Chuck,After reading my entry again I realize that that I was not totally clear on the transceiver bit. We would put a transceiver on the incoming (outside) copper cable the have the fiber patchcord and then another transceiver to connect to the inside copper cable. That seemed to eliminate much of the ground loop issue in most of the older buildings.
Cory "For all the ways in which technology has failed to improve the quality of life, please press three." --Alice Kuhn On 09/17/2011 02:24 PM, Chuck Stickelman wrote:
Cory, I absolutely agree with the idea of using fiber between buildings or on exterior runs. It is almost impossible to protect against ground loops between buildings. Fiber also eliminates the antennae-effect Larry mentioned earlier. Years ago I worked for St. Ann's Hospital in Westerville and they were building a new extension to the hospital. I asked the project manager about bonding the two buildings to try to eliminate ground loop currents from my data lines. He seemed confused, like he had no idea what I was talking about. Together we went to the point where the new and old buildings met and he placed a small jumper wire between the two structures. In less than 20 seconds the wire heated up and failed. The manager was flabergasted to say the least. He made multiple attempts using increasingly larger wires until we settled on a braided grounding wire, like what you'd find in a car, one that was capable of carrying dozens of amps. The problem was that these two buildings had different services entrances, one of which was about 30' higher in elevation than the other (which implies that the moisture content at the point of ground would be significantly different...) so there was a significant (1-3V) voltage differential between the "ground" points of the two buildings. This voltage was capable of delivering HUGE currents. We ran a lot of Twinax cable back then (long, sad story...) which used a braided shielding around the conductors. This shielding was often grounded on both ends, which in the scenario above meant that there was a huge current on what most people thought was "ground". Not a good thing at all. While this was an extreme example, the same sort of thing is likely to be going on in your house right now, to a much lesser extent. Lightening like what Kory described will take advantage of these paths and push large voltages and currents up through the circuitry via the ground pin. Nasty stuff those grounds loops! Not totally sure, but I believe the National Electric code allows fiber to be in the same run as high voltage lines, which would make it easier to get connectivity in some instances. (Though you'd want to be careful not to crush the fiber...) Chuck On Sat, 2011-09-17 at 11:20 -0400, C Olson wrote:Just FYI on ethernet port protection that I used at various government sites that had outside feeds between buildings and also here in Ohio with Broadband Wireless internet is to use fiber optics transceivers to be the go-between since the electric charge cannot go across fiber. Saved many devices from lightning issues. Cory "For all the ways in which technology has failed to improve the quality of life, please press three." --Alice Kuhn On 09/17/2011 10:55 AM, Chuck Stickelman wrote:I know it's expensive, but what is the value of: * the hardware you're protecting, * your time and aggravation to put things to right, * your data? What can be used other than MOVs? 1. Coils and capacitors can be used to filter out some power-line anomalies. 2. Active components, such as Silicon Control Rectifiers (SCRs) and Triacs can work in conjunction with Schmitt-triggers and resistors to dump excess energy. 3. Spark gaps and other devices can be used to as lightening arrestors. In practice I believe that high-quality protectors use a tiered approach: the first stage attempts to shunt the nasty stuff like lightening to ground before it can damage anything, the next stage then uses active components to get the rest under control, and the last stage uses caps and coils to filter the output. (There could even be an output transformer at the very end of the process, such as what is used in the Ferrups product line of UPS.) To have a complete system, you'd probably want to implement a whole-house approach with additional protection of key equipment. It is going to cost to do it right. Anything less and you might be better off doing nothing and banking your savings to buy shiny new hardware every year or so... Chuck On Fri, 2011-09-16 at 17:23 -0400, Kory Pounds wrote:Chuck, I took a look at the surge protector at your link. It is a bit expensive for my tastes. However it looks good. It covers just about everything. If it doesn't use MOV's, then what does it use? I also saw that it had ethernet line protection like we talked about. Kory On Fri, Sep 16, 2011 at 2:13 AM, Chuck Stickelman <cstickelman@xxxxxxxxxx> wrote:My point about MOVs is that they are *BAD*. Not good in any way. Avoid at all costs, even if they have the indicators. Many surge protectors have network and phone line protection. A quick Google search for "Ethernet Surge Protection" shows multiple devices. I do not know whether or not any specific model uses MOVs or not... Here's an example of a unit that might have some value... (http://www.tripplite.com/en/products/model.cfm?txtSeriesID=829&txtModelID=2813) but notice the MSRP of $112... Chuck On Thu, 2011-09-15 at 18:24 -0400, Kory Pounds wrote: I researched surge protection myself and read about MOV's. Yes, they are good but they do indeed degrade as they take hits. And there are nicer surge protectors with indicators that show the current level of protection that it is providing. However, how does this protect an ethernet line? I have not seen a surge protector with an "ethernet in-out" along with the regular outlets, phone and cable protections. In the story that I gave, the upstairs ethernet line was hit with a surge and not the actual power wiring going up to that level or anywhere else. My brother's laptop still worked, including the wireless part of it, but the NIC went out. How could this have to do with a traditional surge through the power cable? If the surge started in my brother's laptop, then the whole thing would have been damaged as the surge went out of the ethernet port and down the line into the basement to the router. If I am missing something here, then please let me know! I am not an electrician or systems/network guy. Is it actually possible for it to start int hat laptop but ONLY hurt the NIC as it traveled out? If that isn't the case, then we are not talking about traditional surges through the house's wiring. If there is not a device that is specifically designed to protect from ethernet surges ONLY, then I am willing to buy a couple of cheap Netgear switches that I am seeing at Amazon.com right now. I would rather have that blow out than my nicer dual-band Cisco wireless 4-port router or someone's ethernet port in a laptop/PC. What does everything think about this? I know I am not the only one who wants to protect our expensive equipment from being damaged like this. Kory On Thu, Sep 15, 2011 at 6:07 PM, Chuck Stickelman <cstickelman@xxxxxxxxxx> wrote:I want to expound on what Larry has said about surge protection. You can certainly find inexpensive products that will claim to be surge protectors. Avoid cheap solutions for this! (Other than Larry's recommendation for sacrificial hubs, I love that one!) There are cheap devices called Metal-Oxide Varistors (MOVs) that can absorb electrical surges. Unfortunately, every time a MOV takes a hit it is less able to provide protection, until ultimately, they die. Some manufacturers have spent extra money designing LEDs that tell you when the MOV is dead and the unit needs replaced, but most don't bother. (Not to mention that many people would fail to check the LED status or ignore the warning...) Manufacturers of high quality surge protection often brag about the fact that their products use no MOVs. Look for that. Also, you *may* be able to tell a little about the quality of the product by the size of their equipment guarantee. (The trick here would be to read the fine print... and know what exclusions exist...) In the end though, as Larry mentioned in his first response, lightning is a whole subject itself. It might not be possible to identify and protect every path that lightning can take to get into your network. Chuck On Thu, 2011-09-15 at 22:28 +0100, DiGioia, Larry wrote:The answer is, there is no affordable product that will protect your equipment. A good Ethernet surge suppressor goes for $$85 and up. I have one in my outdoor transmitter shack, on the CAT5 line that goes to the top of my 110' tower, where there are multiple wi-fi panels. It also requires a good ground, meaning an RF ground - which would be copper strap attached to a 6' ground rod less than 10' away. This is probably not practical for you. My tower recently took a direct hit - destroying the panels, and a key component of my antenna rotor. But it did not make it into the house. (I also have coaxial surge suppressors, and rotor control line surge suppressors, all mounted on a solid copper bar, which is tied through that 3" copper strap, to a network of ground rods. Altogether, I have 20+ ground rods around the house...) I would recommend that you consider using sacrificial hubs - old hubs or switches at strategic points. They tend to not transfer hits to other ports, although this is not a certainty. But for 0 - $10 or so, I would say they are worth it. And there's always wi-fi... -----Original Message----- From: ncolug-bounce@xxxxxxxxxxxxx [mailto:ncolug-bounce@xxxxxxxxxxxxx] On Behalf Of Kory Pounds Sent: Thursday, September 15, 2011 5:06 PM To: ncolug@xxxxxxxxxxxxx Subject: [ncolug] Re: How can only the ethernet ports blow out on the wireless router??? Everyone, After reading Larry's reply, how can I protect against this? Is it even possible to protect against this? I really do not want to have to bother with unplugging ethernet cords every time we think there might be another storm. What about if we do not realize a storm is about to enter the area? Is there something that can protect an ethernet line from surges? What about utilizing some cheap hub where the long ethernet line from the second floor can plug into right before the hub plugs into the router next to it in the basement, so that the cheap hub blows out instead of the nicer, more expensive wireless router? There has to be an answer to this. What do you all make of this? Thanks, Kory On Thu, Sep 15, 2011 at 7:29 AM, Larry DiGioia<larry@xxxxxxxxxxxx> wrote:Lightning is a whole subject in itself. As a ham radio operator, I am quite familiar with this, and I have also seen many episodes like your in the workplace. What is happening is that the ethernet (CAT 5) wire is acting like an antenna, picking up the lightning like a radio wave. Lightning IS in fact, RF. The results vary at different points and places because of differences in lengths of the cable (resonances) and differences in the potential with respect to ground of other connected equipment. NOT being plugged into AC power would help in this case. You may also see ethernet ports fail later due to "degradation" as opposed to outright destruction. On 09/14/2011 11:15 PM, Kory Pounds wrote:Ok everyone, This is a total weird one to me. Maybe you can help me figure out exactly why this happened. Let me explain: This happened during that violent storm we had during the night a couple of weeks ago, with plenty of lightning, etc. At some point our power went out for only about 15 seconds or so (and I heard a "snap" in the background, too). Anyhow, after the storm passed through, I found that the 4 ethernet ports on the wireless router (a single band Linksys) failed and would not work at all. However, the router still powered up and the wireless part still worked. I could still access the internet wirelessly through it. I know that it was not a power surge because the DSL modem was fine and both items are plugged into the same power backup/surge protection battery. The router and modem are in our basement. My bedroom is in the basement and that is where I have my laptop most of the time, with it plugged in by ethernet when it is there. My laptop was fine after this incident, no blow-outs or anything else. My laptop stayed plugged into the power and ethernet all during that night. I verified my laptop by plugging its line into the DSL modem and I accessed the internet. I unplugged it and I accessed it successfully by wireless as well. So I knew my laptop was fine and the problem was with the router. However, my brother has an ethernet line that runs from another of the router ports up to the second floor of the house, where he has a desk where he often works online with his laptop. That night his laptop was up there and plugged into that ethernet line. In the morning he found that his ethernet port/NIC card in his laptop was blown. He could only access the router through the laptop's wireless part. How in the world could this have happened?! Why would only the ethernet ports on the router be affected? Why would my brother's NIC be affected on the second floor and not mine in the basement? Thanks everyone! Kory "The Perl Guru" Pounds P.S. I ended up swapping in a nicer, dual-band Linksys router that I had packed away as a backup. I definitely DO NOT want some similar tragedy to happen to that one as well! Also, is there any sort of "surge protection" for ethernet lines (if it is needed at all)??? To unsubscribe send to ncolug-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field.-- Studio - D Productions l o n g w i r e . c o m "The art of being wise is the art of knowing what to overlook." William James To unsubscribe send to ncolug-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field.To unsubscribe send to ncolug-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field.To unsubscribe send to ncolug-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field.
To unsubscribe send to ncolug-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field.
