Cable tester or qualifier for IP cams?

sdyfgasd

sdyfgasd

n3wb
Jun 13, 2017
18
2
Up until now I've been using a basic toner and a pin tester for my wiring. I'm working on a site with 36 cams on a PoE switch. About 9 of those cams have constant signal loss events (hundreds a day per BI). I've replaced a couple of them, assuming it was the cams. The issue persists. I suspect it's the wiring. But a basic continuity tester won't do. I'm looking into professional cable testers from Fluke, Platinum Tools, IDEAL, etc. However, I don't know, if a fancy $500 copper tester will find my faults or, if I need a $2k qualifier. What do you use?
 
sdyfgasd said:
I suspect it's the wiring. But a basic continuity tester won't do.
Click to expand...
It will if you also check that the colour sequence of the wires in the RJ45 connector properly match the T568B standard, such that the twisted pairs are correctly maintained.
And that the cables aren't in close proximity to sources of electrical noise such as fluorescent lighting wiring or mains wiring.

sdyfgasd said:
I suspect it's the wiring.
Click to expand...
Quite a common source of problems is not adhering to the needed wire colour sequence.
If the twisted pairs are not maintained, signal intergity is severely compromised.
 
  • Like
Reactions: fenderman, looney2ns, tigerwillow1 and 1 other person
CCA or solid copper? How long is the run?
You could set up a ping, whhich runs continiously, to check for drop-outs.

Good luck,
FastB
 
  • Like
Reactions: c hris527
It depends on your expectations on your ROI. Do you run enough networking cable to justify a Fluke network tester? My company does since we are, after all, cabling installers :) However, the darn things come with a hefty price tag.
I'd advise hiring local telecommunications outfit to come certify the cabling. Much cheaper than purchasing of a Fluke. We work with many local IT guys that we may visit a couple times a year to certify (mind you...Cat6 certification means printing out the PASS test results in PDF format and giving to customer) a couple or bunch of cables at one shot.
Price should be per hour for labor (again, that depends on your region). Here in Northern Nevada, you would expect around the $100 range. Can easily test 30-50 cables per hour in an office environment where the cables are marked/labeled correctly. A little slower if warehouse atmosphere were renting a scissor lift is needed.
I do not know what cheaper testers accomplish for Cat6 certification. Do they test for NEXT? DB loss? Noise? Identify where the problem is?
 
Last edited:
  • Like
Reactions: c hris527
Holbs said:
I do not know what cheaper testers accomplish for Cat6 certification. Do they test for NEXT? DB loss? Noise? Identify where the problem is?
Click to expand...

That's partially my question. Most of them list cross-wire, shorts, opens, distance to the fault. But for crosstalk, it seems I would need to drop $8k for a high-end Fluke. I'm not doing that.
This is a pot farm with greenhouses, warehouse, storage containers, open grow. The cameras are in very different environments. Wiring is all over the place, too: some in conduits, some suspended overhead, some tacked to walls or tied to scaffolding. Some outdoor rated, some riser. The longest run is about 310ft... Don't know for sure, as that cable doesn't have markers (and I don't have a device, yet, that would tell me the length).
 
Hire local dudes for it.
To make things more economically feasible, someone (yourself or onsite employee) will have to identify the wiring. Since currently not identified, fastest way is to snapshot the POE switch port link lights. Have someone disconnect the RJ45 from camera and see what port link light on the switch goes out. There ya go....Cable #1 :) Repeat.
 
Are the cameras connected directly to the Poe switch and not through a patch panel first? What switch is being used? Have you cut the existing connectors off and crimped new plug connectors on?
 
sdyfgasd said:
How can there be any signal, if the wires are mis-matched?
Click to expand...
The camera connection uses two 2-wire signal circuits, each carried on a twisted pair using a differential connection. Noise spikes, interference, and crosstalk affect each wire of the pair equally, causing them to be cancelled out. If half of a circuit is carried on one twisted pair, and the other half on a different twisted pair, the cancellation of the bad stuff is lost, and the risk of crosstalk is increased. Used as a DC circuit, the pairing doesn't matter. For carrying high speed data it's critical.
 
Last edited:
  • Like
Reactions: alastairstevenson
iperf to check the bandwidth on the cable? Connect a laptop to both ends, run iperf and see what you get for numbers
 
Something I discovered in relation to the comment above: "And that the cables aren't in close proximity to sources of electrical noise such as fluorescent lighting wiring or mains wiring. " I've always adhered to this principle, hey, it's on the test.

Last year we had a big remodel on our house. I had previously installed my Cat6 cable in the ceiling running from my garage to the router. The electrician, unbeknownst to me as I was not there at the time of install, ran a 100 amp, 120 volt AC cable right next to my cat6 Ethernet cable. Not only that, he used cable ties to strap the Ethernet cable to the 100 amp cable! I flipped out when I saw that. I went to my computer fully expecting no Internet. But I had Internet. I did a speed test. Excellent streaming at 155 mbps. So, why didn't the 100 amp AC cable destroy my Internet signal??
 
ngcjim said:
Something I discovered in relation to the comment above: "And that the cables aren't in close proximity to sources of electrical noise such as fluorescent lighting wiring or mains wiring. " I've always adhered to this principle, hey, it's on the test.

Last year we had a big remodel on our house. I had previously installed my Cat6 cable in the ceiling running from my garage to the router. The electrician, unbeknownst to me as I was not there at the time of install, ran a 100 amp, 120 volt AC cable right next to my cat6 Ethernet cable. Not only that, he used cable ties to strap the Ethernet cable to the 100 amp cable! I flipped out when I saw that. I went to my computer fully expecting no Internet. But I had Internet. I did a speed test. Excellent streaming at 155 mbps. So, why didn't the 100 amp AC cable destroy my Internet signal??
Click to expand...
Because computers can work perfectly fine with dropped packets due to interference. An unbuffered video stream coming from the camera can be jittery, stop, etc
 
fenderman said:
Because computers can work perfectly fine with dropped packets due to interference. An unbuffered video stream coming from the camera can be jittery, stop, etc
Click to expand...
ah. This makes logical sense. My day job is to run data cabling for computers and such. Not for IP camera's. I might have to test this out and see.
 
A network card sends an ACK packet for each datagram it receives. If the sender doesn't get that ACK packet, it resends the data. There is no ACK equivalent in a video stream.
 
And consider the AC current.
More current, bigger field.
And if feeding a motor, BIG field changes when motor turns on/off.
 
You must log in or register to reply here.
Top Bottom