For some reason, my Internet connection in Bedford is half the speed it used to
be: 28.8 instead of 56K. Right now, no DSL from anyone. No cable, either, yet.
Satellite, yes. But I'm truly not up for a $600 equipment investment, plus a
monthly dish access charge of at least $40.
I ran across a thing called NetZero, and am considering their "high speed" dial
up. Supposedly 5 times as fast as my current speed, which is still lousy,
Is anyone using this service? If so, what kind of results have you had, how
happy are you with it, does it make NG access easy or difficult?
"If you want to know what God thinks of money, just look at the people he gave
it to." Dorothy Parker
I'm not using it, nor would I ever unless, God forbid, I move out of range
of cable (which I have now) or at least DSL. But just so we're clear on the
"5 times faster" thing, the way they supposedly achieve that is that
NetZero's servers cache web content so that, for instance, when you're going
to www.nra.org, you don't actually have to get all the way to the NRA web
servers to retrieve the information, because it's likely already cached at
one of NetZero's servers. Now, if you do a lot of downloads, you're
screwed, because those will come down at the same, slow rate as before.
Regarding your current connection, have you tried other 56K numbers to dial
up, assuming there are any? For completeness, I should also point out that
the connection speed that is reported is only the initial speed that was
negotiated between your modem and the ISP's equipment. Both ends are
constantly negotiating the best speed that they can get. For an additional
helping of completeness, you never got 56K. The best you can do, in theory,
is 53K, because that is what the speed over an analog line is limited to, by
law, though I'm not sure if I ever saw a reported connection speed over 48K.
In any case, 28.8K is dreadfully slow and I hope you get it fixed, because I
sure couldn't deal with it.
it isn't an inside wiring issue. a knowledgeable tech from
a DSL provider explained to me years ago that one of the
phone line pairs is hooked up in a different manner to the
phone system, but I must say that I can't recall the
technical details. suffice it to say, the phone company
doesn't guarantee a fast connection using modems, so my
complaints to them resulting in no repair to the slow line.
I'm sure if there are any telecommunications pros here,
they could explain it to you (and refresh my memory).
If the homeowner has only one phone line coming into the house, and
extension jacks throughout the house, and some locations experience slower
connects, then it is indeed an inside wiring issue. The problem can be
caused by a number of factors including...
- polarity reversal (the red and green switched at the jack)
- cheap wire - this is very common. Most folks just use cheap Radio
Shack wire to run their extension jacks and they run them like they would an
electrical outlet (in parallel). Cheap wire is not twisted and is
susceptible to line noise which will result in slower negotiated speeds
between modems as well as more retries on packet transmission. These days
one should spend the little extra and use twisted pair for all home phone
wiring. It's not uncommon for people with two line phones in their homes to
hear cross talk between the phone lines with cheap wire.
- cheap RJ11 jacks - junk jacks will build up an oxide coating on the
pins which results in poor signal - the problems are obvious.
The telephone companies all use structured wiring schemes so all of their
connections between the Central Office and your residential network access
are hooked up in a consistent manner. It would be possible, but most
unlikely for the phone company to have hooked up any of your home lines in
anything but a structured manner.
Still, it comes down to the simple observation - if one location in your
home works well and another does not, then it's your in home wiring that is
Amen. I remember back in the days when directly connected CRTs
would run 9600 on an RS232 line to a minicomputer as long as it
wasn't over 50-100 feet or so. We bought 1000' of shielded,
twisted pair and, just for the heck of it, ran a terminal
through the whole 1000' (which was still wrapped around its
spool). Worked fine :-).
Wire, connectors, soldering, etc., quality DO make a difference!
Likewise - many years ago, when I was a tech, I had a hospital for a
customer and they ran an RS232C line 1500 feet using twisted pair and never
dropped a bit. Like you said, every manufacturer was holding to the spec of
50 feet, but they didn't bother reading the spec - they just installed the
wire and presto.
Curiosity is nibbling. If I can find one, I have 2-3 56K modems around here. I
might slip one in this machine in my spare time (whazzat? I will NEVER move
again) and see what happens. Or bring my wife's machine (the one I used to run
off this line) down here.
Changing dial-up numbers makes no difference whatsoever, by the way.
"It is even harder for the average ape to believe that he has descended from
H. L. Mencken
because the guy at the DSL provider knew from my description
that there is a common problem with some phone lines as they
are connected to the system. wish I could recall the
details. there are two pairs and one of the pairs can be
connected "differently", but like I said 3 times already, I
forget the terms he used to describe HOW it's connected that
causes the top speed to be 28.8 or less. shoot me for
forgetting, but it's been about 4 years since I talked to
for those who continue to insist it is an "inside" problem,
let me mention that ALL (that's right, ALL of the inside
wiring) was bypassed to check the speed of the connection
right where the wiring comes up from underground to a point
about 15" above the ground, on the outside wall of my
garage. so can we put that notion to bed, once and for all?
just for the sake of argument, did you use the same phone cord you were
using before? if you tried it it right where it comes to the house, with
two different phone cords, with all the rest of the house wiring
disconnected and only the one connection going to the modem, i think you can
be sure its not inside. if you only used one phone cord i think that cord
is still suspect..
Because he _knows_ what he is talking about. <grin>
Because it is *NOT* an inside wiring problem.
He has two _different_ phone lines ( each with their _own_ phone number).
One is on a bare copper pair, all the way back to the C.O. -- on *that*
line, he can get 56k. Heck, if the distance was short enough, he could
even get DSL over it.
The other line goes through a 'pair gain' device, where it is *multiplexed*
with other phone lines, *before* it gets to the C.O. Then, at the C.O.,
it is split back out to the original "lines", and connected to the switch.
On *THIS* kind of a physical connection, you are _lucky_ if you can get 28.8k.
a top speed of 19.2, or 14.4 is *not* uncommon. (Note: you _cannot_ run DSL
over this kind of a connection.)
To get a modem connection speed above 33.6k (or 28.8k, if your modem doesn't
do 33.6), there must be _exactly_*ONE*_ analog/digital conversion stage in
the _entire_ circuit from your modem to *their* pseudo-modem. *IF* there
are two, or more, such conversions, then it is simply *not* possible to get
a connection speed faster than 33.6k.
Agreed - they just say it's voice grade and tough luck. But if
I mention the problem to a maintenance guy while he's out for
something else, he'll usually switch the pairs or "change the
loading", whatever that means. Maybe impedance?
Well, 'yes and no'. <grin>
Long runs of wire have significant 'distributed capacitance', just from
the proximity of the two wires to each other.
One can counter-act the effect of that capacitance, by hanging some 'coils'
(known technically, oddly enough, as "loading coils") on the pair, so that
the inductance 'cancels' the effect of the capacitance.
For optimum signal clarity, the amount of inductance supplied by the coils
has to match the capacitance -- obviously. Too much of _either_ one results
in degraded audio. "How much" is 'just right' depends on several things;
mostly the _length_ of the wire run, but, in addition, the -size- of the
wire, the construction of the wire, and what it runs through/next to --
just to name a few of the factors.
Smaller coils, installed at more frequent positions down the length of the
wire, do a better job of keeping the signal clean than do a smaller number
of larger-inductance coils installed farther apart.
email@example.com (Robert Bonomi) wrote in
And among other things, these loading coils need to be removed in order for
DSL to work. Necessary, generally speaking, but not sufficient. Distance
has to be right. No 'bridge taps'. No other equipment in the lines, often
used to solve other problems.
Getting DSL to work in the first place took a lot of very talented
And some of them are pretty good woodworkers (he says, trying to bring some
topicality to the thread again.)
Hope the cable modem works well for you, Dave.
Hm, if that's how they're calculating it, then it's even more iffy
than I thought. Any ISP is going to be on a very fat pipe compared to
a dialup user. The time lag saved by caching the data in a local proxy
server is minimal - they do it mainly to decrease having to re-fetch
the same object dozens or thousands of times a day. Makes sense from
a bandwidth usage standpoint, but only buys a tiny fraction of the total
time compared to a bandwidth download of the same object.
Yes; there's only so much data that can go down the line at one time,
no matter how you look at it. The garden hose won't put out the
I saw a 52K connection just last night on dialup, it was the first time
I saw anything over 50. May have been a change in how it's reported,
I'm spoiled by my home connection, that's for sure. I get to someone's
house where they have dialup, and if their system needs work, it's
actually almost always quicker to drive it over to my place, install
a NIC, and use my feed to go get the patches or drivers or whatever.
Then, I play in the shop while the stuff is downloading, so it's
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It 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.