful as we want our children to be with their possessions we can be quite destructive to cables. We'll bend them, step on them, pull on the cable instead of using a gentle twisting motion to remove the RCA, we won't perform any kind of periodic maintenance that looks like cleaning the connectors, and then one day we'll notice interference or some sort of a bad picture. We won't damn ouselves, but we will damn the cable. We'll come to this forum, to that forum, and we'll buy the best cable that's out there and voila, all the problems have gone away. We'll sing the praises of this new cable, and brother, sing you better after you've dropped a hundred or more. Yes there's a no questions asked return policy but having to eat s/h both ways and probable insurance charges can quickly put a damper on your enthusiasm.
So how much do you or should you pay? Well if you're a fan of the 10% rule, knock yourself out. If you're into specifications and popular brands then buy the highest spec'd cable you can. If you're into buying a cable that can do the job and then some and you're running a couple of meters or so well you just might find there's a whole bunch of places on the web selling 100 megahertz cable that's 3 dB down at 100 feet. A couple of meters is somewhere around $10. Yes $10. Why so cheap? Well it's mass produced for one. The manufacturer isn't advertising and instead selling to resellers. It's being made using cable from lesser known (sometimes even well-known) companies but they're buying a couple of million feet. Costs go down you know. Yes it's 75 ohms. Maybe the question is why are some companies charging so much?