Any non-sinusoidal signal has harmonics. With a square wave (the simplest case, the harmonic amplitudes are inversely proportional to the order of the harmonic. Even that "spike" can be analyzed into a series of harmonics. Look up Fourier analysis .
Again, do a Fourier analysis. Any repetitive signal can be analyzed into a series of sine waves. In the case of a 1 V square wave, it works out to a
1 volt sine wave at the fundemental frequency, 1/3 V at the 3rd harmonic,
1/5 V at the 5th, 1/7V at the 7th etc. Notice a pattern here? As you increase increase in frequency, the harmonic amplitude is proportionally smaller and smaller. Now with a 10 PPS dialing, you'd have to consider harmonics up to the millionth and more, where the level of those harmonics is going to be so low, that you won't be able to measure them through the background noise.
Again, look at the harmonic content of the dial pulses. You'll find no significant harmonic energy at the frequencies used by ethernet.
Precisely. That's why sharing an ethernet cable with a protected phone line is no more risky than having a modem connected to the computer.
The tower has three (IIRC) very heavy ground leads. In a severe storm, they can glow red hot. The equipment room is in a Faraday cage, but that's to prevent interference from all the TV & FM broadcast transmitters located on the tower along with the microwave and VHF/UHF communications systems.
While the cable are buried where I currently live (buried cables can also have lighting induced problems), in the past I've lived in areas with overhead lines. Also, many years ago, I used to work in Northern Ontario (where I saw those flashing gas tubes) on systems connected to open wire lines, where you've got pairs of wire, not cable, running for hundreds of miles. The storms in that area were often very severe. One night, I thought I saw some fluoresent lights flashing outside, during a storm. When I looked the next morning, I discovered that those "lights" were in fact power line insulators that were arcing over.
As for your area, what have they got against grounding? Lightning induced surges are generally common mode, which a protector across the pair won't protect against.
Quite so. I have also seen some bad installations. Any installation connected to the outside world through wire, has to have a good ground connection. A poor ground system is an open invitation for lightning induced damage. However, to get back to my point, running phone lines through an ethernet cable is no more of a risk than connecting the computer to a modem.
If we're talking about dial pulses interfering with ethernet, we're talking about harmonics. As I mentioned, even those spikes are some sum of various harmonics of something. What are the frequency components of those spikes? Without doing the analysis, you can't say. What I can say with a high degree of confidence, is that any spikes generated by some device attached to a phone line, are unlikely to have any significant energy in the frequency range used by ethernet. As I mentioned in a previous note, ethernet signals are in the range of several MHz, with 100baseT running in the vicinity of 125 megabaud, which has a fundamental frequency of 62.5 MHz. One thing that's been obvious for decades, as experienced with radio communications, is that such impulse noise from electrical devices, lightning etc., is virtually nonexistant at those frequencies. You will not find any such interference from "spikes" at those frequencies.
The original claim was interference from the phone line. There are 3 desired signals on a phone line. The main one is voice, with energy primarily below 4 KHz, 20 Hz ringing current and in some situations, pulse dialing, typically at 10 or 20 pps. A pure sine wave has no harmonics. Every other signal does. This means that everything on a phone line is either a sine wave of the sum of a series of sine waves. There is no other possibility.
Once you have determined the various frequencies involved, you can then determine:
a) the coupling of those signals between pairs and b) any interfering effect of those signals on the desired signal.
Any other "conclusions" are simply nonsense.
Harmonics can only be generated when a signal passes through a non-linear device. Inductive or capacitive coupling does not produce harmonics. Semiconductors, corroded connections etc., do. But then you'd again have to analyze the circuit, to determine what harmonics are likely and again, the amplitudes from anything on a phone line are unlikely to produce significant energy at the frequencies used by ethernet.
Agreed. I wasn't saying shared sheath is dangerous to equipment, just that some packets might get ruined from interference and that shared sheath was designed around 10baseT, not 1`00baseTX.
Uninsulated, obviously, but should have ceramic stand-offs. I would think cooling would be an issue.
I'm well aware of this. But harmonics are not necessary for interference. Harmonics tend to persist and produce noise. Digital baseband signalling is highly resistant to low level noise, but will fail on very short spikes that wipe one bit. Fortunately, higher protocol levels can take care of low error-rates.
There may be none in the signal itself, but induction transforms signal energy. Again, I'm not talking about harmonics but about single events, perhaps doubling on one previous reflected event.
Such as??? Anything you can imagine, other than constant DC is composed of sine waves at one or more frequencies. Any wave form, such as square wave triangle wave, etc., are composed of sine waves at a fundement and harmonic frequencies. Noise impluses consist of short duration sine waves (sometimes called "damped" waves, in that they rapidly decay), no matter what the pulse shape. "White" noise is simply a collection of sine waves at many different frequencies. Can you tell me about anything, that can't be shown to be composed of sine wave functions???
Here's something that might provide you with some idea of what's involved.
formatting link
If you're referring to the snow on a TV screen, when there's no TV reception, you are looking at noise from other sources, than electrical appliances etc. However, this doesn't rule out the possibility of interference to the TV at stages after the tuner. For example the universe is filled with radio noise, from a variety of natural sources, including the after effects of the big bang.
Also, while ethernet carries digital data, the signal is in fact analog, in that a carrier is modulated by that data. 10 Mb ethernet uses Manchester encoding and 100 Mb uses something called "multilevel threshold-3" signalling, where there are multiple levels, depending on the data bits. If you were to look at it with a scope, you'd see a complex analog waveform. Also, if it wasn't an analog signal, that is some combination of sinewaves, with equal enery above and below 0 volts, it couldn't pass through the NIC transformers, without distortion. Incidentally, you can take a simple modulation method, such as Manchester encoding and show that it is composed of various sine waves.
I'm well aware of the harmonic structure of music instruments etc. However, the fact remains that if you have a pure sine wave, it doesn't have harmonics. Harmonics are only present, when you have a non-sinusoidal waveform, such as square wave etc. If you can't accept that, then I suggest you take some math courses.
Where on earth did you hear that? Please read up on Fourier. He demonstrated that *ALL* waveforms are composed of some series of sine waves. That "perfectly square wave" is composed of a fundamental sine wave and odd order harmonics, where the amplitude of each harmonic is inversely proportional to it's order. Again, go back and take some math classes and have your instructor explain it to you.
As I recall Maxwell's equation and the (Flemming's) left hand rule, they relate to the relationship between a moving magnetic field and induced current. Other than that, what do they have to do with a waveform?
You must have one incredibly noisy vacuum cleaner or the interference is entering at a later stage.
Actually, with 10baseT, you have data that starts & stops with each frame and a sync signal at the beginning of the frame. There is also a link test signal between frames. 100baseT sends a continuous signal. The clocking is derived from the signal and the sync signal is used to determine the bit, within a frame.
Again, what about it has something specific in regards to harmonics? Does it somehow generate harmonics? Or is it simply a method of describing the relationship between magnetism and electricity? Yes, fields change directions, which is the fundamental principle on which transformers, motors etc. work. Now, you claimed, in a previous message that Maxwell would explain the function of square waves. I haven't seen anything from you about that. How does Maxwell (or Flemming) explain the composition of a square wave??? Fourier certainly does.
Why only harmonics? There are other kinds of interference.
Au contraire. You see snow crackles all the time. They just aren't very objectionable in an analog transmission so long as they're not as continuous as a vacuum cleaner! But ethernet _IS NOT_ analog.
Even 20 pps has a "wavelength" of 10,000 km in TP cable. There will be no harmonic reinforcement of interpulse noise. Even rising edge and trailing edge are probably too far apart.
formatting link
This is one way (not the only way) to deconstruct signals. But a perfectly square wave requires extremely high frequency components at very significant power to sharpen the shoulder and produce the very quick decay.
A near-square wave can also be analysed analytically from first-principles (Left Hand Rule & Maxwell's Equations). I'm suggesting these are more appropriate than the simplifications used for carrier-wave (your harmonics?) analysis.
Yes, I'm aware of how _regular_ waves can be built this way. But digitial baseband transmissions are not regular. They are a close to random pattern of ones and zeros.
No, I'm refering to the effect sometimes seen of a noise generator (vacuum cleaner) producing dropouts (snow) on a tuned TV channel. 60 Hz interfering with MHz (transmissions) or kHz (raster).
True for things like 10broad3. Not AFAIK in 10baseT, 100baseTX or GBE.
I've seen a sorry mess that looks more digital than analog.
Which have to start and stop. AFAIK, the encoding is to carry clock so there is no question "Was that 5 zero bits or 6?"
_EVERYTHING_! When electrons/holes move, they create magnetic fields. When they change directions, those magnetic fields change directions. This produces the changing magnetic field which induces current. The waveform matters because it is the rate of change of current that controls the rate-of-change of magfield which controls sets the current induced.
Which is a major risk.... I never ever use internal {PCI bus} modems. An external modem w/ serial port gives you an excellent sacrificial lamb to Murphy.
Induce a large spike onto the Ethernet pair on your machine, and hope that the standoff insulation saves your butt.
If you think the protector the telco provides will save your Ethernet; keep spares around. Using the same cable is like painting a bullseye on your butt and going to Pamplona. It works, and you may be luckly for a long time...but...
"Going on/off hook (or pulse dialing) might be, especially if there isn't parasitic capacitancy or deliberate contact capacitors to round the shoulders. You also have to cope with "bounce" from mechanical contacts. "
and
"Harmonics? We aren't talking carrier-wave! All it takes is a single spike to ruin one bit, then the 12 kbit packet is shot."
and
"But an old mechanical dialer might be a different story. I think a guy from Finland posted with his tests."
In those comments, you implied that common activity on a phone line would cause interference. Every desired signal on a phone line is below 4 kHz and ethernet signaling is in the several MHz range. This means that the only way one of those signals on a phone line could interfere with ethernet, would be if it were to generate significant energy (harmonics) in the frequency range that ethernet is operating in. The only way that could conceivably happen, would be if those signals generated harmonics in the range that ethernet is sensitive to. Your comment about "All it takes is a single spike to ruin one bit, then the 12 kbit packet is shot." shows you do not understand the nature of what's on the wire. What is that pulse? If it's a square wave of very short duration, it's still composed of sine wave components. Also, what on a phone line could generate signals in the several MHz range? Dial pulses won't do it. Hanging up the phone won't do it. The voice energy won't do it. Defective ADSL equipment, operating without a filter might. Please explain how the interfering signal would be generated and propagated, in a manner that stands up to mathematical or electrical engineering scrutiny.
Remember, you not only have to generate such a signal on the phone line pair, you've got to get it to couple into the ethernet pair, as a differential mode signal, that can pass through the line transformer and whatever filters are in the NIC. Just claiming "All it takes is a single spike to ruin one bit, then the 12 kbit packet is shot" doesn't say much, unless you can explain how. What does that spike look like, on the ethernet pair? What frequency range does it occupy? What amplitude at the ethernet receiver? Further, how could any phone device that generates such interference get FCC approval for connection to the phone line? If it's producing such interference to ethernet, it might also do so to various radio services.
Most computers are connected to a power line and phone or cable TV line or both. Any of those three connections can bring hazardous voltages to the computer. My point has always been that simply sharing a cable, with no electrical connection between the phone and ethernet pairs, is not likely to create a significant increase in risk. Any harmful voltage or current on the phone pair, has to be coupled into the ethernet pair, which will result in significant attenuation of that harmful voltage or current, simply because twisted pair is designed to avoid such coupling. Then it has to overcome the protective circuits of the NIC, such as the ethernet line transformers etc., to get to a point where it can cause damage.
I don't think so. I believe the TV snow from the vac cleaner is coming radiating from the power lines and either leaking thru the TV PSU or is picked up by the antenna/gain circuitry. The interference isn't dependant on distance between TV & vac.
Why do you keep insisting on harmonics? They aren't the only cause of interference.
Fourier describes a synthetic decomposition but says nothing about effects. Flemming describes and Maxwell quantifies the inductive _effect_ [result] of a square edge.
Any signal carrying intelligence involves changing some aspect, such as voltage, frequency or phase. If you have a digital square wave on a wire, you will have a sine wave, at the frequency of the square wave or fundemental frequency, along with another at 3 times the the fundemental frequency and 1/3 the amplitude and another at 5 times the fundemental at
1/5 the amplitude, etc to infinity (assuming infinite bandwidth). While you may have a digial circuit, you have a lot of analog signals. You can then take a square wave at, for example 1 kHz. You will see the various sine waves, as described above, at 1 kHz, 3 kHz, 5 kHz etc. If you then pass that square wave through a perfect low pass filter, with the cut off frequency somewhere between 1 kHz and 3 kHz, you will remove all the harmonics and have only a 1 kHz sine wave left. Further, since ethernet NICs are connected to the cable via a transformer, the signal must be analog.
Ethernet uses modulation to create a signal that can transit the cable, NIC transformers etc. Please read up on Manchester encoding, as used in
10baseT. Incidentally, to pass a signal through a transformer, without distortion, you need a signal that is balance, with equal energy above and below 0V. If there is any DC bias in that signal, it will be lost. For mathematical proof of this, read up on differential calculus.
I'll describe pulse dialing, as it's a simpler case than the hook switch, because it's a periodic function. With pulse dialing (assuming 50 % duty cycle for simplicity), you will typically have a 10 pulse per second, or 10 Hz, square wave. That square wave, as I have described before, is composed of a sine wave at 10 Hz and a series of harmonics. This is a fundamental fact, proven both mathematically and in experiment. And again, as has been irrefutably proven over the years, the amplitude decreases in proportion to harmonic order. There is absolutely no way, for a square wave at 10 Hz, to generate significant energy levels at ethernet frequencies. It is physically impossible. Now when you take the hook switch contacts operating, you have a similar situation WRT Fourier analasys, but since it's not a periodic function, the calculations become more complex.
So, yes, contacts can generate noise at higher frequencies, but by the time you get to ethernet frequencies, there is simply not sufficient energy to cause interference. In order for such pulsing to have sufficient energy in the range of frequencies used by ethernet, it would have to be of extremely short duration. You will then find circuit capacitance and inductance will not permit sufficient bandwidth for such a pulse to exist. That pulse will then be much broader and lower amplitude, due to the low pass filtering, caused by that capacitance and inductance.
Situations such as dialing and hook switch operation are very easy to repeat. Either the phone creates such interference or it doesn't. Again, if the phone managed to produce signals capable of interfering with internet are produced, they won't get far, due to the low pass filter effect, that exists in every circuit.
You have yet to describe any source of such interference. If you still believe that pulse dialing or hook switch operation can, you're proving that you don't know what the hell you're talking about.
True on the power; true for some modems [internal vs external; shared vs. separate power supply, etc] and never true for cable -- the cable talks to an external cable modem {in every case I have seem..} and it uses Ethernet to talk to the PC.
And that's where we disagree. Yes, it's separate pairs, but in CLOSE proximity to each other, esp. at each jack. And yes, the telco has a protector, one whose original function was to keep them from being sued over folks who were on the phone when the pair took a hit; NOT to protect IC's in end-use equipment.
And the 1500v isolation of a NIC transformer is a joke when you talk lightning-induced surges.
As I said, you can likely get away with it.....but...
What do you mean by " reflection and reinforcement"? If you're referring to standing waves, they don't change the frequency of the signal. What they will do is create voltage (or current) maximum and minimum nodes at periodic intervals along a transmission line and then only if the line is not properly terminated.
Please explain how a cable can generate energy. Last I heard, you can't create or destroy energy, except in a nuclear reaction.
How does a cable take a signal below 4 kHz, and convert it to something in the range of several MHz?
Harmonics aren't the only way to generate energy. As a rule, harmonics allow energy levels to build by reflection and reinforcement.
My overall point is that the effects of noise are different on digital and analog signals. Digital is _NOT_ immune to noise. It just tolerates low levels continuously with impunity. Bug digital is vulnerable to spikes/high noise levels, even if fairly rare. Analog rides these through but is vulnerable to continuous low levels.
Baseband signalling voltage levels. No carrier wave.
Then why snow on TV from running vacuums? Any sharp edge has lots of UHF components.
Both of these may be mechanical contacts that produce very sharp edges (and bounces).
OK. You have a set of POTS contacts (either pulse dialer or on-hook) that are open but have voltage across them. Then they are mechanically closed (and probably bounce open & closed a few times in a ms or so). Current suddnly starts to flow, rapidly building a magnetic field. This change in magnetic field induces a current on all nearby conductors. Twisted pair is good because the two conductors are balanced (both generating and receiving pairs), but it isn't perfect, and magnetics follow an inverse square law.
If this generated current is sufficiently high (differential) and long, the ethernet tranceiver transformer will register the noise, not the data. This ruins the bit, and the packet.
Extremely easily. I'm talking about intermittant noise, not continuous interference. Hard to detect transients, especially in analog signals.
The last time I checked, the cable creates its own energy by acting as an antenna. Nearby radio stations can induce signals that are strong enough to interfere with data signals. This is less true for well balanced cables, but it can still happen. The world's not perfect, unfortunately, as theory would have you believe.
Corrosion and/or rust on a contact can cause rectification and detection of signals, making them mix and/or change frequency. Again, the world's not perfect, unfortunately, as theory would have you believe.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.