Resolved Do I really need to use Ethernet protection with my UPS

My old APC UPS battery died so I got a new APC UPS. It has a provision in the back for a Ethernet cable, my question is do I really need this. I am using a stand alone modem & router. My computer is an older 5 year old dell desktop.

I guess the reason for the added Ethernet connections is to protect my router from a power surge via the modem?

I am also concerned that there will be some signal loss, because of the Ethernet cable going from the modem through the UPS and then to the router, or would the ethernet cable run from the ethernet cable out on the router to the UPS then to my computer.

Seems there are a few different ways to connect the ethernet cable to the UPS. I guess if I do hook it up, the lowest signal loos would be my last option I mentioned.

Again do I even need to do this. My old APC UPS just had a in and out plug for for a phone line, I did use that because it was easy to get at, and I use my fax machine a lot...

 

Run an ethernet cable from your router into the UPS then another one out of the UPS and into the back of your computer.

The idea is to stop any surges from getting into your computer via an ethernet cable.

 

I've lost over the 17 years I've had internet access, two old dial-up modems, one DSL modem, one integrated network (ethernet) card, one satellite dish receiver and one 42" flat-screen TV from power surges through the phone lines caused by lightning strikes.

All our phone lines from the telco 50 miles away to the connection outside our house are underground.

I realize a cable connection does not come through the phone lines but the cable wires themselves are as vulnerable as phone lines to lightning strikes.

So, yes, plug the cable into the UPS. It will only cost you the expense of an additional cable from the UPS to the router or computer. It's the cheapest insurance you'll buy.

 

Thanks everyone, will hook up the ethernet cables tonight...

 

First, I applaud you for using an UPS and certainly, APC is a, if not the leader in the industry. Everyone should have a good UPS with AVR to protect their computer, networking and expensive home theater equipment. And as more people are moving to VoIP telephone services, keeping the network alive during power outages is becoming more of a life or death safety issue.

But note you typically don't have to replace an UPS when they fail - that normally means the batteries just need to be replaced. And that is something most users can easily do. It is much less expensive than replacing the whole UPS - especially since the batteries need to be replaced every 3 or 4 years.

Just don't buy replacements from UPS maker. They charge a bundle for the same batteries you can get for less. There are lots of ups batteries sites out there. Note I have an old 500VA APC UPS that recently failed. APC wanted $45 plus $7 shipping. I found the exact same battery on-line for $21 with free shipping and a 2-year warranty vs APC's 1-year. The other UPS makers do the same thing so can't blame just APC.

And when I say "exact same" I mean it. Peel off the APC label on your battery and you will find it is the same battery. APC is an UPS maker, not a battery maker. That said, I don't worry about the replacement battery maker, only the physical and electrical specs, and the warranty.

I have several UPS and when the batteries fail, I shop the various battery sites as prices vary day-to-day, as do shipping charges.

The connectors come in two types, F1 and F2.

Oh and if the original is a 1270 or 12V7Ah (12 volt 7 amp-hour), you can get 8Ah or even 9Ah, as long as it is still 12V, has the same physical dimensions and same connector type. It will cost a bit more but you get longer run times.

And your local Radio Shack will take the old batteries off your hands for proper recycling. Don't toss them in the trash - keep them out of landfills.

One caution - these are SLA ( "sealed" lead-acid) batteries. They are not suppose to leak. But neither are Duracell or Energizer AA batteries and until Man can create perfection 100% of the time, there will be faulty samples. If you open an UPS and the battery has leaked all over the UPS' circuitry, you may need to toss that UPS if you cannot thoroughly clean it up (a messy, hazardous and corrosive process). But I have only seen that once in almost 30 years of UPS experience.

 

Ethernet already has robust protection typically for up to 2000 volts. If you have a surge so large as to damage ethernet devices, then everything inside the building is at risk including furnace, dishwasher, dimming switches, and all clocks. Surge protection means that surge is not inside the house.

Your phone and cable wires should already have best protection where those wires enter the building. A lightning strike far down the street means a surge is incoming to all appliances IF you have not earthed all incoming AC wires - either directly to earth or via a 'whole house' protector.

Why are computer/internet appliances at so much risk? Due to protection already on cable /phone /DSL wires, then that is a best and destructive path used by a surge to connect to earth. Solution is always about connecting that energy to earth BEFORE it can enter the building.

 

This is a very old debate, isn't Westom. 2000V??? Nonsense! Please show us in the 802.3 standards where it says Ethernet devices provide ANY protection, let alone protection against 2000 volts between modem and a router.

And when providing such evidence, please note the OP is NOT talking about networking over power lines, nor is he talking about Power over Ethernet.

And while clearly you are convinced surges can only originate from the grid, please stop trying to convince everyone else of this falsehood.

Westom, a simple search of westom protection shows you wander all over the Internet spewing your falsehoods that are constantly shown to be wrong.

Just because your UPSs outputs a 200V square wave, a "good" or "proper" UPS with AVR does not. And while an approximation sinewave is not ideal during a power outage, it is fine for most computer electronics and certainly good enough to prevent a PC or server crash - something your whole house protectors you keep spamming about don't.

And contrary to your unyielding belief, a "good" UPS with AVR is not just for power outages.

I note more unsubstantiated claims: Audioholics.com (hover over the red line under the username)

Wilderssecurity

Broadband Reports

DiItYourself

http://forums.steampowered.com/forums/showthread.php?p=22380357
http://www.techpowerup.com/forums/showthread.php?t=143351&page=2

 

You are always quick to accuse. Never provide facts. Empty accusations work if a reader fails to notice your lack of technical knowledge. Made obvious by posts constantly devoid of numbers.

These are facts you should have known. Ethernet ports meet many standards including Telcordia GR-1089-Core intra-building Specifications and ITU-T (www.itu.int) Specifications.

Telcordia required each Ethernet port wire to withstand a 100 amp transients that creates up to 800 volts using an industry standard 2/10 microsecond pulse. Notice facts with numbers. Also must withstand a 2/10 microsecond and 100 amp pulse that creates up to 1500 volts on all Ethernet wires. These standards apply to intra-building equipment.

Equipment that connects externally must meet even more robust requirements. How can this be when you repeatedly post denials of internal protection routine in all appliances? Nasty denials are better than facts with numbers?

Move on to another example of what you never learned. ITU-T requires testing with another industry standard (8/20 microsecond) surge resulting in 1000 volt or 1500 volt pulses.

How do designers meet these numbers? Well if you knew this stuff, you could answer that question with numbers. Isolation voltage of an Ethernet interface transformer is rated at 1500 volts meaning its actual protection is something approaching 2000 volts. Capacitors associated with that interface are often rated at 2000 volts.

Enough numbers? I'm not done. You see I do this stuff and you do not. Viewing one of my old datasheets for a Bourns Ethernet transformer (PT61024), it says, "Isolation HiPot 1500 VAC 1 mA for 1 second ". How can this be when you knew existing protection was never anywhere this robust?

When were you going to contribute something useful? When were you going to fill us with your knowledge?

What were typical numbers for 802.3 Standards? 1500 V at 60 Hz for 60 seconds, 2250 VDC for 60 seconds, and ten 2400 volt 1.2/50 microsecond impulses applied with alternating polarity at intervals of 1 second. Why did you not know that?

Many newer Ethernet designs that include POE are rated for as much as 4000 volts. I provided numbers that were standard for Ethernet a decade ago. After ten years, you still did not learn this stuff? Why should you? Many actually believe you because you post cheapshots, denials, accusations, and not one technical fact. You did not even know that Ethernet interfaces must routinely withstand thousands of volts without damage - more than ten years ago.

 

Accuse? No. Expose falsehoods? Yes!

All you do is toss out your own made up numbers with no basis in fact. If you provide links, they never support your falsehoods - just as that ITU link does not support your 2000V figure. And for those rare cases where your figures are based in fact, they don't apply to topic of the discussion. For example, this topic is NOT about networked devices connected to patients in a hospital. But you pull MOPP figures for such devices, then pretend they apply to a $30 router in a home network, or a $5 NIC in a home PC.

You criticize my technical knowledge but cite figures for networks that have "different isolation and grounding requirements" and pretend they apply to all networks - including home networks. And you pretend they apply to "unexpected" anomalies - the topic here.

No you didn't. You just make up numbers. Where is this 4000V stated as it applies to the typical home network? And again I remind you we are talking about standard home and small office Ethernet networks. NOT medical equipment or large corporate networks.

Do not mix facts! Network isolators as found on medical equipment to prevent unwanted current flowing through patients connected to sophisticated monitoring and regulating devices is NOT, in any way, the same thing as protecting network equipment as found on home or small office Ethernet devices, which is what we are talking here!

The problem with you westom, is you believe you are right, and the rest of the world is wrong. You believe what applies in some cases must apply in all cases. You believe all appliances provide "superior" protection. All modems provide sufficient noise filtering. You believe all power anomalies come off the grid. You believe network isolation is the same as protection. You believe the output power from ALL UPS is so dirty it will harm the connected equipment and that an UPS is only for power outages. You believe that surge and spike protectors don't protect. You believe what applies to DCOs applies to all facilities - including all houses, apartments and small offices.

I provided links to just a few of your westomisms where people can see I am not the only one who sees through your smoke.

The ONLY fact I (and others) consistently agree with you, westom, is the importance of having the shortest path of least (ideally zero) resistance to Earth ground (or a common floating ground - as in the case of avionics, for example). No argument there.

 

This again demonstrates why your posts are only nasty. And devoid of any facts. You do not do this stuff. You even deny numbers provided by the ITU-T, Telcordia, datasheets from manufacturers, and even the IEEE 802.3. Why? Otherwise you would have to admit to being brainwashed by hearsay.

Worse, you do not even read what is written. I never said Ethernet is good for over 2000 volts.

To meet 1500 volt specs, Ethernet interfaces were designed for voltages approaching 2000 volts - in *excess* of 1500 volts. Numbers from 802.3 require protection above 2000 volts. Newer POE designs are something around 4000 volts. Since you are such as expert, then you already have numbers that deny that. So post them. Oh. Only you are allowed to know these numbers? Reality: subjective slurs without any spec numbers.

Debate never existed. I provide facts. You invent and post subjective and offensive denials. All readers are warned about anyone who makes claims without numbers. Driven by ego; not by technical savvy.

Instead of posting accusations from hearsay and advertising, surprise everyone by quoting actual facts. If ITU-T numbers are wrong, then quote specific ITU-T numbers. You cannot. You never designed anything. You never learn from datasheets. You do not even know why that transformer exists and why capacitors are often rated for 2000 volts. Bogus accusations devoid of numbers, reality, or experience are to confuse all others. If you have knowledge, then surprise all. Post specific specification numbers. You are wasting bandwidth while confusing others with nasty and empty denials.

Ethernet interfaces *were* so robust as to withstand voltages approaching 2000 volts. After ten years, you still do not know these numbers? Protection approaching 2000 volts so that Ethernet can meet 1500 volts standards without damage. And to meet 802.3 spec numbers that defined "2250 VDC for 60 seconds" and "2400 volt 1.2/50 microsecond impulses ". Numbers that the naive never learn while posting insults; by pretending to be an expert.

Robust protection has always existed in appliances. Others educated by fears from advertising never knew that. Even that UPS only claims near zero protection - once one actually reads UPS spec numbers.

 

I rest my case.

End of discussion.