Interferon <Repl;y...@This.NG.Please> wrote in message
> ----- Original Message -----
> From: "Hooda Gest" <be@one_with.com>
> > Interferon <Repl;y...@This.NG.Please> wrote in message
> > news:firstname.lastname@example.org...
> > > HP Pavilion 8670c -- PIII 600 MHz 128 MB
> > > The box, new last July, came with a Connexant Softk56 PCI modem
> > a
> > > "Chameleon" board with the PC's sound chips.
> > > The problem mainly is difficulty connecting (2-3 tries, sometimes
> > > followed by line drops 10 minutes or more later. Each new call costs
> > > nickel, and over a month that adds up.
> > > This evening, when I was able to connect, it was as low as 13k (I
> > > disconnected on seeing how slow it was). Right now it's 44k, which is
> > > acceptable.
> > Sounds like your line is unstable or you have problems with some, not
> > of teh ISP's modems.
> Or sunspot activity? I've just posted a query on that. BTW, I stayed
> connected for 10 hr. yesterday at 45,667.
Unstable lines, as I pointed out later in that post, can be caused by a lot
of things. These problems can come and go at any time and last for short or
long periods. That's why I refer to them as "unstable" lines rather than
Sunspot activity would be about a 1 in 100 billion shot as a cause.
> > > Since I used TweakDUN to set my MTU at 1524 bytes (default for Ascend
> > front
> > > ends -- my ISP tech help didn't even know what an MTU was, so I had to
> > > search the net for the info), I've sometimes connected as fast as
> > Changing your MTU has no effect on the modem's ability to negotiate a
> > connection speed or even a connection at all.
> Well, something boosted the max speed I've seen. And supposedly it
> the throughput rate.
Changing MTU size can improve throughput with systems running an older
version of DUN (1.1 or 1.0) under some conditions. It is not needed, or
desired, in systems with Win98 (which has DUN1.3) or Win95 systems with
updated DUN and Winsock.
> > Since you seem to notice a relation between the new furnace fan and the
> > noise, check your phone wiring and see if it crosses any electrical
> > If it does, that may be the source of the problem. If the new furnace
> > required the installation of new electrical wiring, the person doing the
> > work may have caused a problem by running the electrical wiring
> > across the phone wiring or damaged the phone wiring in the process.
> The line from the telco box to my bedroom runs outside of the house and is
> partly shielded by aluminum siding. I had a problem a year ago on this
> of loud static, which they came out and fixed.
So, there is a history of problems of a similar nature.
> > It may have been coincidental to something else, though. If the modem
> > the line with other devices (answering machine, fax machine, other
> > remove them from the line and see if that helps.
> No fax. Only one other phone, in the living room -- a cordless.
Is the phone on the same phone line as the modem? I do not mean attached to
the same jack, I mean using the same phone line. If it is, a cordless phone
can cause the problems you are having.
> > You can bypass your inside wiring by moving the computer to a window
> > the telco interface box and running a phone cable direct to the test
> > the box. So far as I know, there is no way to test for line noise
> I'm going to try your suggestion of a direct cable to the test jack.
> already had a looksee inside the box.
Some static/noise problems can be caused by corrosion buildup on the
connections inside the telco box if the box doesn't seal tightly. That type
of problem is more constant, however.
> > > 3. Can anyone advise me about SHIELDED phone wire? In my component
> > > buff days I used shielded cables to exclude stray noise on the preamp
> > > inputs. Since only two wires are needed for a phone connection, maybe
> > > ordinary shielded 2-conductor between the telco box and my wall jack
> > > decrease line noise that might be picked up by unshielded wires.
> > > have any experience in this?
> > You could switch to shielded and you could try re-cabling with CAT3 or
> > telephone cable.
> CAT3/5 -- That's the phone hookup wire? Does Radio Shack carry it?
Yes, Radio Shack carries it.
> Of course if this is sunspot related, the noise would be mainly picked up
> long lines from the telco CO.
Who gave you the idea of sunspot activity being a possibility? Sunspot
activity affects satellite transmission and radio transmission/reception. It
has no effect on landline telecommunications.
> > > FWIW, I am 10,300 ft. from my telco central office. The phone lines
> > my
> > > house have to pass over/under an electric commuter railway line to get
> > > there.
> > Are you copper all the way or do you ride a SLC? If it's all copper,
> > is plenty of room for problems that could include:
> > Damaged insulation
> > Bridgetaps (or halftaps)
> > or a variety of other problems.
> I don't know -- what's a SLC?
Subscriber Loop Carrier. Sometimes called a DLC for Digital Loop Carrier.
Basically, a T1 digital carrier system for 24 separate phone lines. These
can be good or bad depending on how the SLC is interfaced to the telco
central office switch. For example, I am on a SLC which connects me to the
nearest telco switch some 15 miles away. Since the SLC is fully integrated,
I am able to get 50666bps connects (limited to 49333bps for better
throughput) every time. A bad channel unit (1 of the 24 lines) can cause
problems or a poor connection at the point the line enters the SLC can
create problems. I once fought the telco over random dropped calls on my
voice line because of a poor connection. It took them three weeks to clear
> > > 4. Sometimes I get in to the "Connected to remote computer. Verifying
> > user
> > > name and password..." part of the handshake and after 20 sec or so, my
> > modem
> > > throws up its little hands: "Failed to connect to remote computer.
> > to
> > > establish a connection." What gives here? Is it possible to INCREASE
> > > timeout length?
> > Yes, but the above would probably not be affected. What the above
> > is usually a failure at the ISP (router failed to find a link) or a
> > failure. It does not directly indicate a line or modem problem.
> Yeah, so it's the ISP's fault. Man, lots of variables in play here.
The key in the above is the fact that you get to "verifying login and
password". This means the modems established a connection. Checking post
call stats or the modem log can tell you whether error correction was
established and eliminate that as a cause.
After that, you proceed to verifying proper setup of the ISP parameters such
as fixed DNS address vs allowing the ISP to dynamically assign it.
> > > 5. Regarding 4., is there anything to be gained by buying a modem
> > the
> > > manufacturer of my ISP's rack modems?
> > That simply reduces the chances of inter-operability problems.
> Well, I think I'll ask what brand they use.
Can't hurt. I recommend Lucent chipset modems when the ISP uses Rockwell or
Lucent based equipment and USR if the ISP uses Lucent or USR based
I am not a big fan of Rockwell based modems but they are adequate for most
> > > 6. I would buy a hardware-based modem, not a Win modem. What are the
> > > issues of internal PCI vs external serial port vs external USB? I
> > two
> > > empty PCI slots.
> > There are no issues of any significance. The primary issues are between
> > cheap Windows modems vs hardware modems.
> Well, I have a cheap Winmodem. Are some brands of controller-based modems
> more tolerant of line noise?
Define "cheap Winmodem"... I have a SupraMax HCF modem which works a notch
below my USR Sportster hardware modem. It cost me $15 after rebate.
I have a Winmodem which I use on a slightly impaired line at work that
functions very well. The Supra worked equally well on that line, as did a
Courier and a Sportster.
Hardware modems, on the whole, are less troublesome than controllerless and
tend to perform better under a wider variety of conditions. But that is not
a universal truth.