Greetings: We have a client site that is driving us nuts! We have had more equipment failure issues at that one location than at all other locations we manage combined. We have come to the conclusion that there is some type of RF interference occurring that is causing these failures, but are at a loss how to prove that and track down the source. This client site is a new facility. Incoming power is lightning protected by the power company at the meter, and each breaker panel has surge suppression breakers. Incoming telco lines are lightning and surge protected by the telco at the demarc. The building has excellent grounding. All telephone, network, and computer equipment are on UPSes. All network infrastructure is certified CAT6. We have had the power company monitor line quality and they see no blips that would explain out observed problems. The local telco consistently claims the problem is on the customer side of the demarc -- even when it effects telco equipment. The site has channelized T1 voice, an ATM encapsulated FR WAN, and ADSL local Internet access. We have this identical setup for this client at about a dozen other locations, and none have any problems. The most frequent observed problem is the ADSL line dropping due to excessive noise. The problem seems to occur most at about 0130, 0730, 1530, and 2230 local time -- but not every day and not always at those times (about 20 to 25 failures per week). ADSL failures can last from 5 minutes to hours. We get frequent voice and FR circuits failures -- but of shorter duration (usually less than 5 minutes) and far less frequently (2 or 3 times a week). The voice and FR failures appear to occur concurrently (physically separate circuits), but not at the same time as the ADSL failure. I should also add that there never appears to be audible noise on the lines, even during failures. Equipment failures is what really leads to the conclusion this is probably an RF issue. Among the stuff we have had fail are: 3 DSL routers (cisco 8x7) 1 edge router (cisco 28xx) 1 FR router (cisco 36xx) 1 patch panel 1 telco smart jack (ATM/FR circuit) 1 PBX T1 card 1 patch panel (all jacks went open on the same pair) 6+ NICs I should add that the failures started before the building was occupied -- while we were still testing the basic infrastructure. In fact, we have observed problems from day one of the networks being turned on by the telcos. I should also add some other points: -- We have observed failures when the building had zero power, except for the UPS battery power in the server room, so we don't think that we are getting power spikes from anything within the building. -- The building only operates 0600 to 1800, so many failures are occurring after hours. -- There are no RF sources in the building. -- We are not near an airport. -- The building is steel framed and sided -- and a pretty good RF shield -- cell phone reception goes from 5 of 5 bars to 1 or 2 bars as soon as you walk in the door. Given what I have described, would you think this is an RF interference problem? RF problem or not, how would you track down this problem? We are to the point of bringing in a consulting EE, but I am not sure that most would be equipped to solve this problem; so, what should we look for in a potential consulting engineer? Apologizes for the dissertation! Any and all helpful suggestions welcome. Thanks! Jon Kibler -- Jon R. Kibler Chief Technical Officer A.S.E.T., Inc. Charleston, SC USA (843) 849-8214 ================================================== Filtered by: TRUSTEM.COM's Email Filtering Service http://www.trustem.com/ No Spam. No Viruses. Just Good Clean Email.
Greetings:
We have a client site that is driving us nuts...
...
I should also add some other points: -- We have observed failures when the building had zero power, except for the UPS ..... -- The building only operates 0600 to 1800, so many failures are occurring after hours. -- There are no RF sources in the building. -- We are not near an airport. -- The building is steel framed and sided -- and a pretty good RF shield --....
Given what I have described, would you think this is an RF interference problem?
No... Unless you have a Gigawatt radar parked next door, I'm highly dubious that it's RF-instigated.
3 DSL routers (cisco 8x7) 1 edge router (cisco 28xx) 1 FR router (cisco 36xx) 1 patch panel 1 telco smart jack (ATM/FR circuit) 1 PBX T1 card 1 patch panel (all jacks went open on the same pair)
Make that Terawatt...
6+ NICs
A) All these things say grounding issues. I have to wonder if the building is fed from more than one power entrance. The blown patch panel especially makes me think the router on one end of the Cat5 was being fed from a different power source than the one on the other. (Which pair was blown?) Given the UPS mention, maybe there's a ground differential issue with it. B) The other, less likely, path into equipment is telco. Those mile-long pieces of copper from the CO are also called "antennas" and they covet static. I have no idea where this location is -- are there thunderstorms around? C) One more possibility; perhaps some piece of equipment in-house is putting large spikes on the internal distribution. Twenty years ago, I read of a building where large [50 HP HVAC] and small [fridges] motors would regularly die. The high-tech gadgets of that era, Texas Instruments calculators, would reset themselves seemingly spontaneously. After MUCH work, they found the BIG copier was putting nasty spikes back on the grid. I vote A) 75% B) 20% C) 5% You do need an EE, one prepared to look at the building wiring/grounding grid. -- A host is a host from coast to coast.................wb8foz@nrk.com & no one will talk to a host that's close........[v].(301) 56-LINUX Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433
David Lesher wrote:
Greetings:
We have a client site that is driving us nuts...
...
I should also add some other points: -- We have observed failures when the building had zero power, except for the UPS ..... -- The building only operates 0600 to 1800, so many failures are occurring after hours. -- There are no RF sources in the building. -- We are not near an airport. -- The building is steel framed and sided -- and a pretty good RF shield --....
Given what I have described, would you think this is an RF interference problem?
No...
Unless you have a Gigawatt radar parked next door, I'm highly dubious that it's RF-instigated.
3 DSL routers (cisco 8x7) 1 edge router (cisco 28xx) 1 FR router (cisco 36xx) 1 patch panel 1 telco smart jack (ATM/FR circuit) 1 PBX T1 card 1 patch panel (all jacks went open on the same pair)
Make that Terawatt...
6+ NICs
A) All these things say grounding issues. I have to wonder if the building is fed from more than one power entrance. The blown patch panel especially makes me think the router on one end of the Cat5 was being fed from a different power source than the one on the other. (Which pair was blown?) Given the UPS mention, maybe there's a ground differential issue with it.
B) The other, less likely, path into equipment is telco. Those mile-long pieces of copper from the CO are also called "antennas" and they covet static. I have no idea where this location is -- are there thunderstorms around?
C) One more possibility; perhaps some piece of equipment in-house is putting large spikes on the internal distribution. Twenty years ago, I read of a building where large [50 HP HVAC] and small [fridges] motors would regularly die. The high-tech gadgets of that era, Texas Instruments calculators, would reset themselves seemingly spontaneously. After MUCH work, they found the BIG copier was putting nasty spikes back on the grid.
I vote A) 75% B) 20% C) 5%
You do need an EE, one prepared to look at the building wiring/grounding grid.
That is already half of a solution: Go for fiber. That is imune to both ground and RF problems. Avoid ground connections between the equipment. Replace ethernet with fiber. Break serial lines with optical isolators. Cut the ground wire in your power cords but ground the equipment directly to a metal frame. Avoid ground loops: Between two computers you have a ground connection via the powerline ground. Connect them via RS-232 and you have a second connection via the RS-232 ground. If your power ground is bad then you might run amperes through the RS-232 ground that results in Volts, more than your signal level, maybe. regards Peter and Karin Dambier -- Peter and Karin Dambier The Public-Root Consortium Graeffstrasse 14 D-64646 Heppenheim +49(6252)671-788 (Telekom) +49(179)108-3978 (O2 Genion) +49(6252)750-308 (VoIP: sipgate.de) mail: peter@peter-dambier.de mail: peter@echnaton.serveftp.com http://iason.site.voila.fr/ https://sourceforge.net/projects/iason/
That is already half of a solution:
Go for fiber. That is imune to both ground and RF problems. Avoid ground connections between the equipment.
Replace ethernet with fiber. Break serial lines with optical isolators.
Yes, fiber will solve ground loop problems. And this smells like a ground loop issue. But at this juncture, I don't have enough specifics to recommend ma$$ive changes.
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
I would NEVER tell a client to do this. That could easily kill someone.
Avoid ground loops: Between two computers you have a ground connection via the powerline ground. Connect them via RS-232 and you have a second connection via the RS-232 ground. If your power ground is bad then you might run amperes through the RS-232 ground that results in Volts, more than your signal level, maybe.
I don't recall mention of RS-232,. but yes, this is the classic example of ground loops. -- A host is a host from coast to coast.................wb8foz@nrk.com & no one will talk to a host that's close........[v].(301) 56-LINUX Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433
On Sun, 5 Mar 2006 18:00:36 -0500 (EST) David Lesher <wb8foz@nrk.com> wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
I would NEVER tell a client to do this. That could easily kill someone.
Correct. The safety purpose of the ground cord is to cause a short circuit in case line voltage energizes the case, in which case the breaker will trip. If you cut that wire, the metal frame frame can become hot; unless it's firmly grounded itself, there will be a potential between it and ground. Along comes the next well-grounded person to touch it -- poof! Even if the frame were grounded properly, that's a local ground, which may differ in potential from the breaker box's ground. The neutral wire in the circuit is tied to ground at the breaker box, which means there could be a potential difference between it and the frame. That also creates a potential shock hazard, though presumably not that great. What might be useful -- ask an EE, not me -- is a circuit with an isolated ground. In that case, the ground wire from the power plug is routed all the way back to the breaker panel, and isn't connected to, say, the local electrical box that the cord is plugged into. I've seen computer equipment wired that way in the past.
The isolated grounds are definitely a recommended idea for telco/server rooms... Perhaps an array of them depending on the size power feed we're talking about. I'm assuming it's a sizeable UPS that runs your telco and data equipment (or small server room). The irritation, if you haven't done this step already, is that adding a TRUE isolated ground after you've already built your building and room is not exactly a cheap thing to do. Especially in nice metal framed buildings that like to have a tendency of becoming the nearest path ground themselves. But I agree that it's certainly something as a worthwhile "first path" to look into! Scott PS. I agree it's not good business practice to kill your clients! -----Original Message----- From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu] On Behalf Of Steven M. Bellovin Sent: Sunday, March 05, 2006 6:21 PM To: wb8foz@nrk.com Cc: nanog@merit.edu Subject: Re: Presumed RF Interference On Sun, 5 Mar 2006 18:00:36 -0500 (EST) David Lesher <wb8foz@nrk.com> wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
I would NEVER tell a client to do this. That could easily kill someone.
Correct. The safety purpose of the ground cord is to cause a short circuit in case line voltage energizes the case, in which case the breaker will trip. If you cut that wire, the metal frame frame can become hot; unless it's firmly grounded itself, there will be a potential between it and ground. Along comes the next well-grounded person to touch it -- poof! Even if the frame were grounded properly, that's a local ground, which may differ in potential from the breaker box's ground. The neutral wire in the circuit is tied to ground at the breaker box, which means there could be a potential difference between it and the frame. That also creates a potential shock hazard, though presumably not that great. What might be useful -- ask an EE, not me -- is a circuit with an isolated ground. In that case, the ground wire from the power plug is routed all the way back to the breaker panel, and isn't connected to, say, the local electrical box that the cord is plugged into. I've seen computer equipment wired that way in the past.
At 06:20 PM 3/5/2006, Steven M. Bellovin wrote:
What might be useful -- ask an EE, not me -- is a circuit with an isolated ground. In that case, the ground wire from the power plug is routed all the way back to the breaker panel, and isn't connected to, say, the local electrical box that the cord is plugged into. I've seen computer equipment wired that way in the past.
In the US, the NEC code states that the only place a neutral and a ground should be bonded together is in the primary service entrance facility or where the neutral is created. All subpanels will have isolated grounds and neutrals. If you have three phase service and use a delta (wye without the neutral) to wye transformer to create the neutral, the neutral will be bonded to ground inside the transformer cabinet. Eliminating the neutral is typically done to save money when converting 277/480V to 120/208V (no neutral means a reduced conductor count inside the conduit so smaller conduit can be used since the extra copper for the neutral is eliminated on the input side.) All grounds must be connected to the first metal box or conduit they touch. If you are using plastic boxes with Romex, your grounds will go all the back to your subpanel ground bar which will not meet the neutral until the main breaker panel. More often in a datacenter environment or a commercial facility, the wiring will be BX under a raised floor or BX or EMT with THHN overhead. Either way, the ground is connected inside the outlet box and wired directly back to the breaker panel. The bonding in the box is to ensure there is no voltage potential carried on any metal conduit. My NEC book is at the office now and I'm home, but I'm pretty sure everything I have stated from memory is accurate. -Robert Tellurian Networks - The Ultimate Internet Connection http://www.tellurian.com | 888-TELLURIAN | 973-300-9211 "Well done is better than well said." - Benjamin Franklin
On Sun, 05 Mar 2006 23:30:13 -0500 Robert Boyle <robert@tellurian.com> wrote:
At 06:20 PM 3/5/2006, Steven M. Bellovin wrote:
What might be useful -- ask an EE, not me -- is a circuit with an isolated ground. In that case, the ground wire from the power plug is routed all the way back to the breaker panel, and isn't connected to, say, the local electrical box that the cord is plugged into. I've seen computer equipment wired that way in the past.
In the US, the NEC code states that the only place a neutral and a ground should be bonded together is in the primary service entrance facility or where the neutral is created. All subpanels will have isolated grounds and neutrals. If you have three phase service and use a delta (wye without the neutral) to wye transformer to create the neutral, the neutral will be bonded to ground inside the transformer cabinet. Eliminating the neutral is typically done to save money when converting 277/480V to 120/208V (no neutral means a reduced conductor count inside the conduit so smaller conduit can be used since the extra copper for the neutral is eliminated on the input side.) All grounds must be connected to the first metal box or conduit they touch. If you are using plastic boxes with Romex, your grounds will go all the back to your subpanel ground bar which will not meet the neutral until the main breaker panel. More often in a datacenter environment or a commercial facility, the wiring will be BX under a raised floor or BX or EMT with THHN overhead. Either way, the ground is connected inside the outlet box and wired directly back to the breaker panel. The bonding in the box is to ensure there is no voltage potential carried on any metal conduit. My NEC book is at the office now and I'm home, but I'm pretty sure everything I have stated from memory is accurate.
Yes, I believe that that's correct, though I'm not going to dig out my copy of the NEC right now, either. I chose to leave out the part about separate panels. --Steven M. Bellovin, http://www.cs.columbia.edu/~smb
At 05:32 PM 3/5/2006, Peter Dambier wrote:
David Lesher wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
[ I am not a PE - IANAPE ] I don't think that is good advice. You can't possibly have the as-builts, existing condition, or line drawings in front of you. Safety first. Call the consulting engineers. -M< -- Martin Hannigan (c) 617-388-2663 Renesys Corporation (w) 617-395-8574 Member of Technical Staff Network Operations hannigan@renesys.com
Jon, Peter Dambier wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
As a time-served electrician... *****DO NOT DO THIS***** - it will kill someone. However.... You could try separate earth bonding of each components (ie connecting all the chassis together via a provided grounding terminal using nice thick copper wire), however if there is a significant earth fault even that could be dangerous (think fire) - so get a qualified electrician to do it - if there is a ground fault it will use the chassis and the bonded earths as it's route to ground. Earth faults are often easily detectable by using a digital volt meter (Note: analog volt meters do not work for this unless there is a serious fault). First check for induced and ungrounded 'floating' voltages (any AC or DC voltage above 0.05v should be investigated), then if the DVM is fused, check for any current (amps) between chassis. If you have money to spend before investigation find out if the building has a grounding stake and if not add one... A couple of meters of copper stake which will be connected to either the armoring of the supply cable (TN-S) or to the incoming return cable and installation earth PME (TN-C-S) - likely based what someone else in this threat said. In either type of grounding scheme the structure metal frame could (and should) be grounded (esp if exposed) which is likely to cause the phone RF signal drop. A faulty bonding in the structure (esp as it is steal) can also provide for some interesting ground faults as it is not uncommon to provide localised grounding to building frames. (In the UK where I served my apprenticeship, we were required to provide earth bonding to the copper plumbing system, additional bonding at every exposed fitting - this caused a few issues when plumbers first starting using PVC pipes)... All this said with the faults appearing with no external power and with just UPS supply, ground faults really do not 'fit' the problem - however if a generator is used also, you are in an IT type installation (electrical term 'IT' not 'Information Technology' ;-)) and will have to have a grounding stake on site. Please note, I am trained from the UK - laws and regulations change from country to country - get a local qualified/licensed sparky to do the work or assist you. Regards, Mat
On Mon, 06 Mar 2006 21:17:17 +1100 Matthew Sullivan <matthew@sorbs.net> wrote:
(In the UK where I served my apprenticeship, we were required to provide earth bonding to the copper plumbing system, additional bonding at every exposed fitting - this caused a few issues when plumbers first starting using PVC pipes)...
The US National Electrical Code (which has no national force of law; it's a model code voluntarily adopted by many jurisdictions) now bars grounding to pipes except within (as I recall) six feet of where the pipe enters the building, for precisely that reason.
On Mon, Mar 06, 2006 at 09:49:39AM -0500, Steven M. Bellovin wrote:
On Mon, 06 Mar 2006 21:17:17 +1100 Matthew Sullivan <matthew@sorbs.net> wrote:
(In the UK where I served my apprenticeship, we were required to provide earth bonding to the copper plumbing system, additional bonding at every exposed fitting - this caused a few issues when plumbers first starting using PVC pipes)...
The US National Electrical Code (which has no national force of law; it's a model code voluntarily adopted by many jurisdictions) now bars grounding to pipes except within (as I recall) six feet of where the pipe enters the building, for precisely that reason.
The use in modern times of teflon tape at joints in copper piping makes them unuseable for earth grounds even near the entry point to the building. A long (e.g. 2-3 meters) copper stake must be driven for a proper earth ground, or else a large copper mesh mat if the ground is rocky -- unless you are certain that the copper piping that you want to use extends a significant distance underground and unbroken. -w
On 6 Mar 2006, at 15:06, ww@styx.org wrote:
On Mon, Mar 06, 2006 at 09:49:39AM -0500, Steven M. Bellovin wrote:
On Mon, 06 Mar 2006 21:17:17 +1100 Matthew Sullivan <matthew@sorbs.net> wrote:
(In the UK where I served my apprenticeship, we were required to provide earth bonding to the copper plumbing system, additional bonding at every exposed fitting - this caused a few issues when plumbers first starting using PVC pipes)...
The US National Electrical Code (which has no national force of law; it's a model code voluntarily adopted by many jurisdictions) now bars grounding to pipes except within (as I recall) six feet of where the pipe enters the building, for precisely that reason.
The use in modern times of teflon tape at joints in copper piping makes them unuseable for earth grounds even near the entry point to the building. A long (e.g. 2-3 meters) copper stake must be driven for a proper earth ground, or else a large copper mesh mat if the ground is rocky -- unless you are certain that the copper piping that you want to use extends a significant distance underground and unbroken.
The purpose here is not to use the piping *as* a ground, but to ensure that the piping *is* at ground potential. Otherwise, if an electrical failure causes the pipe to reach a dangerous potential then so does the water in it, then so do the hands you're washing in that water. Thus if there's an electrical discontinuity in the piping it is even more important to earth bond any conductive piping/taps etc. that are on the non-earth side of that discontinuity. The same applies too to gas piping except here the principal risk is static, sparks and the subsequent explosion.
Ian Mason wrote:
On 6 Mar 2006, at 15:06, ww@styx.org wrote:
On Mon, Mar 06, 2006 at 09:49:39AM -0500, Steven M. Bellovin wrote:
On Mon, 06 Mar 2006 21:17:17 +1100 Matthew Sullivan <matthew@sorbs.net> wrote:
(In the UK where I served my apprenticeship, we were required to provide earth bonding to the copper plumbing system, additional bonding at every exposed fitting - this caused a few issues when plumbers first starting using PVC pipes)...
The US National Electrical Code (which has no national force of law; it's a model code voluntarily adopted by many jurisdictions) now bars grounding to pipes except within (as I recall) six feet of where the pipe enters the building, for precisely that reason.
The use in modern times of teflon tape at joints in copper piping makes them unuseable for earth grounds even near the entry point to the building. A long (e.g. 2-3 meters) copper stake must be driven for a proper earth ground, or else a large copper mesh mat if the ground is rocky -- unless you are certain that the copper piping that you want to use extends a significant distance underground and unbroken.
The purpose here is not to use the piping *as* a ground, but to ensure that the piping *is* at ground potential. Otherwise, if an electrical failure causes the pipe to reach a dangerous potential then so does the water in it, then so do the hands you're washing in that water. Thus if there's an electrical discontinuity in the piping it is even more important to earth bond any conductive piping/taps etc. that are on the non-earth side of that discontinuity. The same applies too to gas piping except here the principal risk is static, sparks and the subsequent explosion.
I think it is also important to note that NEC 250.52(B) prohibits gas piping as a grounding electrode(1990 or so). The gas pipe ceased as a grounding electrode due to the dielectric fitting at the meter. The gas company did not want a bond around the meter because it defeated the isolation fitting. The presence of gas is not relevant, IIRC. In the old days, it was a big no no (at least according to the hourly wage fellows who actually do the work) to hook the gas line "as" ground other than any incidental grounding which ocurs in a gas furnace as an example. Good place for resources is http://www.mikeholt.com in the forums. Decent community of knowledgeable folk there. Good luck, and "no" do not use your body/fingers/arms/etc to connect various pieces of equipment to see if a voltage exists:-) That's best left to close friends who stand near electric fences. I had problems in the mid 1990's in an older home where the galvanized water supply pipe was the primary ground. Over time, corrosion of the pipe reduced conductivity, and lightening storms toasted a few expensive items (e.g. ISDN gear, sun workstation, etc) before finally driving a few grounding bars into the soil in the basement. Cheers, andy
Randy Bush wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
i strongly recommend that you do this, especially in your 240vac environment. excellent solution to a lot of problems.
randy
I agree, dont propose this to a wood logger :) But yes, I did. I have seen an installation where "ground" was floating somewhere at 110 Volts AC. There was no way to tame it. We had to cut it. Ofcourse we did it not on the wire but in the sockets and we got a reasonable ground before we did. Dont read in the books - and dont tell a lawer :) The soil was extremly dry (not in europe) and the powerline was extremly long... Regards Peter and Karin -- Peter and Karin Dambier The Public-Root Consortium Graeffstrasse 14 D-64646 Heppenheim +49(6252)671-788 (Telekom) +49(179)108-3978 (O2 Genion) +49(6252)750-308 (VoIP: sipgate.de) mail: peter@peter-dambier.de mail: peter@echnaton.serveftp.com http://iason.site.voila.fr/ https://sourceforge.net/projects/iason/
Randy Bush wrote:
Cut the ground wire in your power cords but ground the equipment directly to a metal frame.
i strongly recommend that you do this, especially in your 240vac environment. excellent solution to a lot of problems.
Don't even joke about doing this, please. If there is potential on the grounding conductor, then that problem needs to be corrected as it is a safety of life issue. Even if you cut the conductor and safely ground the equipment in that one rack, you are ignoring the fact that you have very strong evidence of a serious wiring problem in the form of destroyed equipment. Say you do what you suggest, ensure that your rack is well and solidly grounded. And, you're aware that the building grounding wiring is defective. And then someone comes in (maybe you) and plugs in a piece of portable test equipment next to your nice grounded rack. And then puts one hand on the test equipment (plugged into one of the defective outlets) and the other on your well-grounded rack. Especially in the 240 volt environment. There is a serious, potentially fatal, wiring fault in that building. Get it fixed properly. -- Jay Hennigan - CCIE #7880 - Network Administration - jay@west.net NetLojix Communications, Inc. - http://www.netlojix.com/ WestNet: Connecting you to the planet. 805 884-6323
David Lesher wrote:
Given what I have described, would you think this is an RF interference problem?
No...
Unless you have a Gigawatt radar parked next door, I'm highly dubious that it's RF-instigated.
A) All these things say grounding issues. I have to wonder if the building is fed from more than one power entrance. The blown patch panel especially makes me think the router on one end of the Cat5 was being fed from a different power source than the one on the other. (Which pair was blown?) Given the UPS mention, maybe there's a ground differential issue with it.
B) The other, less likely, path into equipment is telco. Those mile-long pieces of copper from the CO are also called "antennas" and they covet static. I have no idea where this location is -- are there thunderstorms around?
C) One more possibility; perhaps some piece of equipment in-house is putting large spikes on the internal distribution. Twenty years ago, I read of a building where large [50 HP HVAC] and small [fridges] motors would regularly die. The high-tech gadgets of that era, Texas Instruments calculators, would reset themselves seemingly spontaneously. After MUCH work, they found the BIG copier was putting nasty spikes back on the grid.
I vote A) 75% B) 20% C) 5%
Re: C, I think (hope) the power company has already eliminated internal power spikes. The have recording monitors on both the building feed and on the distribution panels for the office area. The only thing in the building that pulls more than 20A is a couple of 4 ton A/C units and the server room UPS. Re: B, The "telco lines are antennas" concept is why we were thinking RF. The location is in the SE US, but no thunderstorms this time of the year. Re: A, I don't remember which pair was blown. Since it was a new patch panel (less than a week old) we initially wrote it off as a manufacturing defect -- but now we are not so sure. Everything in the server room -- which includes PBX and all networking equipment -- is on a single UPS. The CAT6 in the building terminates at PCs. Each PC has its own separate UPS. A floating ground or ground current was among my first thoughts -- but a few things about it bugged me and made me look elsewhere: 1) How could a bad ground cause DSL line noise that ia inaudible? Also, the noise is on the telco side, not the LAN side. 2) Why would it be blowing DSL routers that are isolated from the LAN by a switch and another router? And, all of this equipment is in the same rack, on the same ground, and on the same UPS. Well David, thanks for your thoughts. I guess the next step is to find an EE that want to tackle this challenge without asking for an open-ended purse! Jon Kibler -- Jon R. Kibler Chief Technical Officer A.S.E.T., Inc. Charleston, SC USA (843) 849-8214 ================================================== Filtered by: TRUSTEM.COM's Email Filtering Service http://www.trustem.com/ No Spam. No Viruses. Just Good Clean Email.
On 5-mrt-2006, at 23:37, Jon R. Kibler wrote:
1) How could a bad ground cause DSL line noise that ia inaudible? Also, the noise is on the telco side, not the LAN side. 2) Why would it be blowing DSL routers that are isolated from the LAN by a switch and another router? And, all of this equipment is in the same rack, on the same ground, and on the same UPS.
This makes me think of a place where they used copper lines that ran alongside rail road tracks, and each time a train came by the leased line modems would go haywire. Electrical trains put a lot of current in the ground...
Jon R. Kibler wrote:
Greetings:
[snippage]
Given what I have described, would you think this is an RF interference problem?
No. Many of the devices mentioned are not particularly RF sensitive. Those that are will recover when removed from the interference source unless you're talking about levels that are harmful to humans. A *PATCH PANEL* ??? Short of putting it inside a microwave oven, I can't think of a means of damaging it with RF, particularly from any distance. Google "Inverse square law". If you turn the switch off and the fluorescent lights stay on, then you indeed might want to look into RFI.
RF problem or not, how would you track down this problem?
I'm 99.9% sure you have a grounding problem. Verify that your power and equipment grounds have no significant potential difference. Likewise your telco ground, and the metal building itself. Is the entire building fed from a single electric meter?
We are to the point of bringing in a consulting EE, but I am not sure that most would be equipped to solve this problem; so, what should we look for in a potential consulting engineer?
NEC grounding specification compliance, some who knows the difference between a groundED and a groundING conductor and is familiar with static and lightning protection issues. -- Jay Hennigan - CCIE #7880 - Network Administration - jay@west.net NetLojix Communications, Inc. - http://www.netlojix.com/ WestNet: Connecting you to the planet. 805 884-6323
RF problem or not, how would you track down this problem?
To start with I would install some cheap equipment that is more likely to fail so that you can INCREASE your failure rate and get some more data. Maybe consumer grade DSL switch/routers or something like that. Also, talk to radio experts (ham radio) about how to measure the field strength. It is entirely possible that there is some kind of accidental waveguide that channels RF into your facility under the right conditions. Grounding could also be a problem if somebody is pumping lots of volts into the ground nearby. By the way, your timetable sounds like a factory source. Something is done on every shift change, and then maintenance does it once more during the night shift. --Michael Dillon
participants (15)
-
Andrew C Burnette -
David Lesher -
Ian Mason -
Iljitsch van Beijnum -
Jay Hennigan -
Jon R. Kibler -
Martin Hannigan -
Matthew Sullivan -
Michael.Dillon＠btradianz.com -
Peter Dambier -
Randy Bush -
Robert Boyle -
Scott Morris -
Steven M. Bellovin -
ww＠styx.org