I have a network/PoE mystery on my hands

105437

BIT Beta Team
Jun 8, 2015
2,160
1,052
I have two long (280’) Cat 6 runs to my mailbox. Through testing, it appears that one of the runs is bad. The thing that is throwing me for a loop, is that when I use either run and plug the camera in, I can hear the camera power up. When I unplug the camera I can hear it click and power off. I’m trying to understand what possibly could be happening to not allow one of the runs to get the camera up to the point where I can connect to it via a web browser. I have used a network cable test tool, and both runs test successfully, with no shorts, breaks or miss wires.

Thanks!
 
That is a long stretch and approaching the theoretical , so you could be right on the cusp of the camera limits for power and the camera is right at threshold so it doesn't connect. It could be getting enough power to cycle on, but not enough to transmit video back.

Have you already set the camera up to be able to see it on your network and can you see the camera on your network with a short cable?

Does the same camera work on one cable and not the other? If so, it could be a kink or interference at that distance for the one camera.

Or it could be the POE port is not sending out enough to both at the same time. Try one cable, one camera, and only one POE port in use.
 
Yes, I’m able to ping and see the camera via web browser using shorter cables from each port from the patch panel. The same camera works on one long run, but not the other. I terminated new ends on the cable that fails using 568B.
 
If you haven't tried, switch the long run that doesn't work to the port that works to confirm it isn't a power issue with a specific port for a longer run.

Just have one camera connected to the POE switch for testing purposes to rule out a power issue with the POE switch. At that distance, it could be trying to push more wattage through and thus going over total capacity of the switch.
 
As an Amazon Associate IPCamTalk earns from qualifying purchases.
What switch do you have? Could be a switch problem not able to supply the required power constantly or could be related to more demand for example at night if the IR LEDs come on?
 
  • Like
Reactions: sebastiantombs
Thanks, I’ll plan on doing some more testing later this afternoon. I’m using a Uniquiti UniFi 24 port POE switch. I know that switch supports POE+ on 16 of the ports, I need to figure out which 16 ports.
 
Do you have a cable tester to ensure continuity on all pairs end-to-end? If only the RX or TX pairs are compromised, you would not be able to access the camera GUI despite it emitting sounds of being powered.
 
IIRC, some of those Ubiquiti UniFi POE switches support not only 802.3af/at but can provide passive +24 VDC for legacy UniFi and other Ubiquiti devices if configured as such....I suggest you insure one port didn't get by inadvertently configured as passive +24 VDC instead of the desired 802.3af/at.
 
Do you have a cable tester to ensure continuity on all pairs end-to-end? If only the RX or TX pairs are compromised, you would not be able to access the camera GUI despite it emitting sounds of being powered.
Yes, my cable tester checked the individual pairs as well... no issues. Here's a simple drawing of how this is run.

IMG_3232.jpg
 
IIRC, some of those Ubiquiti UniFi POE switches support not only 802.3af/at but can provide passive +24 VDC for legacy UniFi and other Ubiquiti devices if configured as such....I suggest you insure one port didn't get by inadvertently configured as passive +24 VDC instead of the desired 802.3af/at.
I'll need to read up on that since I'm not sure how to check that in the Unifi Controller.
 
Personally, way too many places for problems and too many places for losses. Check the run that doesn't work at each of those RJ45 junctions. Make sure they're OK as well. Plug the longest cable you have into the demarc and make sure the camera suns on that. One of the longer runs could have a kink or bend in it that is killing data but not power but you'd need a TDR to be able to see that.
 
Personally, way too many places for problems and too many places for losses. Check the run that doesn't work at each of those RJ45 junctions. Make sure they're OK as well. Plug the longest cable you have into the demarc and make sure the camera suns on that. One of the longer runs could have a kink or bend in it that is killing data but not power but you'd need a TDR to be able to see that.
I used my cable tester from the patch panel to the mailbox and also from the Dmarc to the mailbox, all tests were good. The Unifi Controller "sees" both cameras and shows they are both using ~3.5 watts.
 
Either cam not working on cable 2 must indicate a problem with that cable, no?

I think somebody mentioned it earlier, check that that cable isn’t kinked or worse have a cut or cable clip through it etc.
 
  • Like
Reactions: sebastiantombs
it may depend on the quality of the cable tester. Most cheap under $50 cable tester check for continuity, but do not run frequency data up and back. You can have cracks in the cable that will pass a continuity check but not a data check. Replace the cable or rent a high quality tester.

What is the manufacture and part number of the long cable run ?
 
Continuity, while important, isn't the whole story. Kinks, bends, connectors, nicks and cuts all effect data transmission. Back when I was installing we used a TDR on every drop to verify length and prove data capacity. I don't do that in my own installation, at home, but want to stress strange things can happen that make you scratch your head.
 
Either cam not working on cable 2 must indicate a problem with that cable, no?

I think somebody mentioned it earlier, check that that cable isn’t kinked or worse have a cut or cable clip through it etc.
The problem is that the cable from the Dmarc to the mailbox is in a conduit, underground. I'm likely just going to end up pulling the cable out and installing a new one.
 
  • Like
Reactions: SouthernYankee
it may depend on the quality of the cable tester. Most cheap under $50 cable tester check for continuity, but do not run frequency data up and back. You can have cracks in the cable that will pass a continuity check but not a data check. Replace the cable or rent a high quality tester.

What is the manufacture and part number of the long cable run ?
The tester I used is a Southwire that was under $50, could be it's not seeing the whole picture. I can get a nice Fluke from work to use.