microsoft.public.windowsxp.general added, since a couple weeks ago I
asked them about readjusting my TCP parameters. At most I changed
MTU then to 1492 from 1500 using Dr.TCP, but another program or
webpage continued to show it as 1500. So I don't konw if I changed
it or not, or if I change it and it reset, and my speed didn't improve
or get worse. . Nothing else was attempted to be changed, nor did I
use the Verizon optimizer, which claims to change those parameters
correctly, but I don't know if it gives a way to change things back. .

But now the problem seems to be solved!!!!

On Thu, 22 Dec 2011 11:25:58 -0800, (Dave Platt)
wrote:

In article ,
micky wrote:

Will flat, modular, outsde the wall phone wire run DSL for 50 feet?
Or do I need thicker wire?

I have consistently slow download speeds on low-speed DSL based on the
Verizon line-test and www.speedtest.org, and the one weak spot in my
home wiring could be the wire from the outdood NID to the Verizon
modem. I"m not using the wiring that came with the house***.

Right now, I"m using some *indoor* but outside the wall phone wire,
the flexible kind that's meant to work with modular plugs and go from
the wall socket to the phone. But I have it from the NID, running
loosely up the side of the house*** to a window which is shut but not
very tightly, to the hall and back into the next bedroom with the
modem, a total of about 50 feet.

Is OUTSIDE THE WALL wire inadequate for such a llong length, 50'?

Maybe. Maybe not. I wouldn't use it myself unless I had no
alternative.

The only way I can see to really test it, is to move your DSL modem to
your demarc point (NID) and connect it there,

That's a great idea that you and Mike both had. I wish I had been
pursuing this when it was warmer out and the days were longer. A
warm day like today I had to work in the morning, and the weekend will
be colder.

and try a download with
all of your house wiring disconnected.

I didn't realize until you said this that after my last rearranging, I
have only one base station phone next to the modem (and 3 wireless
phones) connected, None of the original house wiring has been
connected for months.

Jumping ahead, I found your post as a whole encouragement enough to
think the wiring might be the problem, so since it was warm, I just
replaced the wire with the round somewhat stiff in-the-wallf stuff,
instead of the jack to phone wire that I had been using for a 50 foot
run.

Using www.speedtest.net, that I had used before. My upload speed was
50 to 85% of what it had been, but my download speed was 2.5 to 3
times what it had been with the same test!! Over 6 new tests.

Using the Verizon provided test -- I wonder if this works for those
who don't have Verizon --
http://my.verizon.com/micro/speedtest/broadband/
my upload speed was the same as it had been but my download speed was
about 3.3 times what it had been with the same test, over 4 new tests.

Hoop-de-doo.

Speedtest.net also measures ping time, which started out very high but
eventually got quicker than it had been. I presume it only pings ..
when? Once at the start of a down or upload? So that's not a major
factor, right? And I don't really care about upload times because I
upload far less, mostly just text for newgroups and email, and I'm not
sitting their waiting, usually.

So everything is great now!!! The wire seems to really make a big
difference. I should have pursued this diligently when I first got
DSL, 4 years ago. (I hate to admit it, but I've been depressed)

I wonder if it slowed down my dial-up too, before then.

They say I should get 500 to 1000 KB download speed. And finally, I'm
in that range instead of 220 at most.

speedtest.net tests
Ping DL in KB UL in KB
410 550 070
494 600 110
221 540 080
98 590 070 Different Baltimore server.
300 520 110 Comcast
91 670 110

Verizon tests
.738 140
.. 734 141
734 140
740 138

I'm thinking that the Verizon values are consistently higher because
there are fewer links in the data chain, right? The data doesn't
have to go via the net to or from comcast.


At each end of the round 4 conductor wire, I attached a surface mount
modular telephone jack that I had saved, and used an 8-inch piece of
modular wire to plug into the NID and the modem. At the NID, I put
the jack box in a ziploc bag, and tomorrow I'll tape the bag even more
shut than it is now.

I didn't cut the wire yet, until I saw if it worked, and I have about
20 feet or more folded back and forth in 10-inch sections. I suppose
that might be lowering my speed also, right?

I read your post and Mike's and if there is anything i should answer,
I plan to do that tomorrow, but right now I want to find a movie or
Youtube thing I can watch, or one of those Republican debates that I
missed when they were live amd my DSL barely worked. .

I watched one video segment from C-span.org before measuring my
speeds, and it was flawless.

Thanks to you both, and William. And Paul and Paul and Char and
everyone from XP who helped me before.


Then, try it with a 50-foot
length of your flat modular cable, and see how your speeds behave.
Start reconnecting devices in the current house wiring, and see what
happens.

With many DSL modems and routers, you can get status reports on the
DSL connection (either via UDP datagrams transmitted by the modem, or
by accessing a built-in web-server interface). The figures I see on
my modem include the upstream and downstream bit-rates, the signal
levels and/or signal-to-noise ratio, rate of CRC errors or other
packet problems, the number of DSL renegotiations and rate changes,
and so forth.

If you find that you've got lousy performance (either raw-DSL, or poor
download rate) with your modem connected at the demarc, and no house
wiring at all attached... start griping at your phone company / ISP.
If you find that the 50-foot length of modular cable hurts performance
(maybe increasing the signal attenuation?), replace it.

You might find that the presence of one or more of your existing
phones hurts performance. This could be the case if you haven't
installed microfilters on those devices (they could be loading down
the DSL frequencies). It might be the case even if you *do* have
microfilters... you could get better results if you installed a
full-sized box-mounted "DSL splitter" right at your demarc, and run
separate lines to the DSL modem (from the unfiltered port) and the
house wiring (from the filtered port).

As to the wire itself... my recommendation these days is to do all
inside phone wiring using real network-grade cable... CAT 5 or better.
I no longer use the old non-twisted four-wire "station cable" for any
new runs, and I replace it with CAT 5 when convenient. There's one 6'
length of modular cable between my DSL modem and the unfiltered jack
for it, but the run from the jack back to the demarc is CAT 5.

The reduced attenuation, and better interference rejection of a good
twisted-pair cable is worth the trouble, I think... after all, DSL is
low-frequency RF and deserves to be treated in the same way as (e.g.)
10BaseT Ethernet.