Ethernet POE concern

Joined
Mar 21, 2004
Messages
29,615
Location
Near the beach in Delaware
As I have mentioned before I support the WIFI at a church camp where there are many Unifi access points on telephone poles throughout the cam grounds. I had one section not working and I found a RJ45 plug going into a POE injector that was a little burned. Not melted to a crisp discolored and some of the plastic next to the gold connectors fell away when I touched it.

Replaced the RJ45 plugs on each end of the 200' outdoor cable along with the POE injector. The POE injector tested good but I felt it's jack was suspect.

One the other end of the 200' cable was a Unifi outdoor switch (POE powered) that worked fine once the repairs were done.

The last thing I cannot account for is that the 200' cable (which I think is CAT6 but not positive) passes at 10 Mbs but fails at 100 Mbs and above as tested on a Fluke LinkIQ tester.

The switch and connected access points have been running at 1Gbs since yesterday over the repaired cable.

I would have expected a lightning surge to have fried the switch and POE injector.
 
The last thing I cannot account for is that the 200' cable (which I think is CAT6 but not positive) passes at 10 Mbs but fails at 100 Mbs and above as tested on a Fluke LinkIQ tester.
The switch and connected access points have been running at 1Gbs since yesterday over the repaired cable.
🕵️‍♂️ That's interesting. Have you tried to determine if packet loss occurs over the repaired cable? It's one way to rule out a faulty cable.
 
Last edited by a moderator:
Cable crimped-on RJ45 connector causing a high resistance and getting warm maybe? I really hate crimped-on RJ45's, they can be so unreliable.
 
Cable crimped-on RJ45 connector causing a high resistance and getting warm maybe? I really hate crimped-on RJ45's, they can be so unreliable.

Probably better to put an RJ45 jack on the end of the cable and run a patch cord to the device. If it's outdoors, enclose the RJ45 jack in a weatherproof electrical box.
 
Probably better to put an RJ45 jack on the end of the cable and run a patch cord to the device. If it's outdoors, enclose the RJ45 jack in a weatherproof electrical box.
Why do people like RJ45 jacks better than a crimped plug? I use high quality RJ45 connectors and a high quality crimper.
 
Why do people like RJ45 jacks better than a crimped plug? I use high quality RJ45 connectors and a high quality crimper.

Because they're far more reliable. I've put 1000's of RJ45's on the end of data cables but would still opt to install a jack on either end and use factory made patch cables.

We recently installed a pair of CHP boilers in work (think a 4 cylinder car engine powered generator that runs on natural gas and then a cooling jacket around the exhaust manifold that is used to provide heat to the building for hot water and heating purposes) and we had employed a data specialist to install cables from the central 'brain' to each piece of CHP equipment and terminate every end using RJ45 crimps. I had to go back up to the site after 6 months and replace every cable with new cables with factory terminated RJ45's because the crimped ends were just vibrating themselves to bits all the time. Maybe if the RJ45 ends were potted with some kind of epoxy compound they might have stood a chance, but they certainly didn't last.
 
I need to bring the end of the cable inside the building and then quickly plug it into a Unifi surge protector. Then into a POE injector or a switch with a POE port.
 
Why do people like RJ45 jacks better than a crimped plug? I use high quality RJ45 connectors and a high quality crimper.

I've never liked using RJ45 plugs since at least the time I saw a bag of RJ45 plugs that were labelled as being for solid wire but were most definitely not.
 
we had employed a data specialist to install cables from the central 'brain' to each piece of CHP equipment and terminate every end using RJ45 crimps. I had to go back up to the site after 6 months and replace every cable with new cables with factory terminated RJ45's because the crimped ends were just vibrating themselves to bits all the time. Maybe if the RJ45 ends were potted with some kind of epoxy compound they might have stood a chance, but they certainly didn't last.

Solid wire doesn't like vibration. Using a stranded patch cord to make the connection to a piece of equipment with a lot of vibration is the right way to do it.
 
The Ubiquiti "passive" system only sends power on two of the four cable pairs, so it is not suited for high power as the high-power IEEE standards like bt which use all four pairs-- thus less than half the current on each connector pin. (Less than half because less voltage loss in the cable further increases power transfer efficiency.)

Waterlogged twisted pair cable will pass all DC tests but have improper RF impedance and high RF attenuation, and fail to link. Gel-filled cable should be specified for permanent outdoor installations.
 
  • Like
Reactions: Y_K
The switch in an outdoor enclosure looks to be drawing 11W of POE for itself, a downstream switch and two APs.
That doesnt sound right. usually ap's are around 10w each and switches depends on switch.
What gen equipment is this?
11w for 2 switches and 2 ap's seems REALLY low.
 
The Unifi access points in question seem to be drawing 3.5W. If you look up the spec it says a higher draw. But I think that is the maximum. I am going by what the Unifi switch tells be for power draw. Not measuring it with a meter.

The switches are small POE powered switches 5 port.
 
Because they're far more reliable. I've put 1000's of RJ45's on the end of data cables but would still opt to install a jack on either end and use factory made patch cables.

We recently installed a pair of CHP boilers in work (think a 4 cylinder car engine powered generator that runs on natural gas and then a cooling jacket around the exhaust manifold that is used to provide heat to the building for hot water and heating purposes) and we had employed a data specialist to install cables from the central 'brain' to each piece of CHP equipment and terminate every end using RJ45 crimps. I had to go back up to the site after 6 months and replace every cable with new cables with factory terminated RJ45's because the crimped ends were just vibrating themselves to bits all the time. Maybe if the RJ45 ends were potted with some kind of epoxy compound they might have stood a chance, but they certainly didn't last.
I think all the network cable people crimp is solid wire. Which would not be great for anything with vibration or that flexed
 
The Ubiquiti "passive" system only sends power on two of the four cable pairs, so it is not suited for high power as the high-power IEEE standards like bt which use all four pairs-- thus less than half the current on each connector pin. (Less than half because less voltage loss in the cable further increases power transfer efficiency.)

Waterlogged twisted pair cable will pass all DC tests but have improper RF impedance and high RF attenuation, and fail to link. Gel-filled cable should be specified for permanent outdoor installations.
This is outdoor rated network cable with a messenger wire to allow us to string between telephone poles using a messenger drop clamp. I think gel-filled is for burial.
 
This is outdoor rated network cable with a messenger wire to allow us to string between telephone poles using a messenger drop clamp. I think gel-filled is for burial.

I've seen gel-filled cat5e/cat6 messenger wire cable before.

And the telephone company here (was Contel, then GTE, then Verizon) used gel-filled cable for both aerial AND buried applications. The techs from the former Bell Atlantic areas don't like dealing with it, from what one of them told me.
 
Yes. One is inside in an attic. The other is in an outdoor enclosure that should keeps things dry unless there is a monsoon.
OK. Are either of them under tension? What about vibration? I assume there could be some harmonics from wind?
 
Back
Top