Member Discussion

Is It Possible To Run PoE More Than 300 Feet?

I am supposed to be spec'ing out a plan for surveillance in a housing project. They want IP cameras, and the longest run can be about 400ft.

There is no local source of power, closer to the camera. I wanted to put a secure lockbox at the midway point, but they said due to security concerns, they don't want me to. As far as I know, the max distance is 300ft. Is there any trick I can use to boost the distance by another 100ft.


If there is no way of using PoE extender, I would consider PoE over coax for the runs that are over 300ft.

This from Veracity can do it.

They also have POE over coax as Jack suggests above.

400ft exceeds maximum *ethernet* distance,not just PoE.

You run the risk of collisions due to timing errors if you exceed the Ethernet distance spec.

Vigitron also has extender options.

These are smaller then Veracity units if you need to put them in boxes.

Thank you. What is the benefit of using an extender vs coax? or the other way around?

You can get PoE over coax at cheaper price for the models that provides less mbps output. Plus, PoE over coax is a transparent device, therefore you are less likely to run into risk of port security lock down by corprate routers in some situation (this is not a common issue thou).

However, you might need to pay a little bit more for the PoE long range extenders. but they seem to offer longer distance with higher mbps output and high power PoE compatiblity. So if you are using high power PTZ cameras. you might want to consider those PoE extenders.

I would proably be putting in non PTZ 3mp IR cameras. So for that the PoE over coax whould be good enough?

Then both PoE over coax and the long reach extender will work for you. You might want to check the power consumption on your Ipc plus the consumption from the extender that you r planning to buy, to make sure you have enough power budget on ur PoE switch, sometimes you might also need 802.3at instead of the standard normal 802.3af which only provides upto 15.4 watts. The easiest the way to spec out is to speak to the tech support from your local distributor.

One other note. If there is a requirement for outdoor ft6(or plenum) cable for the project. Then coax cable would be lot more expensive. In that case, you might want to weight your options.

...PoE over coax is a transparent device...

Never used POE over coax, so can't say about those, but I don't remember any Cat X injector/splitter/extender having an IP address if that's what your implying. But then again I may just have never had to access it. How do you mean they are not transparent?

By transparent device, I meant there is no additional mac addresses added to the data flow and less likely to be detected as a individual network device. (I cann't say for all the models there are out there thou)

Injector/splitters/mid-span hubs and as well as the normal extenders does not have those issues as well.

But some long range extender as well as some of the fiber media convertors may have this type of issues. One of my customer was working on a project for car rental company, where both media convertor and long range extender were on their do not supply list. however, this a very uncommon issue, I'm just throwing it out there in case this ever becomes an issue for anyone else.

Transparency is to be expected since midspans etc. are not doing any switching. But once you are going to exceed 100M, you need to store and forward the frames, and double hop it to where you are going. So they probably contain 1-port switches. And so therefore must have MAC addresses. Never considered that. Good call.

Thing is though, you would think that any long range converter would have to be a miniswitch of sorts, just because of the additional latency, whether the device was catx or coax, so a little mystery still there for me at least.

Basic L2 switches are protocol transparent on a network, they don't have any per-port mac addresses that are discoverable in the sense that end-point hosts can recognize or count the number of switch hops in a network.

Also, you really can't have a 1-port switch, what would be the point of a device with an ingress port and no egress port? :)

Switches (and hubs) are transparent from the perspective that they do not alter Ethernet datagrams in any way. If the two hosts communicating with each other are on the same LAN/subnet, the frame received by the endpoint will essentially match the frame transmitted 100%. The impact they have is potential latency, though these days the latency is generally not an issue in the way it was in the days of the 3-4-5 rule.

You can transmit Ethernet packets well in excess of 100M as measured by the cable distance between two hosts, you simply cannot do it on a single twisted-pair cable segment without either adding a repeater device (switch), or a media converter.

Also, you really can't have a 1-port switch, what would be the point of a device with an ingress port and no egress port? :)

A loopback plug is a 1-port switch, (or at least a 1-port hub), no? :)

But of course that's not what I meant when I sloppily said 1-port switches, and should have said something like, they each contain 1-port of a (shared) switch.

As for the "and therefore the switches must have MAC addresses" folly, there is really no wiggle room at all. I was, of course, just dead wrong. But I do know better than that, I just spaced and thought that if routers must have ip's then... But of course no one could accuse a router of being transparent...

Anyway now that I've fumbled the ball away, maybe you could be so kind as to hazard a guess as to what Jack is referring to above, with coax vs cat5 extenders in regards to their transparency...

I would like to point out that I'm not saying that all of the long range Ethernet extender would have this type of issue or none the PoE over coax device have it.

The issue was raised by my customer's client over the use of long range Ethernet extender (media converter). Apparently they bad experiences with some models which they have used before where additional mac address popping up on the same port resulting port security violation. Unfortunately, since this is second hand information, forgive me for not being able to provide much technical details on it.

I personally don't think that this would be a common issue. But it might become a topic during the quoting process. I've also double checked and noticed that few of the manufacture such as Alex's suggestion vigitron already stated that their current models are completely transparent to the corporate IP network and higher layer protocols.

So, I guess if you are planning to offer long range Ethernet extender or any type of media conversion device (including PoE over coax), you should always check those specs with your distributor. if corporate network security is a big factor.

Amy suggestions of a good inexpensive poe over coax, or ethernet extender?

Talk to Jack Zhan

He can help

Jack I will call you Monday :)

Thanks Alex, If you don't mind please edit out the company name.

I forgot to change my profile and take out the link to the site. I'm in no way trying promote anything:)

hah, Thx buddy

Both Rukmini's and Alex's suggestions seem nice and the specs are amazing. If you are in the US, you should give those companies a call.

Those both appear to be cat5 extenders. Can you please suggest a poe over coax device?


Rukmini's link also has PoE over Coax. under Ethernet over Coax Devices section

Vigitron has many options for POE over coax, 2-wire and UTP cables.

Axis also makes an inline POE extender that does not require a seperate power source.

I will say I've done a single 1.3mp camera over POE at 125 meters and it works fine. Been up there for a year now with no problems.

I know I'm not supposed to, but the customer wasn't giving me a lot of options. Didn't want to pay for a POE extender, didn't want to pay for us to re-run the cable (it was originally because of them we had to go the long way, so we weren't going to re-run it for free).

Was it passive POE or active POE? Just asking because an active PSE looks for a certain resistance range to determine if a PD is on the other end.

You would think that the added resistance from the extra cable might (at some point) prevent the PSE from identifying a PD correctly. Not that it couldn't actually power it, just that it might refuse to. I believe that 125 meters is one of the longer ethernet runs I've heard of, did you get a chance to run a network analyzer on it?

Is there anyone out there that can top 125M catx fast ethernet cable? (POE or not)

I did not get the chance to run any sort of analsys on it, and I no longer work for the same company I did when I installed it so I can't just run over there and do that anymore.

It was cat 6, running a single Axis P3364-VE camera.

Looking around the web I see several people claiming > 150M and a couple > 170M, wow.

For me, I remember terming cat5e? at about 120M and it didn't work, I had to keep shortening it by 5M and re-terming the end and retrying... I got down to 105M before it worked at all, and 100M before it worked well. Sometimes an experience can have undue influence on you, because after that day I was pretty damn convinced that it was something to due with propogation delay and so it really couldn't be beat by much. But turns out it's just good old signal attenuation...

"Is there anyone out there that can top 125M catx fast ethernet cable? (POE or not)"

just plugged a Geovision BX320D-1 (Poe) into 160m (Cable IQ verified) crappy Clipsal Cat 5e (not even 23awg) and it's booting and showing on the system, it IS however logging port down errors every few seconds on the PoE switch (HP 5120-48-POE)...

Pretty sure you could easily get 120m on the Cat5e rubbish and I recon using the good clipsal Cat6a I might get 160m :)

I'll hunt through some of our boxes and see if we have sub 200m but more than 150m on any of them for a test.

Actually I stand corrected, that port that camera is connected to is NOT listing errors, it's another network device, the Camera seems stable.

will leave it in for the day and see if it stays stable.

better get to checking out that other network device (wifi router)

Too bad you can't make it a little bit longer, and then a little bit longer, you can only start with too much and go the other way.... :(

For what it's worth, my camera on the 160m cable, since the initial plug-in, has not logged a single Milestone frame drop nor any additional port errors on the switch.

The problem is that you're testing it the wrong way.

Ethernet cable distance is based off packet propagation times, factored to avoid causing excessive collisions on the network.

Switches help tremendously compared to the hub'd networks of the old days. A good majority of data stays off the ports and cables of connected devices because the switch is only sending packets to the port of the machine the packet is destined for (based off the MAC address). With devices commonly being on dedicated switch ports these days its much easier to have a cable that is just a little bit long work with no noticable issues.

You might find that your over-length cable works fine for a while, until your network utilization increases and now you start getting more traffic, particularly if broadcast traffic starts increasing for any reason (since broadcast packets will generally be sent to all ports on a switch). The higher traffic load will increase the chances of late collisions on the segment with the long cable. If that segment has something like a camera, which is putting out near continuous data you can get retransmits alongside the new datagrams, and you really start increasing the chances of even more collisions and delayed or lost data.

If you are running twisted-pair cable beyond the spec, the only way to really ensure it has a high liklihood of continuing to work is to change NOTHING about your network (including typical utilization) after that segment is working acceptably.

Fair call, the HP Switch statistics are showing two "inErrors" out of the hours it's been operating (Milestone showing no errors, switch showing no logged port disconnections)

I still go with "you could get away with it"

personally I would not, instead I'd run a pair of cables over to the area, then throw in a sealed, UPS backed cabinet in, and run everything out of that.

Agree generally, but disagree specifically with

...factored to avoid causing excessive collisions on the network.

and also

The higher traffic load will increase the chances of late collisions...

since collisions are an effect of CSMA/CD. And in this case we are assuming a full-duplex 100BASE-TX link from a layer 2 network switch to a device via a single segment cat5e cable with no repeaters or media converters. I suppose one can run fast ethernet as HDX and CSMA/CD, but does anyone?

I've run PoE with good 23awg cat6 way over 100m no worries, my suggestion, if you're not sure, grab a barrel of cable, run off 150m, throw the rj's on it and let it run for a few days, assuming you have a smart NVR/VMS it'll tell you if your dropping frames.

Absolutely. Take a look at Veracity Outreach. We have worked with integrators on runs of over 1000' with no packet loss.

Hi We have test a product from Planet Technology for up to 1200feet and it works perfectly. The products are Poe-E101 and E201, for Poe and HPOE. I am pretty sure that Planet is not the actual manufacturer but they have the product available for less than $50.

Mobotix MX-2wire, they can be used on any cable. Up to 500 meters depending on cable used.

Scheme Mx2wire+

We have used the OutReach MAX XT on many installations without fail.

We have a few suspended in mid air for runs spanning poles, we have a couple buried in equipment boxes as well. If I remember correctly they even pass PoE+ and we have one in use with an Axis Q6045-E.

OUTREACH Max XT Outdoor Ethernet and POE Extender

Best of luck.

NTV Tbus is a good option as well. it is well within there spec.