dslreports logo
Ethernet Turn You On?
Power hardware via an ethernet jack
If the underneath portion of your desk looks anything like ours, we're guessing you might find Techweb's exploration of Power over Ethernet (PoE) a promising solution. While for a few years it has primarily been explored for business hardware (there is an IEEE 802.3 Study Group), it's a solution that could find it's way to your home office. Some users aren't waiting for silly standards and commercial solutions and have hacked their hardware themselves to power it via ethernet (obviously not recommended for the clueless).
view:
topics flat nest 
russotto
join:2000-10-05
West Orange, NJ

russotto

Member

Yeah, it works, but...

POE is horrendously inefficient... sending 12V DC over those skinny little wires must waste a lot of power.

Be nice to replace all those wall-warts with one providing e.g. 48VDC and use solid-state inside-the-box voltage converters from there, but it probably won't happen anytime soon.
67845017 (banned)
join:2000-12-17
Naperville, IL

67845017 (banned)

Member

Additional Interesting Reading

»www.poweroverethernet.com/

AkumalDave
Life's A Beach
Premium Member
join:2001-04-20
Minneapolis, MN

1 recommendation

AkumalDave to russotto

Premium Member

to russotto

For the right application, it's o.k.

We've been powering our office phones via PoE for some time now. It sure beats having two cords connecting to each phone (one wall wart and one data cable). Plus, PoE allows us to centralize our UPS / power protection for all telephones.

Dave

DaSneaky1D
what's up
MVM
join:2001-03-29
The Lou

DaSneaky1D to russotto

MVM

to russotto

Re: Yeah, it works, but...

How is it inefficient? Do you mean in general or just in specific circumstances? If you mean in general, then I have to take issue with you.

I use it for the specific use of powering my AP. Since wireless performace can be greatly affected by the location of the AP, I have my located in a place (high on a wall in my closet) that covers my entire property. It would have been inefficient for me to run a power outlet through the wall and 7 feet off of the floor just to do this.

AthlGrond
Premium Member
join:2002-04-25
Aurora, CO

AthlGrond

Premium Member

Re: Yeah, it works, but...

I think he is saying that ethernet wires suck up more power than if you just used a wall outlet.

Not sure how big a deal that line loss would be. (Probably negligible in your case.)

Greg_Z
Premium Member
join:2001-08-08
Springfield, IL

1 edit

Greg_Z

Premium Member

Re: Yeah, it works, but...

Believe it or not, you can send 115vac over 24awg wire. The key is very little Amperage. Since POE is using milliamps, and not Amps, you can send over a 22-24awg wire.

fireflier
Coffee. . .Need Coffee
Premium Member
join:2001-05-25
Limbo

1 recommendation

fireflier

Premium Member

Re: Yeah, it works, but...

The maximum voltage across the conductors is limited by the dielectric value of the insulation between them. Raise the voltage too high and you punch a hole in the insulation and arc to the neighboring conductors. Standard AC house wiring is often rated 600V. I've seen insulation damage in this stuff from lightning hits to the house before.

Copper 22AWG wire has an ampacity of 2.1--that's probably assuming bare copper. It will be less for insulated.

Copper 24AWG has an ampacity of 1.3--again probably bare.

If you assume 1A and run it at 48V, you've got 48W on the device end (neglecting the losses of conversion from 48V back to something lower). The POE 'standard' lowers this wattage even more. Lastly (and I'm not positive about this part) I believe the POE white paper 'standard' utilizes TWO pairs for power, not just one.

Also, that's 48V DC. Wouldn't want to run AC parallel to data lines like that.

AthlGrond
Premium Member
join:2002-04-25
Aurora, CO

AthlGrond to Greg_Z

Premium Member

to Greg_Z
I was just explaining what the original poster was talking about, I'm not disbelieving that this technology is practical (or in existence )).

Actually I think it sounds like a really good technology, and can't wait to see more products supporting it.
russotto
join:2000-10-05
West Orange, NJ

russotto to DaSneaky1D

Member

to DaSneaky1D
Most "power over ethernet" applications nowadays simply run the voltage required by the DC IN of the powered device over an unused Ethernet pair. That voltage can be as low as 5V and is typically 9V or 12V. For devices using equal power, the lower the voltage, the more the amperage. The greater the amperage, the greater the power lost to resistance in those skinny little Cat-5 wires. It's not really a big problem with APs, but I find it unaesthetic .

It appears the Power-Over-Ethernet (802.3af) STANDARD does what I've suggested and uses 48V (the obvious choice as it's used in telephony). But that doesn't seem to be widely implemented yet.

Augustus III
If Only Rome Could See Us Now....
join:2001-01-25
Gainesville, GA

Augustus III to russotto

Member

to russotto
White it may not power a 21" CRT it sounds pretty good for all those pesky devices such as hubs..

Greg_Z
Premium Member
join:2001-08-08
Springfield, IL

Greg_Z to russotto

Premium Member

to russotto
It is already happening russotto.

Steve
I know your IP address

join:2001-03-10
Tustin, CA

Steve to russotto

to russotto
said by russotto:

POE is horrendously inefficient... sending 12V DC over those skinny little wires must waste a lot of power.
Huh? PoE uses 48 volts.

cat5man
@comcast.net

cat5man

Anon

If you can run one cable...

If you can run one cable, you can run two.

Reading some of the reasons people want to run power through an ethernet cable, I've come to the conclusion that most can be done safer, with less hassle, and without violating the ethernet spec by:
1) an extension cord (that's what they're for!), or
2) a low-voltage cable rated for the power requirements, bundled along with the ethernet cable.

I've done both with success, and for those people who claim an extension cord is a hassle (or just not allowed for code reasons), I recommend the low-voltage cable run. It can be tied or taped to the cat-5 being pulled, and it's much more robust. One can offer more power safely, and the power cable can be powered from a different room or area. There's just more options this way, and it feels much simpler than the hackery involved in getting power through a cat-5 without the possibility of damaging sensitive and expensive equipment.

DaSneaky1D
what's up
MVM
join:2001-03-29
The Lou

DaSneaky1D

MVM

Re: If you can run one cable...

PoE doesn't voilate any Ethernet spec because data throughput is not affected at all.

This goes beyond simply "running wires". This is also about convience. If you have a PoE injector and outlet, all you need is power and Cat 5 cabling. Once it works, it works. There is no damage to equipment once it is done correctly the first time.

This set up is extremely portable and reliable. If you have your doubts, that's fine, but don't make assumptions about what is acceptable without backing them with facts.

If PoE wasn't seen as a niche solution, it wouldn't have been adopted by WISPs and VoIP providers, Cisco equipment, ect.

cat5man
@comcast.net

cat5man

Anon

Re: If you can run one cable...

Calm down.

I was talking about those who hacked their existing cat-5 cables (which *is* a violation of spec), instead of waiting for an actual Power over Ethernet standard, with the appropriately rated wiring, or running the power a safer way.

DaSneaky1D
what's up
MVM
join:2001-03-29
The Lou

DaSneaky1D

MVM

Re: If you can run one cable...

There is nothing "calm down" over.

You originial post didn't insinuate anything other than taking aim at PoE, with only your opinion to stand on. I simply provided you real examples of how PoE is used.

Though the standard of PoE has only recently been signed off on, there is no appropriate wiring for it to be used over, other than Cat 5. That is the benefit of PoE.

BTW, Cat 5 is a rating for transmission of speed up to 100mbps. If you leave the data wires alone, it is not a violation of data spec. »en.wikipedia.org/wiki/Cat5

cat5man
@comcast.net

cat5man

Anon

Re: If you can run one cable...

You're way too hopped up over this. Apparently I've ruffled a few feathers.
You originial post didn't insinuate anything other than taking aim at PoE, with only your opinion to stand on.
I believed it to be very clear. I even followed up for you to clear up any discrepancy. And it appears you only have your opinion as well. Mexican stand-off?
I simply provided you real examples of how PoE is used.
Actually, you didn't. But that's okay, I read the articles, and the examples for real-world situations were on non-PoE equipment. That's a big deal since even if you have an injector, you still need to splice off on the equipment end to the appropriate power connector, and then properly ground the unused wires in the cat-5 cable to the equipment and/or make sure you don't run current through the ethernet ports. I never said it couldn't be done -- just that it's more complicated and more prone to failure. Using your analogy, anybody can travel through a wormhole once it's setup correctly (it's the wormhole part that's complicated, in case you missed the joke).
Though the standard of PoE has only recently been signed off on, there is no appropriate wiring for it to be used over, other than Cat 5.
I was not aware it was ratified yet. But, as the first article says, there isn't much equipment available that is ready for PoE. Thus, people result to the hackery described in the second article. I call it "hackery" because these are Rube Goldberg style setups. Please go back to my first post on the topic where I explain the inappropriateness of most of it to solve their simple problems.

As for a real-world example where PoE just does not work due to safety, read my existing reply.
BTW, Cat 5 is a rating for transmission of speed up to 100mbps. If you leave the data wires alone, it is not a violation of data spec.
Not for electricity transmission! I don't think this can be stressed enough. I don't know what the actual PoE spec says regarding amperage ratings, but it can't be much if it still uses cat-5. The gauge is just too high for transmission without incurring a huge power loss over any substantial distance, not to mention the heat generated with larger amperages. To willy-nilly lay this out without a basic understanding of electrical theory is downright irresponsible. If one doesn't know the power requirements and limitations, there's a huge potential shock or fire hazard. Yes, I know that ethernet involves electricity, but this is minimal when compared to providing actual power to a device. USB and Firewire (IEEE 1394) have provisions for this built-in and accounted for, so safety is not an issue. Cat-5 and ethernet was not designed for this usage -- sure, it can be retrofitted, but then the complication increases, the chance for an expensive screw-up increases, and without proper precautions applied (like fuses and short detection) the safety decreases.

I just don't condone the hacks these people in the second article are undertaking. It's quite obvious from some of the commentary that they don't know what they're doing! Some have admitted to destroying equipment. Others have taken the wrong measurements but gotten lucky. This is definitely not for the average BBR crowd. It's just plain irresponsible to post that hacking link here without a bunch of disclaimers.

And if you still don't believe me about safety, go research how much amperage it takes to stop a heart. It's definitely less than those AP power supplies put out.

fireflier
Coffee. . .Need Coffee
Premium Member
join:2001-05-25
Limbo

fireflier

Premium Member

Re: If you can run one cable...



And if you still don't believe me about safety, go research how much amperage it takes to stop a heart. It's definitely less than those AP power supplies put out.



It takes approximately 30 ma AT the heart to stop it (some people claim more--I'm being conservative). This is assuming very favorable conditions.

At the 48V used with POE equipment, human skin is too highly resistive (I'll assume you're familiar with I=E/R) to get 30 ma passing through the heart. Short of cutting your chest open and jamming the open end CAT-5 into your atrium, POE is not lethal.



I was not aware it was ratified yet. But, as the first article says, there isn't much equipment available that is ready for PoE. Thus, people result to the hackery described in the second article. I call it "hackery" because these are Rube Goldberg style setups. Please go back to my first post on the topic where I explain the inappropriateness of most of it to solve their simple problems.



Not sure why the predisposition to assume hackery. The only people who are resorting to "hackery" are the people who don't seem to know POE end solution equipment exists.

»www.newegg.com/app/ViewP ··· 8&depa=0

I'm using them now and they work fine. There's no "hackery" necessary, no Rube Goldberg style setups, nothing. Plug them in, plug the end equipment in. They come in pairs--the injector and the splitter. Nobody has to "ground anything", worry about overcurrent on the conductors, and least of all killing themselves with them. They're $30 for a simple solution and they work for equipment even if the equipment they're used with isn't POE ready (just make sure the voltages are compatible). If anyone's resorting to hackery to do the same thing, it's because they want to, not because they have to--I didn't have to look very long or hard to find these.

cat5man
@comcast.net

cat5man

Anon

Re: If you can run one cable...

t the 48V used with POE equipment, human skin is too highly resistive (I'll assume you're familiar with I=E/R) to get 30 ma passing through the heart.
I have to disagree. Ever had 12v go through your hand and out your neck? Yes I'm quite sure it was 12v. Even used a multimeter to verify (well, after I was shocked -- I didn't know someone had energized the ceiling grid with a misplaced wire). I was lucky.
Not sure why the predisposition to assume hackery. The only people who are resorting to "hackery" are the people who don't seem to know POE end solution equipment exists.
Because I'm talking about the clueless people in the second article, not the people who actually use the real PoE equipment.
Nobody has to "ground anything", worry about overcurrent on the conductors, and least of all killing themselves with them.
Seriously. RTFA. The guys in the second article are dangerous.

fireflier
Coffee. . .Need Coffee
Premium Member
join:2001-05-25
Limbo

fireflier

Premium Member

Re: If you can run one cable...

Just because it went out your neck doesn't mean sufficient current went through your heart to stop it. If 48V were deadly, telephone linemen would be dropping like flies. If you think 12V is deadly, ask your mechanic how many people he knows who have died from a car battery.

Sorry, but you're being paranoid. Even dog shocker collars run at a higher voltage and dogs don't drop dead from it--nor would you.

As far as being lucky with the ceiling grid, the only luck involved was that it was 12V and not 120V. The worst you're going to get from 48V is a burn if it's from a high current source (several amps) and you get yourself or something else across it. If you die from 48V, then perhaps Darwinism has done its job.

Disagree all you want, it's still unfounded concern.
B777300
join:2002-01-02

B777300 to cat5man

Member

to cat5man
said by cat5man:
t the 48V used with POE equipment, human skin is too highly resistive (I'll assume you're familiar with I=E/R) to get 30 ma passing through the heart.
I have to disagree. Ever had 12v go through your hand and out your neck? Yes I'm quite sure it was 12v. Even used a multimeter to verify (well, after I was shocked -- I didn't know someone had energized the ceiling grid with a misplaced wire). I was lucky.
Not sure why the predisposition to assume hackery. The only people who are resorting to "hackery" are the people who don't seem to know POE end solution equipment exists.
Because I'm talking about the clueless people in the second article, not the people who actually use the real PoE equipment.
Nobody has to "ground anything", worry about overcurrent on the conductors, and least of all killing themselves with them.
Seriously. RTFA. The guys in the second article are dangerous.
I touched 220-240V over 4-5 times, its not that deadly as you think, I also touched 10,000v but for very short prieod.

fireflier
Coffee. . .Need Coffee
Premium Member
join:2001-05-25
Limbo

fireflier to cat5man

Premium Member

to cat5man
I'm preparing to use POE for some remote-site network security cameras. The purchase of the paired injector/splitter more than saves on the cost of having to resort to "hackery" (as you put it) involved in extending the low voltage cable from the AC adapter normally used to power the cameras. Extension cords through a tiled ceiling are not code-compliant here and even if they were, I wouldn't do it because it's just damn unprofessional, and finding a means of extending a low voltage cable with a barrel plug end (without resorting to cutting, splicing, and soldering or crimping and finally reconnecting the barrel plug), would likely resort in a shoddy use of electrical tape.

802.3af is a nice standard, and for low power consumption equipment it works quite well. Imagine taking a nice compact PowerDSine 3006 hub with one AC plug to power six devices vs. having to use a klunky outlet strip with six heavy AC adapters running out of it with the wires running up to the CAT-5 patch panel and onward. Thanks, but I'll take the PowerDSine hub. Besides that, you can control POE on some equipment via software and with the press of a key turn something off and back on to reset it. Many IT people would prefer that to trudging out into the cold at 3:00 am to go pull an AC adapter loose simply to reconnect it.

And what's all this about safety? 802.3af POE compliant equipment will not apply 48V to anything other than what it detects is POE so you won't be blowing anything up. 48V is also certainly not going to kill anyone even if you did gnaw into the cable for some reason. No modification to existing CAT-5/5e/6 cable is necessary at all.

POE is a good useful transparent standard for specific needs. With VOIP coming in, you will likely see more of this.

cat5man
@comcast.net

cat5man

Anon

Re: If you can run one cable...

I've already written a reply that addresses your concern.
cat5man

cat5man to fireflier

Anon

to fireflier
Oh hey, one more thing:
Imagine taking a nice compact PowerDSine 3006 hub with one AC plug to power six devices vs. having to use a klunky outlet strip with six heavy AC adapters running out of it with the wires running up to the CAT-5 patch panel and onward.
This can be accomplished the same way using separate low-voltage wiring to a hub. I use rack-mounted switch boxes tied to a DC power source (battery + inverter/charger plugged into standard AC), and those cables then get distributed alongside the ethernet. One switch for each cable, one power source, and no trudging out into the snow and wind to power cycle the security cameras. (In fact, that's what these power. But I need well over an amp at peak usage with each camera, so using cat-5 would add too much resistance, and, well, is just unsafe with that kind of current.) You could probably use the PowerDSine with my implementation no problem. And if you're having trouble soldering and splicing and using electrical tape, then you're using the wrong connectors, my friend.

fireflier
Coffee. . .Need Coffee
Premium Member
join:2001-05-25
Limbo

fireflier

Premium Member

Re: If you can run one cable...

Based on your description, I'm not sure how you're indicating you can remotely (i.e. with software over a LAN) cycle POE-like devices. I'm not sure if by "switch" you mean a network switch (as in hub, router, etc) or if you're talking about an SPDT switch you'd get at radio shack. If it's the latter, then you're still relegated to being in the vicinity of the switch, e.g. if it's at work and you're home at 3:00, you're screwed.

What you seem to be describing doesn't sound like something PowerDSine would work with, but then I'm not exactly sure what you're describing.

As for trouble soldering, splicing and using electrical tape, the lack of necessity for any of the three in a true POE application was my point--you shouldn't need any, and if you do (which you seemed to be indicating in your first post "2) a low-voltage cable rated for the power requirements, bundled along with the ethernet cable. " then there are better ways (see next paragraph).

Most cams, wi-fi ap's, etc. use barrel-style/coaxial plugs and jacks--like this: »www.radioshack.com/produ ··· 274-1568

If your Panasonic camera comes with an AC adapter with this style connector and the adapter cord is 10 feet but your camera is 30 feet away, you've obviously got to do some cutting and splicing. Regardless of whether you solder or crimp the extra 20 feet in, it's not as clean as the POE "standard". If on the other hand, your camera is POE compatible (and the coaxial/barrel plugs match), no soldering, crimping, splicing, or cutting is necessary.

If I misunderstood what you were doing, sorry but your description just sounded weird--I wasn't sure how an inverter/charger/battery was significant to POE.

cat5man
@comcast.net

cat5man

Anon

Re: If you can run one cable...

You seem to be going back and forth between PoE and non-PoE systems when you take me to task for my descriptions. I believe your setup is not PoE. Unless someone has come out with a PoE security camera very recently? If they haven't, then quit saying your installation is better, since it still requires the hack of splitting off the power from the camera end of the ethernet to the power jack on the camera itself. No? You haven't detailed your cameras, why should I?

It may be as simple as two ships passing in the night, but let me crush this misunderstanding right here.

Let's deconstruct your message:
Based on your description, I'm not sure how you're indicating you can remotely (i.e. with software over a LAN) cycle POE-like devices.
First, these aren't PoE anything devices. I was very clear on that. There isn't any power travelling over the cat-5. Secondly, these cameras can power cycle themselves via a command over the network, or, at a rackmount switchBOX (not a network "switch", I clearly said switchBOX, and since I was talking about electricity, I felt this would be doubly clear) where the security guard can manually turn off the power or turn it back on. Sorry, these aren't rinky-dink cameras part of some cheap Radio Shack system.
If it's the latter, then you're still relegated to being in the vicinity of the switch, e.g. if it's at work and you're home at 3:00, you're screwed.
No, the customer is not screwed, because the network is not usually connected to the Internet, and the company (or other entity) can afford a guard or two.
What you seem to be describing doesn't sound like something PowerDSine would work with, but then I'm not exactly sure what you're describing.
Then you haven't read my detailed description of how it works. I won't write it again. And again, you do not detail your system either, yet you tout it as somewhat better... more on that later.
As for trouble soldering, splicing and using electrical tape, the lack of necessity for any of the three in a true POE application was my point--you shouldn't need any, and if you do (which you seemed to be indicating in your first post "2) a low-voltage cable rated for the power requirements, bundled along with the ethernet cable. " then there are better ways (see next paragraph).
Actually, I didn't need to do anything too complicated. I do prepare the wire ends of the low voltage cables, but that doesn't take long. It ends up cheaper than paying for a bunch of extra equipment, safer, and more power-flexible since I can run more power through it than with a cat-5 already used for data.
Most cams, wi-fi ap's, etc. use barrel-style/coaxial plugs and jacks--like this: (RadioShack link)
This style camera is different. But I don't get my parts from Radio Shack -- they never have what I need and if they do, they're too expensive.
If your Panasonic camera comes with an AC adapter with this style connector and the adapter cord is 10 feet but your camera is 30 feet away, you've obviously got to do some cutting and splicing.
I'm talking about 100's of feet. And not Panasonic. And not an AC adapter either, although at the risk of confusion, you could use one.
Regardless of whether you solder or crimp the extra 20 feet in, it's not as clean as the POE "standard". If on the other hand, your camera is POE compatible (and the coaxial/barrel plugs match), no soldering, crimping, splicing, or cutting is necessary.
Oh, I think I see where you're coming from. You've got a splitter for the camera-end where it has a Radio Shack style multi-plug. Is that the case? Do you have a link? However, I wouldn't be able to use it anyway since I need to use separate low-voltage 1 pair cable for power reasons.
If I misunderstood what you were doing, sorry but your description just sounded weird--I wasn't sure how an inverter/charger/battery was significant to POE.
It's just the power source. All the low voltage cables terminate at the switchbox where the power for the systems come in. It makes power distribution and troubleshooting very easy. And because it's separated from the ethernet, it doesn't require access to the data room.

Your customers' requirements are obviously much different than my clients'.

And a little more about cable preparation. I usually need to prepare a length of cat-5 for each camera by crimping on the ethernet connectors (since pre-made cables of the length I need are usually too expensive or non-existent). So putting connectors on the ends of a low-voltage cable are just par for the course.

And your cameras likely require coax for the video feed, right? Or do they transmit on a TV channel? The ones I use transmit video via ethernet needing no coax. What I don't understand for you is that if you're pulling coax for your job, and you're putting in cat-5, what difference does a dedicated power cable make? For me, it's just one extra cable with the thickness of cat-5 -- just two cables for one camera.

I wish you luck with yours, and I'm glad your power requirements don't exceed what one can transmit safely via ethernet. As I've pointed out a few times before, I don't have that luxury.

Aliens
Premium Member
join:2002-10-09
space-time

Aliens

Premium Member

Might extend the life of copper vs. fiber

PoE is one of the few things that might keep copper ethernet wires around another decade for short-haul work vs. optical fiber, as the cost of fiber patch cables and laser LEDs continues to drop.

silverton
join:2000-08-10
Palo Alto, CA

silverton

Member

Re: Might extend the life of copper vs. fiber

Or practially forever, for that matter. What's the advantage of EVER running fiber for the last ten to hundred feet? Trust me, nobody has been more enamored by the promises of fiber optics than I have, but until 10Gbps to the desktop isn't enough, copper will probably own the PC connection. That said, the all fiber optic Ethernet First Mile »www.ieee802.org/3/efm/ can't come soon enough.

We've primarily found PoE of great convenience in setting up AP's around cubicle farms. It's an affordable and simple way to not blow up already overloaded cubicle power strips.

insomx
Premium Member
join:2003-01-26
Canada

insomx

Premium Member

Re: Might extend the life of copper vs. fiber

I use PoE for my WAP. I have it in my attic, so 50 feet of extension cord is not worth it. I compenstaed for the resistance of the wire by haveing a transformer that has a higher output voltage.

Steve
I know your IP address

join:2001-03-10
Tustin, CA

Steve to Aliens

to Aliens
I dunno, "Power over Fiber" couldn't be far behind...

*snicker*

insomx
Premium Member
join:2003-01-26
Canada

insomx

Premium Member

Re: Might extend the life of copper vs. fiber

NO way that will ever happen. You cannot transfer electricity in light form. Maybe at really high intensity's but the loss would be so great or the fiber would melt.

Steve
I know your IP address

join:2001-03-10
Tustin, CA

Steve

Re: Might extend the life of copper vs. fiber

said by insomx:

You cannot transfer electricity in light form.
"Sarcasm detectors" not working there, John?

"snicker" should have been your first clue

insomx
Premium Member
join:2003-01-26
Canada

insomx

Premium Member

Re: Might extend the life of copper vs. fiber

Hehe, I sorta realize that now. That was a long day that day.

Maybe they could transfer electricity of high frequency sound waves, eh?

whizkid3
MVM
join:2002-02-21
Queens, NY

whizkid3 to Aliens

MVM

to Aliens
"I dunno, "Power over Fiber" couldn't be far behind..."

Actually, you can supply power via wireless! Thats right - you don't even need cables to supply power!

It hasn't really caught on, because it requires a high-powered beam of microwave energy to the receiving device. Just don't stand in the middle, or put anything that might not enjoy some good-ole microwave heating behind it.

insomx
Premium Member
join:2003-01-26
Canada

insomx

Premium Member

Re: Might extend the life of copper vs. fiber

Well, it might be possible, but it is not safe.

whizkid3
MVM
join:2002-02-21
Queens, NY

whizkid3

MVM

Edgar Allen PoE

My two cents:

- Regarding running low voltage cables, or extension cords - this is a nice solution - for the home. (If the extension cords don't burn it down.)

- PoE is not a 'niche' market - its a phenomenal solution for large networks. In one wireless network I designed there were over 300 APs. The APs were redundant meaning that there were two, on separate networks, at each location. The cost of just running the two different electrical outlets to each location, one on UPS and one on emergency power, with all the frills, was over $300,000.00 USD! The high cost was partly due to that 50% of the APs were ran on fiber and some were over a thousand feet from the nearest electrical power. Bear in mind that this was done in a foreign country with local non-union labor, and can be considered cheap. If it were done in the US with union labor, the cost for the electric alone could have easily exceeded one-million USD. I clearly see an advantage in PoE. This cost of the electrical power would have been much, much less. Not to mention the time savings. Unfortunately, PoE was in its infancy then and switches & APs that took advantage of PoE were not available.

- In building large networks, very often the costs of the electrical installations to support them greatly exceed the cost of the network cabling. This is true especially for commercial facilities that require armored AC power cabling, or conduit runs.

- There are many types of fiber cables that provide copper wires to supply power. PoE plans for the future need to specify a way to take advantage of this.

- One of the biggest factors in supplying electrical power is the length of the run. The lower the voltage, the higher the current must be to supply the same amount of power. Unfortunately, power losses increase exponentially with higher currents & lower voltages. (This is why the power companies send out 13MV over their distribution systems.) That is why PoE specifies 48 VdC. It is virtually the highest voltage that can be supplied while still be considered low voltage. Anything over 50 Volts, must meet must more stringent specs in the National Electrical Code, including that they couldn't be run together with low voltage cables such as Ethernet. A lower voltage would increase the power loss significantly, making the PoE system very inefficient, and require larger gauge cable than standard CAT5.

The power loss (in Watts) is: P = I^2 * R,
where 'I' is the current in Amps, and 'R' is the resistance of the wire in Ohms.

jig
join:2001-01-05
Hacienda Heights, CA

1 edit

jig

Member

some comments

1)why would we want ethernet CABLES at all? isn't all local data going to be transmitted wirelessly in the future? well, maybe not, but certainly wireless data will beat out fiber for cameras and other hard to reach places.

2)instead of using a different type of cable to transmit power (this is for cat5man), you could just run another strand of cat5, twist 4 of the wires together into two sets to end up with enough copper cross-section to have as little or less resistive loss than your low voltage cable run. just a suggestion.

3)poe won't work over distances of 1000's of feet. you'll still have to run power either somewhere along the way or at the end point.

4)current from a 9v battery (the little toy kind) is enough to stop your heart if you get past the resistance of the skin in some way. specifically, stick a needle into each of your thumbs (not directly into your heart) and touch each one to an opposite terminal on the battery. i take no responsibility if you are stupid enough to try this without EMTs nearby and as a result die. 30mV is plenty to kill, but in practice would have a very hard time trying to get past the skin.
jaymay75
join:2005-01-11
Pittsburgh, PA

jaymay75

Member

Re: some comments

Thank God I found this post, I need some help.

I need to wire some Web Cameras either using POE or extending the wire on the power adapter. I'd just run the power line along with the CAT 5 when I pulled it.

The cameras probably will be 10-40 feet away from the adapter. If I use the power adapter and extend the wire 40 ft., will there be enough amps to power the camera at the end of te 40ft run. Will it hurt the camera in anyway. Where are the best places to buy the wire and connectors to do this.

Could I do this with PPOE. The camera isn't PPOE compatible, but I was thinking of running the power through the CAT 5 cable. Then at the end of the CAT 5 cable, I will have to tap into the pair carrying the current with a barrel connector to power the camera. Then I'd have to put a send the other pairs to the RJ-45 jack.

I was looking at the Dlink DWL-100 device for POE. This devices outputs 5VDC, 2.5A. The camera I'm using is an Axis 205. This camera only needs 5.0-5.5 VDC / min 400mA. Will the D-link device work to power this camera, or does the D-link device output too much power.

Any and all help would be really appreciated.

jig
join:2001-01-05
Hacienda Heights, CA

jig

Member

Re: some comments

it is my belief that you would probably want a POE camera to go with the POE dlink device. but, rather than do that, i'd just buy a regular camera (like your axis 205) and normal network gear (minimum 100Mb system) and run things completely normally cabled.

first, to run the camera you can use ANY dc output that supplies 5v as long as it supplies the minimum current (400mv). it doesn't matter if the ac-dc converter is rated at some higher current output (like the 2.5A you mentioned); as long as the voltage is the same (or pretty close), then the camera will only draw the current it needs.

lets assume you stick with the axis 205. what i would do is run two cat5 cables, one for data and the other for power. the one for data would have the regular rj45 ends in place and just plug in. the power cable would have half the conductors twined and soldered to form a positive lead, the other half the negative lead, then spliced into the regular power cable+adapter that you got with the camera. you can use regular lamp cord instead if you want (cheap at home depot). or, if there is a power plug somewhere closer to the camera you can just plug it in there.. either way, you should be able to power the camera over these types of cables over 40ft without having to worry about voltage drop across the cable length.

you could also try wireless cameras..