From owner-freebsd-questions Tue Nov 6 12:53:34 2001 Delivered-To: freebsd-questions@freebsd.org Received: from mtiwmhc26.worldnet.att.net (mtiwmhc26.worldnet.att.net [204.127.131.51]) by hub.freebsd.org (Postfix) with ESMTP id 4F0A637B418 for ; Tue, 6 Nov 2001 12:53:23 -0800 (PST) Received: from MOBILE2 ([12.76.84.254]) by mtiwmhc26.worldnet.att.net (InterMail vM.4.01.03.27 201-229-121-127-20010626) with SMTP id <20011106205321.CSJU4964.mtiwmhc26.worldnet.att.net@MOBILE2>; Tue, 6 Nov 2001 20:53:21 +0000 Message-ID: <005e01c16705$076bb080$0100a8c0@MOBILE2> From: "SNF" To: "Bob Johnson" , Cc: References: <3BE84211.88CBE92D@eng.ufl.edu> Subject: Re: Welders causing dial-out to fail Date: Tue, 6 Nov 2001 14:53:02 -0600 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG > > lines are. There is only about 5' of phone line from the proxy to the main > > telco jack. The welders are on the other side of the building: > > > > telco conn. > > +-----+-----+-------------------------+ > > |proxy| | | > > +-----+ | | > > | | | > > | | | > > | office | | > > | | | > > | | | > > +-----------+ | > > | | > > | welders | > > | | > > +-------------------------------------+ > > > > I don't know whether this may mean different things to different people, > > but it doesn't (from my viewpoint) seem to support the "shield the wire > > from proxy to telco will solve the problem" angle. It does make a difference what type of wall is between the welders and the equipment you are refering to, but unless it is a 2 ft. thick concrete wall, resistance welders spew out enourmous amounts of RF (at least compared to the MIG welders we have) and can have some funky effects on equipment up to 100 ft. away. (Physically, I don't understand how this makes sense - the physics behind the last statement I made don't seem to add up, but experience argues against my gut instinct.) Anyway, I would definitely suggest using shielded cable (I would even consider high end CAT5E under circumstances I will comment on below) - even for 5 feet. > It may also be that the noise is coming in through the power system, > although that would sort of require that there be some powered phone > system equipment to feed it into the phone line. I doubt that it is > getting into the phone line through the computer power supply. It > is conceivable that the phone line is run close to a power line and > picking up noise that way. If you have the phone line bundled > with power lines for neatness, try moving it several inches away > from them. > > Also, if you _do_ decide to try shielded phone line, the usually > recommended installation is to connect the shield to the ground > rod at the phone company access point, but don't connect the > shield to anything else. > > Finally, this could be caused by a bad ground on either the power > wiring for the building, or the phone system. Have both of them > checked. They _should_ share the same ground system, but they > might not. This is something I hadn't thought of - which can be more harmful than the welders. On a different job, I had a case where the power to a building we were working on was fluctuating between 90 and 120 VAC. I was dealing with some pretty weird issues and after working on the problem for a few hours (some dial-in lines and 10/100 switches were randomly losing data packets, thus slowing or completely stopping network communication to that building - again, randomly), I remembered reading an article on how sensitive networking equipment (modems included) can be to slight power fluctuations. Whereas a computer sys. may not seem to appear to be effected by such fluctuations, network gear will respond pretty quickly. In the end, we ran down to the power plant for the building and found that the circuit for our equipment was going out (it was an electrically regulated circuit) and after we replaced the circuit, the problems went away immediately. If your ground, on the other hand, is the cause of the problems, then two solutions exist. The best is to have a better ground installed or find a UPS that does line noise conditioning or actually converts AC to DC and then back to AC for equipment use. The latter is quite expensive and I have found that for a small shop, something like an APC Back-UPS pro works fine. With proper power supply, a well constructed network will be fairly immune to a lot of the RF a welding facility may throw at it. In most cases, I have found that quality CAT5 cabling alone will do the job (they aren't jobs I engineered, but worked on after installation) - I would have prefered the shielded because I don't know what the long term effects of heavy RF are on the sheath and physical structure/metalic properties of the cable. The only case where I absolutely had to install shielded or in some cases fiber, was where welding robots were involved... But there are very few places that do that type of work anyway. > > The ceiling of this place is metal girders, and some of the electrical > > runs through the ceiling, so I'm wondering if the ceiling is getting > > charged or something and acting like a huge antenna that just transfers > > the interference throughout the rest of the building. > > Not so much like an antenna, but a transmission line. Once the noise > gets into the power line, it travels just like phone signals travel > through the phone line. Anything connected to the same power system > may pick up the noise. Agreed... Good luck... The best advice I can give is to think of the solution that makes the most sense from a financial and infrustructural standpoint. Going completely all out (i.e. redundant power systems isolated from the power source and thinking that shielded cable is "absolutely" required) doesn't make sense because the investment in equipment might not ever be recovered when considering the benefit of a well running system. On the other hand, if one has to constantly maintain something that should really run on its own once installed, then going a little further may in fact help and be an over all better investment. SNF To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message