POE Lighting

Status
Not open for further replies.

310 BLAZE IT

Senior Member
Location
NJ
Seems like this would be great for residential.. it doesn't make much sense running #12-2 wire for a few bedrooms with less than 100W of lights, and you would be far under the threshold for distance on PoE. Maybe an idea for next generation canless LEDs
 

cvillej17

Member
Location
Pittsburgh
Occupation
Technical Advisor
Seems like this would be great for residential.. it doesn't make much sense running #12-2 wire for a few bedrooms with less than 100W of lights, and you would be far under the threshold for distance on PoE. Maybe an idea for next generation canless LEDs
It absolutely is, as long as there is an early adopter of tech as the homeowner. Some knowledge of networks and IoT devices is extremely helpful, but for the most part, once the system is installed, it just works. For a home, it's safe and adapts well to the significantly electronic environment. PoE Texas makes a PoE powered combination network jack and USB-C wall receptacle, so power and data are available for most modern electronic devices. Throw in some table lamps converted to DC using DC bulbs which could run off a PoE lighting node and it's a slam dunk.
Software makes it easy to zone and control light fixtures. It's simple to add some of the commercial features like motion sensors, daylight harvesting, and timeclocks. The UI can be manipulated from a smartphone.
On top of everything else, it can be powered by renewables like solar and wind without the DC-AC-DC conversion, making it that much more efficient.

I think whoever is first to market with a really solid UI that incorporates all the advantages of PoE technology will retire a very wealthy person. In my mind it makes too much sense not to catch on and take off.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
I think there is some chance down the road of most rooms having receptacles that have 120VAC, a couple USB charging ports, and a couple POE ports for data and lamps that can be powered from them.

However, I think the codes will need some tweaking before this is a common choice.
 

d0nut

Senior Member
Location
Omaha, NE
Seems like this would be great for residential.. it doesn't make much sense running #12-2 wire for a few bedrooms with less than 100W of lights, and you would be far under the threshold for distance on PoE. Maybe an idea for next generation canless LEDs
Except you are trading a $20 circuit breaker for a $1000 PoE switch that should probably be replaced every 5-10 years because it will no longer be supported.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
Except you are trading a $20 circuit breaker for a $1000 PoE switch that should probably be replaced every 5-10 years because it will no longer be supported.
I think you are probably looking at receptacles with a power supply that can supply a couple USB charging ports and inject power into POE jacks. I don't think there is any need for a POE switch. Just bring a data cable in and you can inject the power right there. No need for a POE switch somewhere.

Could even have an ethernet switch built in to the receptacle. The chips to do this can be had for a few dollars these days.
 

cvillej17

Member
Location
Pittsburgh
Occupation
Technical Advisor
Except you are trading a $20 circuit breaker for a $1000 PoE switch that should probably be replaced every 5-10 years because it will no longer be supported.
But in exchange you're also trading heavy gauge copper, an electrician, and death for a system that runs devices more efficiently, safely, and longer with the ability to be actively monitored and adjusted. Plus I'm sure you have no qualms with replacing your multi-hundred dollar smart phone every couple years. A $1000 switch today will be $100 in ten years and more reliable, like everything else that's based in advancing technology. How long do you think it'll be before that $20 breaker is replaced by $100 breakers that can be remotely controlled? And how many breakers are in a modern panel, anyway?
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
Seems to me that this whole PoE thing is growing into the "floobie-dust" realm, especially trying to push more and more power over 24ga conductors and plugs not designed to carry current. Seems the IT crowd doesn't understand that. I would suggest a new "Power with Ethernet" cable, a CAT 6 cable that has a pair of 18ga (or larger) power conductors molded on the jacket in a figure 8 or Siamese fashion. Then a new style "RJ-45" plug could be developed that would accommodate the power conductors outside of the data.

-Hal
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
Seems to me that this whole PoE thing is growing into the "floobie-dust" realm, especially trying to push more and more power over 24ga conductors and plugs not designed to carry current. Seems the IT crowd doesn't understand that. I would suggest a new "Power with Ethernet" cable, a CAT 6 cable that has a pair of 18ga (or larger) power conductors molded on the jacket in a figure 8 or Siamese fashion. Then a new style "RJ-45" plug could be developed that would accommodate the power conductors outside of the data.

-Hal
the whole point of POE is to use the existing stuff and not have to reinvent the wheel. it works pretty well as long as what you are powering is within the limited capability of the wiring.
 

Flicker Index

Senior Member
Location
Pac NW
Occupation
Lights
You are correct - some fixtures come with integrated drivers, so "simply remove the AC driver" was an oversimplification and technically incorrect. I don't know enough about each LED fixture manufacturer's build process to say how they integrate their components. The concept for a PoE fixture in general, though, is to build a PoE LED fixture that does not have a driver, but instead only an LED array, which is what HE Williams and others do with their production AC fixtures.
You aren't just "technically incorrect". You're grossly incorrect. You have no idea what you're talking about yet you call yourself a technical advisor? LEDs are not resistive elements and you can't just hook up raw LED elements to a voltage source and be like voilà. A standard PoE power supply is a voltage source. You can't just hook up raw LED elements to a low impedance source. A current source power supply wouldn't be a standard PoE power supply. It would be an external LED ballast with an RJ45 connector.

There are rudimentary LED flashlights that use a bunch of LEDs in parallel connected across three AAA batteries in series. The impedance of battery becomes the LED ballast causing batteries to heat up and this becomes part of the ballasting loss. Dinky little interior LED lamps for automotive interior lighting is often just 3 or 4 LEDs in series with a series resistor as a ballast, but the regulation is terrible and if you scale it up, the ballasting loss would be unacceptable. 25-30% ballasting loss is fine for dome light, but not for luminaires that emits thousands of lumens. It will cost you big time in output lumen regulation and lumens per watt.

Low production volume alternative voltage LEDs sometimes use mass produced LED arrays intended for 120v lamps and pair it with a boost converter based 12/24v input LED ballast, although these special application devices don't have DOE efficacy standards to meet and because of the transformation ratio, the lumens per 12v side input is usually worse than 120v lamps. If you're going to do PoE power supply, you'll have to add the power supply loss as well.

Also, you mentioned using a UPS. You know, APC SmartUPS line-interactive UPS in the 1,000 to 1,500VA scale has a standby power consumption of like 40W, or 350kWh/year or about the same as a high efficiency refrigerator.
 
Last edited:

Flicker Index

Senior Member
Location
Pac NW
Occupation
Lights
led board.jpg

Here's how they do some consumer LED luminaires and lamps these days. Both luminaires and lamps use pre-made boards like this.
This one is made from seven 18V LED elements and one 9v LED element all in series. Using two different elements seem peculiar, but they do this in order to match the voltage closely with the intended line voltage so that it can be used with a simple series electronic rheostat linear ballast. This one is specifically designed for 120v AC 60 Hz in US/Canada. It is quite sensitive to line variation, so it wouldn't be appropriate for Mexico where the voltage is 127v due to their derivation from 220Y/127V or other places where voltage is unstable.

The big connector simply puts a 46uF capacitor across the string. 120v AC goes into the little connector. The 4-pin thing is a rectifier. The 8 pin DIP is an electronic rheostat LED ballast which requires line voltage to be matched closely to LED array. This whole thing is compatible with phase-dimming.

To use it in with 48v DC input, the array would have to be made completely differently or paired with a 48v input, dimmable, 10W, constant current 60mA output LED ballast. Such LED ballast is going to cost quite a fortune because of low volume being made. This is especially true when 90-95% efficiency is to be expected.
 

cvillej17

Member
Location
Pittsburgh
Occupation
Technical Advisor
You aren't just "technically incorrect". You're grossly incorrect. You have no idea what you're talking about yet you call yourself a technical advisor? LEDs are not resistive elements and you can't just hook up raw LED elements to a voltage source and be like voilà. A standard PoE power supply is a voltage source. You can't just hook up raw LED elements to a low impedance source. A current source power supply wouldn't be a standard PoE power supply. It would be an external LED ballast with an RJ45 connector.

There are rudimentary LED flashlights that use a bunch of LEDs in parallel connected across three AAA batteries in series. The impedance of battery becomes the LED ballast causing batteries to heat up and this becomes part of the ballasting loss. Dinky little interior LED lamps for automotive interior lighting is often just 3 or 4 LEDs in series with a series resistor as a ballast, but the regulation is terrible and if you scale it up, the ballasting loss would be unacceptable. 25-30% ballasting loss is fine for dome light, but not for luminaires that emits thousands of lumens. It will cost you big time in output lumen regulation and lumens per watt.

Low production volume alternative voltage LEDs sometimes use mass produced LED arrays intended for 120v lamps and pair it with a boost converter based 12/24v input LED ballast, although these special application devices don't have DOE efficacy standards to meet and because of the transformation ratio, the lumens per 12v side input is usually worse than 120v lamps. If you're going to do PoE power supply, you'll have to add the power supply loss as well.

Also, you mentioned using a UPS. You know, APC SmartUPS line-interactive UPS in the 1,000 to 1,500VA scale has a standby power consumption of like 40W, or 350kWh/year or about the same as a high efficiency refrigerator.
UPSs exist and people buy them to mitigate the effects of a power outage. I'll bet in every instance they've checked to see if they'd rather use (or are allowed to use) a diesel generator that requires testing every month and decided a UPS is a better solution. What's your point?
My second paragraph to Rock86 was a description of the node that runs a PoE fixture. Did you miss that section when you replied to me initially? Sorry you spent time laying out your flashlight knowledge and such, but I said there is no AC driver. There is a DC driver. It's called a node in several of the PoE systems on the market. Each node is custom programmed to tailor the current to the LED array. It is precisely an external LED ballast with an RJ45 connector. Here is a picture of an LED array for a PoE fixture. Pretty straight forward (and HE Williams has nearly 1000 different PoE light SKUs built exactly the same way, so apparently it's a thing). Clearly you know a bunch about the technical details for AC lighting and LEDs. Let me know if I can help you with the PoE control part.
1641849592135.png
 

cvillej17

Member
Location
Pittsburgh
Occupation
Technical Advisor
I have not seen that at all. Just lighting mostly, and it is well suited for that, at least in residential settings.
Actually, network cabling IS powering entire commercial buildings, although the main power transfer method is not using a PoE standard.
Volt Server uses pulsed DC power on a network cable to transmit 2000w up to 300m while staying under NEC Article 725 / CEC Rule 16-200 wiring methods and Class 2 wiring techniques (according to them - don't shoot the messenger). It's primarily power transfer but there is low bandwidth data in the cable, as well.
https://voltserver.com/
They don't use RJ45 jacks on the cable.
 
What's mudding the waters is that there's real, actual, Power-over-Ethernet (which includes the Ethernet part), and there's Power-over-CAT5 which is just using the 4-pair wiring method and is emphatically NOT PoE.

If ya ain't got the Ethernet, it ain't PoE.
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
Volt Server uses pulsed DC power on a network cable to transmit 2000w up to 300m while staying under NEC Article 725 / CEC Rule 16-200 wiring methods and Class 2 wiring techniques (according to them - don't shoot the messenger).
On the face of it, that's a nonsensical idea, as the Class 2 rules limit the power to 100VA, as I understand them. Do they have any plausible argument otherwise? Are they trying to exploit some limited duration surge exemption in the listing standard for Class 2 devices, and then repeatedly pulsing the DC power above and below that 100VA limit?

Cheers, Wayne
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
On the face of it, that's a nonsensical idea, as the Class 2 rules limit the power to 100VA, as I understand them. Do they have any plausible argument otherwise? Are they trying to exploit some limited duration surge exemption in the listing standard for Class 2 devices, and then repeatedly pulsing the DC power above and below that 100VA limit?

Cheers, Wayne
Maybe they are running 100 VA in 30 conductors simutaneously.
 
Status
Not open for further replies.
Top