49 Kbps limit on phones?

Someone told me that the standard phone line can not carry more than 49Kbps. Is this true, if not, then is there are a cap on the ability of the phone lines?

thanks.

64k is a default phone line (DS0). ISDN is basically 2 phone lines and so 128k. DSL is a different beast in that the card at the central office where your phone line eventually winds up is built differently from a typical phone line card.

There’s likely overhead in there pushing you below the maximum too.

The limitation is regulatory, not technological. The FCC has (or at least had) an imposed 53Kbps maximum designed to keep bundled telephone wires from interfering with each other. Dedicated lines like ISDN or DSL can get away with higher rates.

Heck, there isn’t any physical reason you can’t crank 1gbps or more through a standard telephone line, but if everyone tried it, there’d be so much cross-talk that the whole system would be rendered useless.

Physically the phone line is capable of carrying way more than 49 Kbps as is proven by the mere existence of DSL which uses the phone line.

Now, as a practical matter, each line is different and the limits will depend on the distance from the central office, the quality of the line etc.

Single channel ISDN is 64 kbps, but it’s digital. You can’t get 64 kbps with an analog modem.

IIRC, 56k modems take advantage of the analog to digital conversion that takes place on most phone lines[sup]*[/sup], and they don’t work as well with other lines. I doubt modems will ever get any faster than 56k, and they won’t even get up to 56k unless the FCC changes the limits on the amount of power you can send over a phone line.

  • The conversion means that volume is quantized - all volume levels will be rounded off to certain increments. If you listen to your modem connecting, you’ll notice a tone that starts out very quiet and becomes louder and louder for about a second. That’s the sound of the modem figuring out exactly which volume increments it can use.

At some point in the connection (poss. unless you’re calling someone connected to the same CO as yourself) your connection is converted from analog to digital. Each analog connection is allocated 64kb/sec in this conversion (1/24th of a T1). Obviously you can’t push more than 64kb over a 64k line, at least not without compression. Once you count overhead, conversion loss and such, you’re down to 56k (regulated down to 53k as mentioned above). The physical wire has no such limitation, which is why DSL, which originates at the CO, can achieve such high rates.

Running with Scissors appears to be “running with scissors”, and offers enough truth to be convincing, but ultimately, falls short of the right answer. Grey is mostly right - 64 Kbps is a DS0, which is part of the digital signal hierarchy, but it is not necessarily a “default phone line”. ISDN Basic Rate Interface (BRI) is divided into two 64 Kbps channels (b channels - data bearers) and one 16 Kbps channel (d channel - used for signaling), a total of 144 Kbps. But ISDN BRI is quite rare these days, and is quickly become an endangered species.

sailor’s description is perfectly accurate - about xDSL technology. DSL requires special equipment at both ends of the line, not just a standard phone set and phone switch.

And you can’t get more than 64 Kbps from a 64 kbps channel, even with compression. Compression is simply an algorithm that improves effective throughput, not transmission rate (although when modulating an analog waveform, the distinction is meaningless).

A DS1 (aka T1) is a 1.544 Mbps digital pipe. In order to get DS0 channels on the DS1, you can divide it into 24 data channels OR 23 data channels and one signaling channel. The latter is known as ISDN Primary Rate Interface (aka PRI). With the former, you must take 8 Kbps per channel for signaling, and you end up with 24 64 Kbps channels, but each channel is only capable of transmitting 56 Kbps of data. There is no “conversion loss”. (Please note: the details of this paragraph are a bit different outside North America)

But I get the impression the OP is really asking about modems, and not “phone lines”, or at least digital transmission schemes. Modems convert the digital (binary) language of your computer to analog waves for transmission across phone lines. Analog modems can be quite forgiving and can operate on practically any phone line in the world, unlike DSL. The limit of an analog modem on a standard telephone line is derived from an early standard for telephony equipment.

Modems of the 80s and early 90s were full duplex, transmitting the same amount of data in both directions simultaneously. Modern modems are asymetrical, approaching a theoretical 53 Kbps limit (of the standard, not a regulation) in the downstream direction only, and limited to 33.6 Kbps in the upstream direction.

For all the bloody details, this link does a good job of describing it all (along with other pages of the tutorial on ADSL). It provides the formula for calculating the limit, and also explains about the digital capacity of twisted pair phone lines.

So, if someone was really talking about the lines themselves, I wouldn’t rely on that someone for information too much. If it was in reference to using analog modems across phone lines, they were essentially correct.

Just a note, modern modems use the v.92 standard which allows an upstream of up to 48kbps. This is one of several changes from the original v.90 standard.

Clearly, slim2none didn’t bother to read all the bloody details. To believe it is a sham perpetuated by the landline phone companies would be to believe that every phone company in the world is part of a coordinated conspiracy which requires them to invest billions of capital to enable DSL services across infrastructure, that in many places, is approaching 100 years old.

As for Alereon’s point, you can read more here and here.

I see. And all the phone companies and all the governments of the whole-world-wide-world are participants in this sham. Yes?

Can you explain in detail how can a program in your computer push more than 56kbaud through the phone line? Can you give us specific examples of such programs?

I am not impressed with your terminology. You aren’t an electrical engineer aren’t you?

Again, all the goverments of the world are participants in this scam, aren’t they? Europe, America, China, even Saddam Hussein, they all agreed that the consumer just had to be screwed?

We all look forward to your technically detailed explanations.

All of them except Saddam Hussein. That’s why he had to be taken out!

Yikes. I thought that DSL worked by pumping a different frequency through the line, to use a technologically-retarded layman’s term, one above that which interfered with normal voiceline commerce but which was necessarily fainter, hence the distance constraint.

And furthermore, I thought that the 53Kb constraint on “regular” modem data transfer was implemented due to the possibility of fire hazards. Or at least someone was sold on that idea, thanks to U.S. Robotics’ lobbyists.

Since this is a definite “tell me I’m wrong” thread, please tell me why this is wrong–it’s kind of important.

Sofa King, DSL (and its variant ADSL) does, indeed, use higher frequencies. As has been said, the copper twisted pair will normally be able to carry much more than 56K. It is once you get to the central office where you have the problem. So, if you have an ADSL modem at each end of the line, before the signal enters the switching equipment, then you can get much more throughput. Again, if the line is very long or very bad quality, then maybe you cannot get DSL to work but normally the bottleneck is at the centrl office, not the capacity of the twisted pair.

Thanks for the answer, sailor. Could the distance still somehow be related to the resistance in the wire, like one of those fancy UDMA cables for your hard drive?

And while we’re at it, what’s up with the twisted pair? Was it originally for redundancy or something? It seems to me that such a line would naturally create its own interference.

SofaKing, the link I provided in my first post touched on both of those issues - the DSL issue on subsequent pages.

The “fire” issue is perhaps exaggerated, but generally accurate. Back before 1969, no one but the phone company could connect a device (like a phone) to the telco network. In '68, the now infamous Carterphone decision began to break up the telephone monopology, arguing that companies other than Western Electric (a division of Ma Bell) could supply telephone equipment and connect it to the telephone network. To protect the multi-billion dollar investment of the telephone company switches and cables from dumbasses that might try to pass high voltage down a small guage wire,the phone companies installed filters where the copper lines enter the Central Office. Without the filter, it would be possible for an end user to start a fire in a telco CO (one of the worst telephone service outages in the US was due to a CO fire). But the bigger risk is having someone fry the electronics (without fire) on the interface card that terminates the copper wire on the CO switch.

As for twisted pair, you need two wires to form a circuit. When you phone is on the hook, the circuit is open. When you pick up the phone, the circuit closes. The network detects this change, and applies dial tone.

The twisting is done to reduce interference.

Hope that helps.

Yes, longer distance means worse conditions due to resistance, capacitance, noise and other factors. But, as I said, in general terms the wires are capable of carrying much more that 56K. The problem is that the switching equipment and all the transmission equipment was/is designed for voice and limits the bandwidth to that of a voice channel because it would make no sense to have greater bandwidth when all you are transmitting is voice.

You need to be able to send a ring signal to the phone. You need to be able to dial from the phone to the switching station. Then you need voice in both directions. You’ll also need a return path since the earth works as a fairly poor return for electrical signals of this type (not that it’s never been tried). That’s 5 wires.

Very few people outside of us engineering geeks have any appreciation for the creativity required to get it down to 2 wires.

There’s a 48 volt DC bias (current limited) on the line. When you pick up the receiver it closes a circuit which causes current to flow, so the switching office knows you picked up the phone. The ring signal is alternating current shoved onto the same 2 wires. You dial out by either breaking the 48 V DC connection (the equivalent of briefly hanging up and picking the phone back up very quickly) to make pulses (pulse dialing), or by sending mixed tones of 2 specific frequencies per button (tone dialing). The voice coils, which modulate the current from the 48 VDC source, are a really tricky arrangement that allow the incoming signal to go to your ear at a much louder volume than the outgoing signal from the mouthpiece, which is on the same wire.

There’s no redundancy to it. Instead it’s a bunch of crap all squished down to 2 wires, and is a very elegant solution considering what was available with technology at the time.

Another thing they started doing very early on was multiplexing signals so that you could carry multiple voice channels over a single wire. This drastically reduces the number of wires that need to come out of the switching station going towards other switching stations. In the early days they used simple analog multiplexing, which was simply modulating the base signals up to a higher frequency and mixing them together on the same line. These days it’s mostly done digitally, although there are still many old analog sections in most phone systems. It’s the same idea though, and the bandwidth of each voice channel has to be limited or else it won’t fit in its frequency allocation for the multiplexing (or in the digital bit stream if it’s digitally multiplexed).

You end up with a 56k limit because of the freqency multiplexing. Think of it as being very similar to the limits of TV broadcasting. Channel 3 has to use standard frequencies for video broadcasting. If they tried to double their frequencies (to maybe broadcast a higher resolution TV signal) then they would interfere with channel 4. If they tried to increase their bandwidth by 4 times, then they would interfere with channels 4,5,6,and 7. It’s the exact same thing with voice modems. If you try to increase your bandwidth, you can easily increase the bandwidth to the switching station, but you can’t increase it beyond that because it exceeds the voice channel allocated for it. You would start interfering with other voice channels.