I read that PoE will damage Hubitat. Why is that? If you go onto any corporate network it is all PoE to the desktop, mainly for the hard phone. In the home, the quality cameras are PoE, which is great when you have the wiring. I am concerned as I add a PoE switch into my network, how could it damage other devices? Why would it damage my Hubitat?
Newer poe systems (ie intelligent ones that only push power when Poe devices are detected) won't damage it. A lot of us use Poe splitters for
Power. Warning is more of a "Better Safe than Sorry" thing
1 Like
Yes but also the terms say you could potentially void your warranty so be careful!!!
https://docs.hubitat.com/index.php?title=Registration_and_setup
This is what bothers me with that item in the terms of service, POE is intelligently delivered based on standards and I have been doing POE since the early days with VoIP phones in call centers.
1/2 of my ports on my network are POE 802.3at. Following standards during link negotiation determines if power is to be applied to the pairs. I do have managed switches and only turn on POE based on port profiles or individual port configuration but any ethernet device that follows the standards should be safe plugging into a POE port.
With that statement in the terms of service tells me that the Hubitat may have compliance issues. I also use POE splitters for most of my USB powered devices (Hubitat, Pi's, Camera's etc.)
It may be more of a CYA thing than compliance.. much easier to issue a blanket statement like this and then deal with things on an individual basis.. but that is just my take.. I'm NOT affiliated with Hubitat, Inc other than being a community member and "fanboi"/"fanperson".
2 Likes
We have seen a number of hubs that were fried by plugging in POE. Standards are nice, but real world experience suggests that this should be avoided. Hubitat Elevation ethernet port does not accept POE.
2 Likes
not all PoE is intelligent. there is passive PoE that will send out voltage and can/has damaged NICs
2 Likes
None of these devices accept POE and have ethernet port: Laptop docking stations with ethernet, desktop computers, Connected TV's, Roku's, Synology NAS, HP Printers, Tablo OTA DVRs, Home Run receivers, Ooma VoIP ATA, etc. etc. etc. None of them have the TOS that says plugging the device into a POE port voids warranty. It may to be prudent just to put into the instructions that the Hub isn't powered by POE and call it a day. Or maybe add that if it's determined that your Hubitat was damaged by POE warranty is void?
Agreed, I haven't had to use any passive POE injectors but defiantly they do exist.
This language was added after we replaced a number of fried units under warranty. People don't read instructions. Generally, one puts disclaimers into warranty language for things that won't be covered by the warranty. Given that we live in a click OK and go world, where else would this belong, if not the TOS?
Or, are you suggesting we should just eat the cost for those who don't read?
2 Likes
Never would I ever ask you guys to eat the cost. You have to run your business as you see fit.
This is probably something that should have been investigated very carefully.
If one simply plugs a hub into a PoE capable switch, then there is the potential for power on the hub's Ethernet port. This is something that ALL ethernet devices have to work with without exception. Damage could only occur if there is a very serious defect in either the switch or the hub implementation of Ethernet. Possible, but pretty damn unlikely. FWIW, I doubt the "not compatible with PoE ports" phrasing is even permissible under the licensing terms of the Ethernet standard.
On splitters, unless there is an egregious design defect, a splitters should be able to pass PoE power through to the downstream port. So even if you plug a PoE device downstream of a splitter, it simply does not work. With no PoE power is available to the Hubitat so there is no way to fry the Hubitat's port.
I suspect HE hubs were damaged by being plugged into passive POE supplies - those that provide a continuous voltage on some wires in the connector. Many end devices get damaged this way.
However I've never seen a device damaged by a reputable active POE source of 802.3af, 802.3at or 802.3at POE+. But there could always be poor devices that don't implement those standards correctly, especially the cheap inline devices. All my HE hubs and indeed all my devices are connected to POE/POE+ capable switch ports.
3 Likes
Probably. Most of those devices are designed to work with a very specific downstream device, and generally guarantee adherence to Ethernet standards at the upstream end only. Not sure why anyone would plug a device such as Hubitat into one though.
Yes, mine as well. I actually don't have any non PoE switches any more. 
A couple of my hubs run on PoE Texas splitters. Reputable, solid gear, used in a lot of home automation installs. I know a lot of people use the UCTRONICS splitters as well.
The thing about Ethernet is there is no mate first / disconnect last built into the connector. On disconnect you can get some unintentional current paths that can take out transformers and terminations, especially if the manufacturers reference designs are followed blindly.
POE is a negotiated standard, no current except for signal current connects first, then the POE protocol negotiates the level of current to be delivered. This is a pretty good article on how this works:
PoE Negotiation of the Power [Power Step-Down of the PoE Protocol].
This type of issue would not pertain to the Hubitat connecting to a PoE switch as the Hubitat is not a powered device.
The only way that I can think of getting unintended power here would be if you could unplug a running PoE device from a jack and plug the Hubitat in its place within a few hundred milliseconds. I doubt that it is physically possible.
Agreed for most Active PoE network switches…not so much for Passive PoE switches which will fry nonPoE devices.
We are going to do a bunch more testing with various POE devices.
7 Likes
I agree with you. Something must not have working as it should have been to cause the failure of the hub. It is possible that if the splitter was disconnected from the ethernet while powered, voltage could have got on the data lines.
