I don’t know anything about digital coax cable service offered by cable companies. I was asked today if I could add a cable outlet behind an entertainment center. To “split” off of a digital signal can one effectively use a coax splitter commonly used in analog cable service? Or if not, how does one extend or otherwise “add” an outlet in such a service? I assume RG6 quad shield coax is appropriate as well?
If at first you don’t succeed, try using a hammer next time…everything needs some extra persuasion from time to time. -ME
Replies
There is basically nothing special about the "passive" components in a digital system vs analog -- the cable, splitters, connectors, etc are all the same. In fact, the signal's essentially the same, it's just that a given TV channel may have been converted to transmit digital rather than analog data.
There IS a difference with regard to cable amplifiers, if you will be using a cable internet modem or some sort of cable box that communicates back to the home office. The amplifiers have to be designed to permit the "backwards" flow of this data.
Of course, some cable companies will mumble stuff about how the cable has to be special, but they're mostly blowing smoke. You just need to use good quality stuff and be careful about balancing any splitters used.
(Trivia item: Only cable, cellular, and satellite modems implement "broadband". DSL and plain old CAT5 local net wiring are not "broadband". Has nothing to do with the bandwidth, but rather the transmission technology.)
Thanks, Dan....sounds like as long as I use a high end splitter it'll be fine. A coax cable runs nearby about a foot away. I was hoping to just split off it, sounds like that's all I need to do.If at first you don't succeed, try using a hammer next time...everything needs some extra persuasion from time to time. -ME
Even more important than the quality of the splitter is the quality of the connectors (like pretty much all wiring things involving the movement of electrons). I asked about this a while back, and the genearl consensus is that the crimp and seal connectors are the best.Digital transmission will occur up to something like 4000 Mhz (not sure exactly). But the thing to do is to check every use of the coax (TV, modem, closed circuit camera, etc) and find out the maximum speed. Then use connectors, splitters and cable that is rated at least that good.Here the cable company will sell users the proper everything that meets their standards, and my experience is that their requirements are the highest, as compared to say closed circuit monitors.The advice to homerun everything if possible is spot on. Ideally, run everything back to a distribution center.
Those are Snap-N-Seal, not crimp. Some, like the ones from F-Conn, actually slide a conical sleeve over a softer area and more or less swages it tight. Thomas & Betts are very common and can be bought all over. I would recommend connectors that are made of something other than polished nickel- those tend to slip off and are hard to tighten when your hands are hot and sweaty or if there's any oil on your fingers or the connector.. Most cable is sweep tested to 3GHz and there's nothing in general cable signal that is more than about 2GHz. Ditto the home run and central location, preferably having runs of similar lengths.
"I cut this piece four times and it's still too short."
Yeah, snap and seal, that's what I meant. Still seems like they crimp in a way, to me, but definitely they are called snap and seal.
I would recommend connectors that are made of something other than polished nickel- those tend to slip off and are hard to tighten when your hands are hot and sweaty or if there's any oil on your fingers or the connector..
About 10 years ago, I was trimming out a commercial job with voice/data/video plates. I was still using the crimp-type F-connectors then, and would leave the center conductor long, trimming it after crimping. I was on my second plate (of about 30), and sure enough, the connector slipped, with the center conductor going straight through my finger. I finished the plates that day, but it sure slowed me down...
I'm using the T&B system these days, and they're a lot easier (and safer) to install.
Bob
I'm kind of partial to having the stinger slide under my fingernail, but that's just me.
"I cut this piece four times and it's still too short."
Dan, I'm curious as to how you consider some technologies braodband, and some not.I'm in the computer field, but I don't understand your statment, care to expound further?A medium to large guy named Alan, not an ambiguous female....
NOT that there is anything wrong with that.
"Broadband" is the process of transmitting multiple signals simultaneously, each on a separate frequency as with broadcast radio/TV. "Broad" because a substantial "width" of the radio spectrum is occupied."Baseband" is the transmission of a single signal without modulation.There's a grey middle ground that DSL occupies -- not baseband since it's modulated, but not truely broadband either.Early on, a sort of broadband based on TV channels over coax appeared to be the front-runner for high speed computer interconnections (the niche now occupied by Ethernet). Basically, each computer on a network would have it's own private transmission channel, and someone wanting to receive would tune into that channel. The adapter card actually included a primitive TV tuner.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
You're applying a limited definition here. Tryhttp://en.wikipedia.org/wiki/BroadbandGeorge Patterson
Well, the definition has been "adapted" over the past 30 years or so, and different definitions have been adopted by different disciplines. I'm referring to the original definition.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
I understand your point now Dan.You had me confused thinking on a speed basis.
And that is a good point, a twisted pair can only carry so much data, no matter how new, and how short.Token ring is the coax based tv like network you are thinking of, I think.I learned about the old token ring networks, but I've never worked on one.A medium to large guy named Alan, not an ambiguous female....
NOT that there is anything wrong with that.
No, token ring isn't coax-based and is "baseband" just like Ethernet. The TV-based scheme predates both of those.Another bit of trivia is that "Ethernet" derives from "Alohanet", a radio-based, education-oriented computer network in Hawaii. Alohanet was really a true "ether" network, and the first "wi-fi", but more significantly it was the first practical use of collision-based networking.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
Well, it's been too long since my studies I guess.A medium to large guy named Alan, not an ambiguous female....
NOT that there is anything wrong with that.
Token-ring is baseband, like FDDI. If I remember correctly, ethernet net tosses in packets like little sacks with an address label on them while token-ring attaches packets into a chain lead by a "token" that announces its identity.
<!----><!----> <!---->
Phill Giles<!----><!---->
The Unionville Woodwright<!----><!---->
Yeah, in theory token ring can achieve higher bandwidth utilization than Ethernet. But bandwidth utilization has rarely been the most important concern, and token ring is a more complex protocol, plus the electrical "ring" is fragile.A lot of this became moot, of course, when it became obvious that the star/hub physical architecture was going to win out over the ring or bus. With modern switching hubs the electrical protocol really doesn't matter that much.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
I guess you could consider gigabit ethernet broadband, even though it's just using all the pairs in the cable, the packets and addressing is still the same AFAIK.I wish I could remember what the system was called that we tore out of an old middle school, it was coax, and had some other odd deals, but I think we called it token ring by mistake.A medium to large guy named Alan, not an ambiguous female....
NOT that there is anything wrong with that.
Of course the original Ethernet was coax. Hardly ever used anymore.Before that there were a half-dozen other schemes, mostly involving coax of one sort or another.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
Now that I think about it, one topology used twisted pair to the wall-jack, and then co-ax from the wall to the comm-card. I still have a box of connectors that we used to join a couple of wall-to-machine co-ax cables together so that we didn't have to put the desks right against the wall-jacks. The basic reason is that the early monitors and many attached printers produced so much EMI that they killed their own signal; it also meant that the same topology could be used in any environment (shop-floor was a nightmare for the early stuff).
<!----><!----> <!---->
Phill Giles<!----><!---->
The Unionville Woodwright<!----><!---->
And of course there were all sorts of computer to "dumb terminal protocols before you had "local area networks". Mostly these were "home runs" of coax or multi-wire cable to a "controller" somewhere (a box sized anywhere from modern "floor model" PC to a 2x6x12 foot monster, handling from four to maybe 32 terminals). Between controller and computer there was generally a 2K baud telephone line running BISYNC or SDLC.The earliest (10-Base-5) Ethernet was one of the most bizarre PC networking setups, from a cabling standpoint. A piece of RG6X "yellow garden hose" was run the length of the building, to a max of 500 meters, with terminators on each end. Then "vampire" taps were used attach a small transceiver box that converted the signal to a multiwire signal to the computer. Vampire taps could only be installed at designated points on the cable, designated by black bands.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
Ah, ASCII 3101 push-pull terminals - one character at a time through a 56 bps (that's right, no "K") incredably expensive modem. Remember when Gandalf modems first appeared ?
<!----><!----> <!---->
Phill Giles<!----><!---->
The Unionville Woodwright<!----><!---->
Never dealt with a 56 baud unit directly. I know that the old 5-bit, 1.5 stop bit SR32 TTYs were anywhere from 50-75 baud, depending. Used by military, RTTY guys, and the old Teletype cable network. Sounded like a sewing machine on slow, or a very quiet machine gun. The SR33 models that ran a 7-bit (2 stop bit) code with parity (though parity was never checked) were pretty universally 110 baud -- 10cps, more or less. Worked with many of those and was employed as a TTY repairman in college.Saw all sorts of modems. Most of the lower speed (below 300 baud) units were acoustic coupler units -- don't recall the brands. High speed units (generally 2000-2400 baud) ran synchronous protocols, vs the async protocol for the slower units.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
For most of my TTY experience, we used the 8-track (8-hole) Friden system; the 7-track system matched IBM (and other manufacturers' 7-track tape-drives); we used ASCII push-pull for CALL360. I can't now recall what we used to connect to BitNorth (my gateway into the net, long before it was the net - 1970ish).
Real challenge was using card reader/punch units as communicating devices (12-row, 80 column cards)
<!----><!----> <!---->
Phill Giles<!----><!---->
The Unionville Woodwright<!----><!---->
Yeah, the 8-bit (7 plus parity) ASCII system is what was used for the Teletype xSR33/35 series. The 5-bit TTY system was generally called "Baudot", though there were several variations and official terms. IBM used 5, 6, and 7 bit variations of the "BCD" code, a code that roughly correlated with coding for keypunch cards, and then of course introduced EBCDIC with the System/360. They also promoted the aborted "USASCII", a kludge that tried to split the difference between ASCII and EBCDIC.The communicating Selectric terminals used a form of BCD, I believe, that was 6 bits of data and something like 1.3 stop bits. They used an odd data rate -- something like 134 baud.There was also a 6-bit code that was the predecessor of ASCII -- can't remember what it was called.Electrically, terminals were generally either "current loop" or RS232. Current loop was compatible with old telegraph signalling technology and is what Teletype terminals used by default. RS232 was the standard that Bell developed for their modems. I'm not quite sure what "push-pull" was.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
You guys are bringing back memories of my first model 15 with a tube type polar relay TU.Then I graduated to a 28ASR with an ST6 homebrew TU...It had punch tape in and out, It was on a 2M local autostart freq...For grins, if one was up at 3am, you could bring all the guys machines up and ring the bell on em...60wpm Baudot ...
Bud
If the old system had BNC connectors with T connectors and RG 58 (stranded coax), it was a bus system. IBM came up with the token ring system and there are two connections at each node. The path is literally a ring, although it could be rectangular, it just means that the data moves from one node to the next and only the correct computer will respond since they all have unique addresses. Kind of like the kid's game, where someone tells one kid something and they all have to pass it around to the end(beginning) again. A bus is just a series of connected computers on a single trunk line.
"I cut this piece four times and it's still too short."
Yeah, there was a ring network that IBM had before Token Ring. "Rocket", I believe it was called -- used for bank terminals mostly. I never saw one up close, but I suspect it was coax w/BNC.IBM also has the 5250 Twinax scheme, for S/34/26/38 terminals. Connectors looked like PL259 connectors but had two pins inside. These could be "daisy-chained" out to 1000 feet and, IIRC, seven devices.The "thinnet" 10Base2 Ethernet also used coax with BNC connectors. The cables were daisy-chained through T connectors on the back of the adapter cards. Same basic topology as 10Base5, but avoided the separate transceiver and AUI cable.Token Ring, from the get-go, used the funky square bisexual connectors ("IBM Cabling System") that allowed the same cables to be used for Twinax, Ethernet, and some of the old coax terminal/controller protocols. The cables were 4-wire and would automatically "pass thru" to maintain a loop when disconnected.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
daisy-chained through T connectors on the back
Ugh, hated those "bayonet T's". Not because they were bad, but because electricians would move the coax just a bit while working one soemthing else. That movement always seemed to rotate just the one swithc (or heaven help us, a router) out of the backbone, but only mostly. Did not help that the backbone run was about 275' to the server (so that the people at the end of the backbone always had packet collision problems . . . every so often their complaints were real, snagdangabbit . . . )Occupational hazard of my occupation not being around (sorry Bubba)
You wanna see collisions, knock a "Cabling System" plug loose that's carrying 10BaseT. The self-shorting connectors end up feeding the receive data back to the transmit data. "Smart" hubs would just disconnect the cable, but "dumb" hubs would lock up.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
Yeah, the Thinnet Ethernet systems were a lot of fun to troubleshoot. Someone moves their computer and network goes dead. Not always obvious which one, so start disconnecting computers and substituting terminating devices until you find which one it was. Someone earlier mentioned an old network in a school. It could have been ARCnet, which was one of the earlier networks that became fairly popular and used some type of token passing scheme. I studied the system, but never actually worked on one. There were a lot of different types of networking before 10BaseT Ethernet pretty much swept the field. I still have a bunch of Appletalk connectors which could use two wires of an existing telephone line to connect the computers. Not particularly fast, but easy to set up and, for the time, relatively inexpensive. Most of the earlier networking schemes were proprietary and Ethernet got a boost from being public domain. The Ethernet 2BaseT and 10BaseT were not fast enough to interconnect mainframes, and while the 10BaseT pretty much took over on local area networks, the proprietary networks were used where higher speeds were needed. A number of computer folks figured Ethernet would not be able to keep up with increasing speeds. Ethernet used a system called CSMA/CD (carrier-sense multiple access with collision detection) which means that the ethernet card would detect a silent period on the network and then out a data packet. If some other card also sent out a packet at the same time, a "collision" would occur and the card would detect that the data were scrambled and start over again. With a lot of users, collisions can be a signficant problem and it was felt that this would rule out use of Ethernet in on heavily trafficed networks. However, with the advent of higher speeds and the use of routers to limit traffic between subnets, the problem was manageable and Ethernet pretty much took over.Modern network models call for a modular type of design. Thus various protocols should be able, in theory anyway, to work with a variety of physical transmission scchemes. IIRC (and I have managed to forget more of this stuff than I now remember), Ethernet protocols are currently used with not only the original coax cable (which, by the way, had to be a specific impedence which was different from that used by video) and the twisted pair protocols that replaced it (2BaseT, 10BaseT, 100BaseT, and 1000BaseT) but also optical fiber (1000BaseFX) and the WiFi 802.11 series wireless Ethernet. (IIRC, the designation of 10BaseT comes from a maximum bit rate that can be shoved down the wire is 10Meg with a baseband schema and "T" for twisted pair. However, the data throughput is always much less than the rated maximum.) Who knows, maybe we will even see Ethernet over quantum weirdness in the future...And, the beloved Teletype 33 terminal. I wish I had grabbed one of those when I had a chance, but had no place to store it. It was kind of neat to hear the thing chuggig along on idle waiting for some data to come in. I remember they used the sound to introduce a news program back in the days of radio newsbroadcasts as most stations had a Teletype 33 in the backroom to get their news feeds before computers were in common use. Back in my early academic days, we had a Digital PDP-8 computer that had to be booted from switches on the front of the machine and then the data was fed in from a punch tape that was made on the Teletype 33. Things have changed just a mite in 40 years...The original Hollerith character codes (the IBM punch cards) were a six bit codes (no capital letters). I think the CDC computers we used at Berkeley in the late '60s used a 7 bit ASCII code. Of course IBM has used the EBCDIC 8 bit codes on their mainframes while adopting the 8 bit extended ASCII for the PC. Apple also used an 8 bit ASCII code, but to make things difficult for those of us doing data conversion in the early days, they reversed the bit order from that of the PC.More than you wanted to know...
> And, the beloved Teletype 33 terminal. I wish I had grabbed one of those when I had a chance, but had no place to store it. Yeah, I had an ASR33 too. It was a weird bird, though, with a non-standard solenoid voltage.Interesting how the thing was controlled by a single solenoid, and there was no mechanical connection between keyboard and printer.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
EBCDIC punched cards were 12 rows, it became 8 in hex. Second gen stuff ran octal (i.e base 8).
<!----><!----> <!---->
Phill Giles<!----><!---->
The Unionville Woodwright<!----><!---->
I assume RG6 quad shield coax is appropriate as well?
I use RG6 Quad Shield exclusively for CATV and cable modem drops. A broadcast engineer (cheap) friend of mine contends that for the most part, Quad Shield isn't necessary. In any event, I avoid splitters to the extent possible - if you can make a home run to the service entrance, so much the better. Home runs make future changes and troubleshooting *much* easier.
Bob
What I meant to say is that I avoid splitters behind splitters. Of course, there will be some sort of splitter and/or amplifier at the service entrance.
Edited 4/13/2007 9:16 pm ET by bobguindon
Understood, Thank you.If at first you don't succeed, try using a hammer next time...everything needs some extra persuasion from time to time. -ME
If there's not an abundance of RF interference, quad shield isn't needed, provided that the connectors are decent and installed correctly. Poorly installed connectors cause many issues, the worst being signal loss and leakage/RF infiltration.
"I cut this piece four times and it's still too short."
Here is my experience based only on what happened in my own home:
We had regular cable TV and cable internet and it worked fine for some years, although now that I think of it, the cable modem would have to be reset every week or 2. TV worked fine. The CAVT RG6 coax comes into the house and then to a junction box, kinda like a small fuse panel box that I mounted in the wall when building the house. BTW - the quad shield wasn't out yet when I built the house. In the box was a splitter/amplifier and all the coax from elsewhere in the house comes in there, along with 2 phone lines per room and cat5e twisted pair for Ethernet (which has never been used).
Then we got digital cable and a plasma TV. We just couldn't get the TV to work exactly right - it had various seemingly unrelated and intermittent problems. The first cable guy tested the strength of all the signals. In addition the digital cable TV service included a test channel that basically displayed cable signal strength at the digital cable box. That is the simplified version - there is actually maybe 4 or 5 different signal values, and he told me what is the minimum goal and optimum for each. anyway, he said that the signal was a little weak at the TV but it should be OK. The other non-digital TVs in the house all worked fine. I had him come back, and he installed a better amplifier and a splitter - but it was still basically the same thing.
Then, about 2 weeks later I got a customer satisfaction call/survey. I said that the guy had bent over backwards to get it right, but I still had intermittent problems. They sent out another guy, who had more sophisticated test equipment. He said that there was something(?) wrong with the cable drop that went to the plasm TV. Having anticipated this, I had the cable all ready to go, this time RG6 quad since I wanted the best available, and since there is several conduits going to the junction box, I quickly pulled a new wire through the crawlspace and up the conduit to the J-box and at the other end through the floor to the plasma TV. This time he said he wanted the cable on the TV end to go directly to the TV - not through the wall box as it had been connected for several years - and like all the other ones in the house. Now it works fine.
So, I guess the short of it is that this stuff can be tricky and sometimes special test equipment can be required, or at least help. Personally I wouldn't mess with doing this for someone else. BTW - some time ago I was an electronics repair man so I have some background in this stuff.
In my case all this monkeying around didn't cost me anything since the initial digital cable install was included in the purchase/upgrade to the digital service, but let's just say it didn't go smoothly at all.
BTW - looking in the big J-box the main coax comes into the house and then through the 1 to 4 amplifier which says "Antronix ARA4-8" then to the the splitter says "3 way splitter 5-100mhz".
Good luck....
Nine times out of ten a problem of this sort is at a connector, so use the best connectors available, and the fewest connections you can manage.Most common problem is simply with the center wire getting bent or pulling back into the cable so that it doesn't make contact. You can still "radiate" a significant amount of signal through, so you may not notice it, especially on older sets with their better (more sensitive) tuners. (Modern sets are designed assuming they'll be connected to cable and never have to deal with rabbit ears or an unamped rooftop antenna.)
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
The first guy replaced all the involved connections. Some twice. I forget how it works... but I think it's a 2 part connector. Then there is a pliers like tool that forces the assembly together - it's not a crimp. I did notice that they now sell these type connectors at Lowes.
Really what I suspect is that the problem was 2 fold: first that our house is back from the road a bit and they really needed the next size wire up going to the house since the signal strength at the "service entrance" is not the best. I forget the designation but it's like the one we have is around 1/2" in diameter, and the one that would support the distance better is around 1" in diameter. I'd estimate the run to be about 160 ft. Secondly, I believe that the original coax that went to to the plasma TV was somehow damaged - either kinked, or perhaps the guy I had helping me when I pulled all the low voltage stuff shot a staple into it - although I never found one.
As I said before all the initial install worked fine with the analog cable but just couldn't satisfy the demands of the digital system.
My central point is that although this stuff isn't rocket science, it can be a bit tricky and some test equipment can be helpful. Further, having read some of the other circumstances Willy has been in, maybe he doesn't need the aggravation.
Yeah, a good field strength meter is quite helpful, as is a TDR tester.
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin
(But a knowledgeable DIYer who uses the good stuff can likely do a better job than the cable guy. More time, less pressure.)
So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin