Upgrading house from rg 59

[RumRunner]

Coax isn't really manufactured for any particular frequency limit. How it works is the higher the frequency the higher the attenuation which means the shorter the cable. Taken another way, if you know how long your cable has to be, you choose one based on the highest frequency used and how low a signal you can tolerate at that frequency.

So sure, the old RG-59 can be used to 2Ghz but after 100 feet you would be over 16db down. (Most charts don't go over 1Ghz because that's considered the limit of it's usefulness.) This is similar to voltage drop on a pair of conductors, you can supply them with 120V but what is left at the other end when you consider the load. It works the same way- if you want to have the most voltage at the other end you use bigger wire or in this case a bigger lower loss cable.. . . .


-Hal

Most application of coax in homes (not interconnections ) from the street to modem are for broadband.
No. . ., cables are not "custom " made to handle the particular frequency they are intended to be used. You select the highest rated frequency they can handle.

Again, for home use you won't (and you can't) exceed 2000 Mhz because ISPs don't even transmit close to 2000Mhz.
It's more like cables made for 600 volts but you are only using it for 480 volts or less.
Attenuation does occur anytime you transmit voice or data from the source to its destination.
All these issues are really non-issue because of data that ride on higher frequency can be boosted or demodulated any which way you can.

The enormous broadband that we now have-- can be modulated and demodulated and channels can be separated for voice, video and data. Adding to that, the awesome resolutions on video are something we haven't seen before.

That's the improvement over the invention based on Heaviside experiment . . . the physicist and mathematician who first experimented on it.
Losses in the small range is not a big concern as I have said --since the increase in frequency can be utilized to compensate or correct for the loss.

Comparing data that rides on the wire cannot be compared with power that is delivered from power source to an appliance that uses that power.
You can have voltage without (that much) power. The important component is the frequency where you can have data to ride on.
Now, don't confuse this as something like:

How can power travel without current?
This violates OHM’s LAW. Would it not?

Well, data delivery doesn't depend on current. All it needs is continuity for "PACKETS" to be able to travel from the packet creator (the server) to its particular address --embedded in the packet and encoded on the request when the "handshake" between client and server was initiated.
Continuity is assured and can be monitored by a small resistance at the furthest end-- that will cause a very small amount of current to run through the wire. This small current is NOT there to be used in running the electronics at the destination. The signal riding on the wire are those ZEROes and ONES (binary.)
Think of old telegraph DOT and DASH signals.

So, you can discard the thought that data is lost because of the size of conductor or the power (amperes) is insufficient like toasting our morning bread.

Electrical engineering is mostly concerned on how we can run our machines with the power at our disposal--and not about the attenuation in decibels.

We will leave that to the physicists and scientists who have too much time on their hands to tinker with them.

 

[zbang]

Really nothing to do with digital

Yep, and at the physical layer, it's really analog anyway.

 

[rnatalie]

The leakage between houses aren't the problem so much as allowing RF to leak out into the general airspace interfering with over the air things such as Air Traffic Control and other users that happen to be on the same frequency.

 

[GoldDigger]

Although there does not need to be DC continuity at the receiving end of a digital cable transmission line (which allows supplying power in the form of DC over a signal cable), the input impedance of the digital or analog signal detector must be a primarily resistive impedance whose value is the characteristic transmission line impedance of the cable. That means that the combination of conductor resistance and dielectric losses along the cable must be low compared to the characteristic impedance for a large enough signal to be useful to be detected at the receiver. Looking at it as signal power rather than signal voltage is a perfectly legitimate way to characterize performance.

And as for the data being in the form of zeros and ones, a practical system will actually use the digital signal to modulate one or more carrier frequencies to transmit the information. Modern signal encoding will use a combination of amplitude and phase/frequency encoding to pack the maximum number of bits within the bandwidth of a single carrier frequency channel. One useful interpretation of this is that analog signal processing is essential to recover the digital information from the signal waveform on the cable.

 

[synchro]

... Looking at it as signal power rather than signal voltage is a perfectly legitimate way to characterize performance.

Yes. In RF test equipment dBm (decibels relative to 1 milliwatt) is a very common measurement of the signal power present at coaxial inputs.

And as for the data being in the form of zeros and ones, a practical system will actually use the digital signal to modulate one or more carrier frequencies to transmit the information. Modern signal encoding will use a combination of amplitude and phase/frequency encoding to pack the maximum number of bits within the bandwidth of a single carrier frequency channel. One useful interpretation of this is that analog signal processing is essential to recover the digital information from the signal waveform on the cable.

To expand on this a little, in the quadrature amplitude modulation (QAM) that's used in cable, wi-fi, WiMax, LTE, etc., data is conveyed in the form of symbols, where each symbol has a unique amplitude and phase representing multiple bits (i.e., a digital word) of information. The latest deployed cable standard DOCSIS 3.1 can modulate using a "constellation" of up to 4096 QAM symbols, and optionally even more. 4096 = 2¹² and so each modulation symbol conveys 12 bits of information.

In 3-phase power, phasor diagrams are commonly used to represent voltage or current and phase on the "complex plane".
These digital modulations are also represented on a complex plane except that they are referenced to the carrier frequency instead of 50/60 Hz.
So for example, 4096 QAM symbols would be a "constellation" of points on a 64 x 64 rectangular grid in the complex plane, centered at the origin. When the signal is modulated the phasor moves from one point to another at the "symbol rate", conveying 12 bits of information at each point. The phasor has to move in a relatively smooth manner between these symbol points so it doesn't occupy too much bandwidth. The amplitude and phase of the RF signal waveform is a physically-based quantity and is therefore analog. But as time goes by more and more of the signal processing of these waveforms is done in digital hardware as the speed and current drain of CMOS improves. The ultimate goal is just to have an A/D converter on the input of a box and a D/A on the output.

They often do QAM modulation using relatively narrowband modulation on multiple subcarriers instead of doing broadband modulation on just one carrier. This makes it much less sensitive to timing errors and delay vs. frequency, because with narrowband modulation the symbols on each subcarrier will move more slowly than with broadband modulation of one carrier. Having such multiple subcarriers is the FDM (frequency division multiplexing) part of OFDM.

The symbols for 4096 QAM are more closely spaced than say 256 QAM (a 16 x 16 rectangular grid), and therefore it needs a higher signal-to-noise ratio to maintain an acceptable bit rate in the presence of noise. And so the signal power discussed earlier on this thread is important, particularly for the higher data rates where constellations with more symbols are used..

 

[hbiss]

The enormous broadband that we now have-- can be modulated and demodulated and channels can be separated for voice, video and data. Adding to that, the awesome resolutions on video are something we haven't seen before.

One of these days you'll see cable providers dumping discrete video channels and making the entire bandwidth IP. No longer will each channel take up permanent piece of bandwidth regardless of whether it's used or not. All subscribers will need is a browser and the TV channels will be available on the internet from the cable company the same way as Netflix and other services are now.

That's the day when we quit running RG-6 around a house in favor of CAT-5e or WiFi.

-Hal

 

[grich]

One of these days you'll see cable providers dumping discrete video channels and making the entire bandwidth IP. No longer will each channel take up permanent piece of bandwidth regardless of whether it's used or not. All subscribers will need is a browser and the TV channels will be available on the internet from the cable company the same way as Netflix and other services are now.

That's the day when we quit running RG-6 around a house in favor of CAT-5e or WiFi.

-Hal

Our provider has already done that. Ten years ago or so, they laid fiber, and the ONT presented us with a CATV output, an Ethernet output, and a POTS output. About 4 years ago, faced with a head-end rebuild, they instead went IP Video, using Amino decoders at the homes. It worked well, and came with DVR capability, but the price was still high, thanks to ESPN. We dropped that, got Roku boxes and stream stuff that way. Saving almost $100 a month. Each TV point has RG-6 and CAT-5e, with the RG-6 homeruns fed from an antenna in the attic. Internet is 100 down/50 up, more than fast enough for us.

 

[hbiss]

Internet is 100 down/50 up, more than fast enough for us.

With all that new found bandwidth I'm sure they can give you a lot more speed if you wanted to pay for it.

-Hal

 

[grich]

With all that new found bandwidth I'm sure they can give you a lot more speed if you wanted to pay for it.

-Hal

And they can if we need it.

The battery in the ONT's UPS finally went bad last month. I called the provider, a local telco co-op, and asked if they could just give me a battery to swap out. They said, "heck, we'll give you a whole new UPS. Swap it and bring the old one back whenever." They know what I do for a living, they didn't have to roll a truck or deal with COVID protocols, etc. Fun living in a small town.

 

[mikeames]

One of these days you'll see cable providers dumping discrete video channels and making the entire bandwidth IP. No longer will each channel take up permanent piece of bandwidth regardless of whether it's used or not. All subscribers will need is a browser and the TV channels will be available on the internet from the cable company the same way as Netflix and other services are now.

That's the day when we quit running RG-6 around a house in favor of CAT-5e or WiFi.

-Hal

Yep, I have FIOS and 1GB internet, all fiber to the house and then my house is all 6A homeruns to every TV and computer. The issue I have is the STB all still use Coax but the STB all have an ethernet port. I called VZ and asked if my STB could use the ethernet instead of Coax. NOPE. The hardware is designed for future use and is capable but FIOS still uses Coax because that's the most common infrastructure. Seems like we are in that transition period and probably will be for another 5-10 years until there are enough houses with more modern infrastructure.

 

[hbiss]

Seems like we are in that transition period and probably will be for another 5-10 years until there are enough houses with more modern infrastructure.

I think you will see the change to IP video much sooner than that. I also think that they aren't using UTP because they don't want to stock two different types of STBs and it's easier to train installers for one method. Two would totally confuse them.

-Hal

 

[LarryFine]

Plus, ethernet wiring requires more talent than does coax.

 

[mikeames]

Plus, ethernet wiring requires more talent than does coax.

Talent is such a strong word here........ Ill go with patience, focus, effort...
Bending pretty pipe is talent IMHO.

 

[hbiss]

Talent is such a strong word here........ Ill go with patience, focus, effort...

Let's put it this way. When someone applies to be a cable installer, at the front door there is a sign that says "Anyone with an IQ higher than 50 need not apply".

-Hal

 

[steven765]

Quality RG6-quad shielded will solve that. Get a compression tool for making ends. Since it's coming off a router you don't have to worry about the dB loss as the signal will be plenty strong. That said at most you should have 1 splitter and use the termination caps on any unused connections.