What would be the fastest dial-up access speed if phone systemshad 144 dB of dynamic range?
in [DSP]
|
Prev: What would be the fastest dial-up access speed if phone systems had 144 dB of dynamic range?
Next: existance of solution of a (mathematical) problem
From: Tim Wescott on 16 Nov 2009 16:31 On Mon, 16 Nov 2009 15:19:58 -0500, Randy Yates wrote: > "Green Xenon" <green_xenon1(a)yahoo.com> writes: > >> Hi: >> >> The following is not a homework question. It is a question of my >> curiosity. >> >> Currently dial up internet access has a maximum theoretical speed of 64 >> kbps. This is due in part to the 48 dB dynamic range of telephone >> systems. >> >> If telephones systems are updated so that they have a dynamic range of >> 144 dB instead of 48 dB, what will be the maximum theoretical speed of >> dial-up internet access? > > The easy answer is provided by the capacity formula > > C = W * log_2(1+P/N) > = 4000 * 144/log_10(2) > = 1.2 Mb/s .... and is, unfortunately, wrong. Without digging out my capacity formulae, 48dB is equivalent to 8 bits, and 144dB is equivalent to 24. So assuming the original statement were correct you would triple the capacity, to 192kbps. But that would only be if the capacity of dial up were actually 64kbps, which it isn't, or wasn't the last time I looked -- dial up maps to 8 bits (not linearly), except that every once in a while the phone system 'steals' the least significant bit for signaling. You don't notice because they did such a good job with companding and such, but it is really missing, at a rate of something like 1/8 or 1/16th of the time. So really fast dial-up modems are more like 56kbps, on a good day, _and_ they make that happen by figuring out how to hit the bit patterns of the D/A converter at the line card in the central office fairly exactly. Not that you'd actually see 24 bits even if you increased the quality of the D/A converter -- 24 bits of precision is well below the threshold of seeing 2nd order effects, and into the really weird stuff. -- www.wescottdesign.com
From: Jerry Avins on 16 Nov 2009 21:07 Green Xenon wrote: >> Not that you'd actually see 24 bits even if you increased the quality of >> the D/A converter -- 24 bits of precision is well below the threshold of > >> seeing 2nd order effects, and into the really weird stuff. > > > But isn't it possible to build phone systems than can handle 144 dB [24 > bits] of dynamic range? Yes. It is also possible to build phone systems with a 20KHz bandwidth. Who would pay for it? > Also, what are the 2nd order effects? > > I ask because I like the nostalgia of dial-up access but would like for it > to be faster than what it currently is. > > As I write, I'm brainstorming about dial-up internet access in which all > the devices involved have a dynamic range of 144 dB and a bandwidth of 20 > kHz [the same as human audio perception] as opposed to current dial-up > internet devices which have a bandwidth of only 4 kHz. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
From: Jerry Avins on 16 Nov 2009 21:09 Green Xenon wrote: >> Green Xenon <green_xenon1(a)yahoo.com> wrote: >> >>> The following is not a homework question. It is a question of my >>> curiosity. >>> Currently dial up internet access has a maximum theoretical speed of > 64 >>> kbps. This is due in part to the 48 dB dynamic range of telephone > systems. >>> If telephones systems are updated so that they have a dynamic range of > 144 >>> dB instead of 48 dB, what will be the maximum theoretical speed of > dial-up >>> internet access? >> 144dB sounds like more than the electronics, and especially the >> electromechanical parts can do. Do they also increase the sample rate? >> >> The current phone system is mostly a digital 64kb/s system so obviously >> the maximum rate is 64kb/s. >> >> -- glen >> > > A dynamic range of 144 dB doesn't necessarily mean the equipment produces > signals that loud. How loud? Range is difference between maximum and minimum, not either one alone. > Let's say the updated telephone systems have a dynamic range of 144 dB and > a bandwidth of 20 kHz [as opposed to 4 kHz]. What would be the maximum > speed possible of the dial-up internet access? Who cares? Aint gonna happen. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
From: Jerry Avins on 16 Nov 2009 21:38 Green Xenon wrote: >>>> Green Xenon <green_xenon1(a)yahoo.com> wrote: > >>> Let's say the updated telephone systems have a dynamic range of 144 dB > and >>> a bandwidth of 20 kHz [as opposed to 4 kHz]. What would be the maximum >>> speed possible of the dial-up internet access? > >> Who cares? Aint gonna happen. > > I'm just curious, since I like the nostalgia of dial-up Nothing stops you from using dial-up now. 56 Kb is part of the experience. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
From: Eric Jacobsen on 17 Nov 2009 00:34
On 11/16/2009 6:35 PM, Green Xenon wrote: >> "Green Xenon"<green_xenon1(a)yahoo.com> writes: >> >>> Hi: >>> >>> The following is not a homework question. It is a question of my >>> curiosity. >>> >>> Currently dial up internet access has a maximum theoretical speed of > 64 >>> kbps. This is due in part to the 48 dB dynamic range of telephone > systems. >>> If telephones systems are updated so that they have a dynamic range of > 144 >>> dB instead of 48 dB, what will be the maximum theoretical speed of > dial-up >>> internet access? >> The easy answer is provided by the capacity formula >> >> C = W * log_2(1+P/N) >> = 4000 * 144/log_10(2) >> = 1.2 Mb/s > > What do "P" and "N" stand for? In a diode, the P-type and N-type materials result from doping the semiconductor with impurities that affect the conductivity. The amount of each used will have a strong influence on the ability to carry information. -- Eric Jacobsen Minister of Algorithms Abineau Communications http://www.abineau.com |