How could you make a product so limited, and not make those limitations clear in the Specs ?
Let's start with the tech specs from the basic product information page for the GS105PE (removed non-PoE related stuff, and yes - the same information was on the previous format Web):
===
TECHNICAL SPECS
Number of PoE ports : 2 Total PoE Power budget (Watts) : PoE pass-thru: 19W with 802.3at / 7.9W with 802.3af input power Internal/External : ***NO EXTERNAL POWER ADAPTER: This product requires PoE input power (802.3af or 802.3at) to operate; it has no means to power from AC. Max consumption (Watts) : 22W
Is this because of Voltage ( 12V ), or Class 0 ? The Cameras and DVR Power Supply are IEEE 802.3af
Assuming you tried to power the GS105PE from the DVR where one PSE port does 802.3af, providing max 15.4W, with a guaranteed power available of 12.95W at the powered device (PD - that's the GS105PE) the max loss allowed on a full length Gigabit Ethernet connection cable. Taking away the power the switch does require to operate the switching core and four Gigabit Ethernet connections at full length and speed, does leave a mere 7.9W available for the power sourcing equipment (PSE). This is way less than the 12W your PD camera does require. Talking IEEE standards, this allows to drive a single Class 2 device only.
The 12V might be some direct DC voltage which can be used to power the camera externally, this is not relevant here. Just for completeness: According to 802.3af (aka. 802.3at type 1) the voltage at the PSE is 44...57V, the voltage at the PD is 37..57V, the max power is 350mA.
A device announcing as a Class 0 device does _reserve_ 15.4W at the PSE. As this power is not available when feed from the 802.3af PSE, the PoE power negotiation will not power up the port. And it does not matter if this Class 0 device does effectively take 0.1W or all the 12.95W. So yes, that's the correct answer for your question.
Even if I try a 45V POE Injector on Port 5, your Troubleshooting guide says it can only handle one Class 0 device
Whatever PoE inector that is - could be PoE, voltage is in the range I've showed from the standards above. Probably a 802.3at PSE providing max 30W at the PSE, with a max. guaranteed 25.5W at the PD. Either way, an IEEE compliant PoE device does have a well defined voltage range - there is no need to mention the voltage (typical Amazon or eBay crap) - IEEE 802.3at is perfectly sufficient. In the meantime you should understand the GS105PE does require power the switch does require to operate the switching core and four Gigabit Ethernet connections at full length and speed, does leave a 19W available for the power sourcing equipment (PSE) ports. So one Class 0 at 15.4W on one port, plus some which can't be used - don't nag me where the additional 1.5W are going to.
I have 4 other No-Name Brand POE Pass thru amplifiers and splitters ( under $15 ) that work fine
Nothing anywhere near to the IEEE PoE standards. Some simplified stuff hopefully retaining a reliable Gigabit Ethernet link, while being able to bring e.g. 12V DC to the far end. No standard, no security protecting accidentally plugged network devices into the link carrying Ethernet and power, ... nothing to compare.
Here again, to make this reply complete again: There are PoE/PoE+/PoE++ switches in the field where the power budget is less than the number of ports multiplied with the max. power available. Advanced units are sometimes designated as FlexPoE. Here we face either the ability to manage the PoE port priority (low, medium, high), while non-managed are simply using the priority from port 1 (highest) to port n (lowest). If there is more power required, the low prio resp. highest port power is cut until the switch is within it's max. PSE power again.
If the GS108PE would do the same with the two Class 0 devices (while 802.3at powered), both cameras might work to some point, e.g. when more processing power is required depending on image complexity, and the IR LED are coming on. Then one port would cut of the power. That's why it does not make much sense - the average user would not understand why the second camera would go down randomly.
In the Tech Specs, I did miss the "7.9W Limit" on af, but still one camera should run
While Spec'd at 12W, the cameras run below 6W
No, and no.
It's not about what it's spec'd, and it's not what drawing under whatever condition.
It's about the fact that the camera does request Class 0 - this i the only spec applicable (there are some resistors soldered in, and some pulsing do the negotiation - in newer/higher IEEE PoE+/++ standards also makes use of LLDP) - this makes the switch reserving the power for a class 0 device. If not available == no power.
For a reason, the specs are on 6W - I guess again it's about processing and/or IR LEDs. If the camera would operate under all conditions below 6W, they could have defined it to be a Class 2 device (read max 7W from the PSE, 6.49W max at the PD). And now it would get powered.
Obviously they have chosen Class 0 - either because the camera can draw more then these 6W, or by laziness - because this low cost NVR/camera combo maker never had the idea that a more granular specs could be required. Simply because their design intention is on NVR port, one camera. Problem solved for the NVR/camera maker going to market.
There are PoE+/++ devices which are lowering capabilities if the higher product class power isn't available, e.g. acess points operating on lower power of the last few watts are absent.
And even an "at" supply allows only one Class 0 device. This detail is very well hidden
Just adding the info that info might not be sufficient - because the IEEE PoE design details and class levels are not commonly known and it isn't commodity know-how for many newcomers.
We can argue with Netgear that the content of the troubleshooting document (especially the tables) should be made available direct on the product specs page. I have done that btw. a long time ago...I'm not Netgear.
I'm not the only one. Many others here are equally surprised.
Hey you are not alone - I've hit the similar questions scratching my head when planning and testing the GS105PE capabilities for the first time.
Most of the confusion is coming because we're humans having troubles and are mis-matching electrics lessons 1..3 basics with what is IEEE PoE, device classes, and the guaranteed power design.
Just for info, as suggested elsewhere, it does work if you reboot the switch
With "af" source, I can power one device, as specified ( even Class 0 )
This is clearly a BUG !
Can't agree more! A Class 0 device must never be powered when the GS105PE is feed by an 802.11af PSE. Definitively out of specs!. My related complaints are not heard @YeZ
