CATV distribution

The biggest challenge is implementing pay-per-view, which requires two-way communication. Yeah, it could be done with a side channel into the telephone system, but that's old-school.

CATV was ''community antenna television" at one time, which was exactly a distribution of signals effort. That purity is lacking in the present offerings from cable companies. DOCSIS modems and streaming of... nearly everything, isn't a TV function, but IS happening; there's a lot of cable boxes with hard drives inside (buffering video, not always with purchase-this-video recording and playback).

I was interested in how the *video* was distributed. This seems like IP is the only modern answer, regardless of authentication, gating, etc. I.e., you can put 1 drop, 100 drops or 1000 drops with basically the same infrastructure and system design.

I don't think RF scales quite that easily (though the noise floor may be far enough down that it isn't a practical limit on number of "loads")

Why not ethernet?

whit3rd snipped-for-privacy@gmail.com wrote in news: snipped-for-privacy@googlegroups.com:

Just Dangload them thar movies and play 'em from a USB stick! Yee Ha! Them thar smart TVs will play dem files....

I just watched that thar new Djinn movie. It is a really good flic.

for a hotel:

In the old days they did't they just put n decoders in a back office each decoding a different channnel onto a different output channel and then combined those RF signals, and fed them to the distribution amplifiers.

These days it's probably all IP over ethernet.

One would think there would be a turnkey solution as there are lots of hotels, lots of apartment houses, multiunit condos, etc. I've not been able to find evidence of these.

Originally, it would have been a matter of just delivering signal to N units (assume each unit had their own STB). So, an N-way distribution amplifier (where N is reasonably large).

Or, "rebroadcasting" downconverted to normal VHF/UHF channels. This would add some sort of M-way tuner on the front end, a mixer and N-way distribution amp.

The need to support logically "separate" drops to each customer (so each could have his own cable modem) would be a relatively new addition.

No doubt. Even if the STB *looks* like a cable converter.

What are the individual channels like ? Analog ,8VSB digital or DVB-C digital ?

The combined peak -to-average ratio limits the number of channels through an amplifier.

Especially with analog signals, the added noise in each amplifier limits the number of amplifiers in series.

For digital signals, the number of series connected amplifiers can be very large, if after a few amplifiers decode each channel individually, apply ECC and remodulate each channel individually. This allows eliminating the added amplifier noise as well as amplifier distortion.

what about fiber ?

modern TVs come with fiber output/input

I'm thinking of "classic" (analog) cable. Digital transport eliminates a lot of problems that, I would assume, would plague "wide" analog distribution.

Exactly. Imagine taking a CATV feed (or, from a satellite dish atop a 400 unit hotel) and distributing that signal to all of the units therein.

That is usually just for audio input and output. However, many GPON fibre modems operate with three different wavelengths on the fibre. One for internet uplink, another for internet downlink and a third for TV downlink.

John

I see no problem. The classic analog way was to amplify the signal and then distribute it through a many-way splitter, which may be a star or composed of multiple units distributed throughout the building. You'd never (well, almost never) put amplifiers in series.

Jeroen Belleman

Each split loses signal. So, you need to make up for that, "up front" (if you don't include additional distribution amplifiers "mid stream") Note that it's not impractical for there to be thousands of units in a single establishment (e.g., The MGM Grand has almost 7000 units and was opened in the early 90's -- presumably, they have had "in room TV service" since their opening)

A TV tuner needs about -80dBm to function. For 7000 tuners, you'd need -41.5dBm (-80 + 10log(7k)). Allow another 13dB or so for splitter losses (roughly log2(7k) dB), that's -28.5dBm. Add 20dB for

100 channels and another 20dB for headroom. We're at +11.5dBm. Add, say, another 12dB for distribution cabling losses. Final power needed is about +23.5dBm, or a little over 200mW.

Looks like a walk in the park for the amplifier.

Jeroen Belleman

You see only 1dB per split loss? Not *3*? E.g., a 16 port splitter already has you down ~12dB -- you're going to do that 7000-way for 1dB more?

The 3dB drop in a two-way splitter is not a loss! Each output gets half the power, -3dB. There is some real loss too and I generously budgeted 1dB per binary stage. It's usually less.

A 7k-way splitter would have -38.5 dB input to outputs, plus maybe

13 times 1dB for each stage if it's a binary tree.

Jeroen Bellemanm

That's what I was trying to describe, but it was an M way splitter M STBs and an M way combiner in a back room, or a closet. the cable compoany doesn't need to stock expensive hardware that way,

I distribute broadcast tv over ethernet here, or did - the drive in that PC failed and I suspect the PSU may be bad too.

But, does the cable company handle the site-specific configuration? I would have thought that was a separate market that the site owner bought into to suit their needs.

From what I've heard of folks with CATV service, relying on the cable company for "prompt support" is laughable.

I do here, as well. Along with stored video (from a media tank); I've got a really old, antique radio (piece of rosewood furniture) from which I've removed the old "tube set" to make room for the media tank. (SWMBO isn't fond of electronics kit lying around but a nice antique is a different story, entirely!)

Haven't found a way to put the original controls to use, though (the tuning "arm", volume control, etc.)

I have several HD HomeRun tuners hidden away to put the content on the wire and other nodes that can process (commercial detect) and store it. Conceivably, I could support cable/satellite sources, as well, but we just don't consume much "live content".

Presently, I use diskless workstations (X Terminals, etc.) as the presentation end as they've already got the audio and video hardware in place. It will be a real challenge to design something cheaper/smaller so I can afford to defer that activity.

Sadly, no one seems to make a similar device for broadcast radio (AM/FM/FMHD/Xirius) so I've been schooling myself in SDR technology. Again, hard to imagine acquiring enough know-how to beat a commercial offering but it can't hurt to understand what the selection issues would be.