Splitting one PoE-ethernet cable from my NVR5216-16P-4KS2E for two cameras?

[reflection]

Quick follow-up on the 4-port poe extender that I linked. More testing with it and I turned on the IR LEDs and my power draw for the three cameras went over the 15.4W limit for 802.3af. All three cameras shut off and would not come back online.

I thought it would not be an issue because my switch supports 25W+ (802.3at) and the 4-port poe extender indicated that it supports a 802.3at input.

Upon debugging my switch, I noticed that the extender does not negotiate for more power. Therefore the switch was only granting the extender 15.4W by default. When the extender tried to draw more power, the switch was smart enough to shut off the power to the port. The fix was to tell the switch to allow the port that connects to the extender to have more power. After that, things worked fine.

In summary, the 4-port poe extender does not negotiate power based on standards for PoE. Therefore you have to turn off PoE negotiation on your switch port if you need more than 15.4W.

 

[Safetyfirst]

Quick follow-up on the 4-port poe extender that I linked. More testing with it and I turned on the IR LEDs and my power draw for the three cameras went over the 15.4W limit for 802.3af. All three cameras shut off and would not come back online.

I thought it would not be an issue because my switch supports 25W+ (802.3at) and the 4-port poe extender indicated that it supports a 802.3at input.

Upon debugging my switch, I noticed that the extender does not negotiate for more power. Therefore the switch was only granting the extender 15.4W by default. When the extender tried to draw more power, the switch was smart enough to shut off the power to the port. The fix was to tell the switch to allow the port that connects to the extender to have more power. After that, things worked fine.

In summary, the 4-port poe extender does not negotiate power based on standards for PoE. Therefore you have to turn off PoE negotiation on your switch port if you need more than 15.4W.

If we were running cameras in forced day mode with no IR then in theory 3 cameras could work then? So this splitter could be better for indoor use in hallways etc which are always lit or if we have hikvision/ dahua daylight at night cameras with no IR?

 

[pozzello]

Good info, reflection! another solution if the PoE port can't be configured to allow more power would be an external PoE injector at the head (switch port) end. A bit clunkier, of course, but workable...

 

[reflection]

If we were running cameras in forced day mode with no IR then in theory 3 cameras could work then? So this splitter could be better for indoor use in hallways etc which are always lit or if we have hikvision/ dahua daylight at night cameras with no IR?

The PoE splitter works with three cameras. Probably four as well as long as your total power consumption is below 25W. I'm using 16.3W for three cameras (one with IR on) so there is enough power for another camera.

I don't know if it will work with a power input of 15.4W (802.3af PoE) because the description says the input should be 802.3at PoE.

 

[Ratfink11]

I run and terminate my own cables and have never followed T568a or T568b. There are no issues with older Ethernet devices as long as you follow the guidance I mentioned (1-2, 3-6, 4-5, 7-8 matching on both ends). The issue is if someone uses a certain color scheme on one side, and another person uses a different scheme on the other end.

I understand what you are saying. However, every government contract bid we won and completed required the T568a wiring protocol as its a TIA Standard. Its also the same wiring protocol for all pre made network cable. A little research will explain the reasons at TIAonline.org. Im a big fan of Standards and obviously so is the rest of the world. Standards are the reason an inch stays an inch and a millimeter stays a millimeter. If others prefer to make up thier own wiring scheme its just one more issue to troubleshoot if a problem arises. Which is what exactly is happening in this case. Why would anyone not wish to follow a STANDARD to eliminate troubleshooting steps? When I am paid to do a job for anyone I follow Standards and run a network cable for each camera plus extra runs just in case. Im being paid to do the job correctly. Throwing in extra splitters, not following a common worldwide wiring Standard, not testing every cable, or using the correct heads or crimper is what todays generation considers GOOD ENOUGH QUALITY. I know when we complete a job everything works. If there is a problem we fix it then and there. I dont have the time or money to fix it right the 8th time with bubble gum, paper clips, and splitters. There will continue to be gremilins and intermittent issues that will never be found. Fix it right once or keep fixing it forever. Spend 2 hours now and fix it right or months fixing it later.

 

[reflection]

I understand what you are saying. However, every government contract bid we won and completed required the T568a wiring protocol as its a TIA Standard. Its also the same wiring protocol for all pre made network cable. A little research will explain the reasons at TIAonline.org. Im a big fan of Standards and obviously so is the rest of the world. Standards are the reason an inch stays an inch and a millimeter stays a millimeter. If others prefer to make up thier own wiring scheme its just one more issue to troubleshoot if a problem arises. Which is what exactly is happening in this case. Why would anyone not wish to follow a STANDARD to eliminate troubleshooting steps?

Nothing wrong with following standards, but different countries follow different standards. This is an international forum so we can't expect everyone to follow T568A.

My information was more about explaining the wiring sequence regardless of color code standard. There is no electrical difference between T568A and T568B wiring patterns (or the pattern that I personally use at home). The laws of physics don't change in terms of EMI and crosstalk. Knowing the correct wiring sequence is important for troubleshooting if you already have existing UTP runs. The OP is not going to re-terminate all his runs if he doesn't have to. We are trying to troubleshoot the single run that he has an issue with.

 

[alastairstevenson]

There is no electrical difference between T568A and T568B wiring patterns (or the pattern that I personally use at home).

Yes, because the needed twisted pairs are still maintained in both standards.

My information was more about explaining the wiring sequence regardless of color code standard.

A scheme that arbitrarily terminates a cable that consists of 4 twisted pairs will not match the requirement that the RX+/RX- and TX+/TX- differential signals must be sent down a twisted pair in the cable, such that crosstalk and common-mode signals cancel each other out, and signal integrity is maintained as best it can be.
It's a tribute to the signal processing electronics at each end, and to the data integrity checks and repairs by layer 2 protocols that valid data still gets through, despite the signal abuse on the way.

The laws of physics don't change in terms of EMI and crosstalk.

Yes, exactly.
Which is why the wiring scheme is designed to minimise their effect when applied as designed.

 

[reflection]

A scheme that arbitrarily terminates a cable that consists of 4 twisted pairs will not match the requirement that the RX+/RX- and TX+/TX- differential signals must be sent down a twisted pair in the cable, such that crosstalk and common-mode signals cancel each other out, and signal integrity is maintained as best it can be.
It's a tribute to the signal processing electronics at each end, and to the data integrity checks and repairs by layer 2 protocols that valid data still gets through, despite the signal abuse on the way.

I don't think anyone is suggesting that you arbitrarily terminate cables. No one said that. The point in the previous postings were that the pairs have to match the pins properly. Four pairs, 8-pins, this pairing: 1-2, 3-6, 4-5, 7-8 regardless of the color code you choose. I guarantee you that if I made a cable using the sequence blue, blue-white, brown, orange, orange-white, brown-white, green, green-white on both sides will work fine and it is electrically equivalent (not just conductively, but in terms Gauss's law) to a cable done with the pattern defined in T568A or T568B.

 

[alastairstevenson]

I guarantee you that if I made a cable using the sequence blue, blue-white, brown, orange, orange-white, brown-white, green, green-white on both sides will work fine and it is electrically equivalent (not just conductively, but in terms Gauss's law) to a cable done with the pattern defined in T568A or T568B.

Sure - you obviously understand how the wiring needs to be connected to support the signalling.
But there is no technical benefit in not using the recognised standard wiring scheme - and no saving of time or effort in not using it either.

What I wanted to help avoid was a misunderstanding readers might get that provided an ethernet cable is just connected the same way at both ends that it will work properly.
Especially when they validate their crimping with a continuity tester.
We have had a few posts where people have not realised that there are some specific requirements that must be applied, and puzzled over the problems that occurred when they weren't.

 

[reflection]

Sure - you obviously understand how the wiring needs to be connected to support the signalling.
But there is no technical benefit in not using the recognised standard wiring scheme - and no saving of time or effort in not using it either.

What I wanted to help avoid was a misunderstanding readers might get that provided an ethernet cable is just connected the same way at both ends that it will work properly.
Especially when they validate their crimping with a continuity tester.
We have had a few posts where people have not realised that there are some specific requirements that must be applied, and puzzled over the problems that occurred when they weren't.

Totally in sync with you my friend. That's the great thing about this forum is that we all help each other. I agree the best practice for a new install is to follow the standard.

This is a little different because the OP has existing wire runs. I'm trying to help the OP find root cause of his issue. Four worked and one run didn't work that he had been troubleshooting for multiple days.

"Depending on the distance and environment, if you crimp the pairs wrong, you will get funky test results. Perhaps the cable that has the issue is a longer run or passes by some power wires?"

It would be cool if he were able to identify the root cause of his issue.

 

[manwald]

and no, that 4 port switch is powered by and re-distributes PoE.
this one: 4 Port IEEE802.3af PoE Extender for IP camera Extend 120m transmission distance | eBay

also avilabel on Ali express and Amazon...

I just want you tell you that I had success with the solution @pozzello mentioned

I tried two different solutions, this and this

Using the two cheap splitter-cables didn't help me, because my NVR didn't find the two cams.
But I had instant success with the 4-port PoE-splitter (BTW, I paid 20 USD). It took a couple seconds for my NVR to find both cams, but it's running flawlessly now.

So now I can use the other cable, that is free now, for my Unifi AP AC Pro, because the WiFi of my router really sucks.

Thanks for all your help, guys!