Speaking of PoE plus Apps in the IoT and Hot Buttons Galore!

Posted: December 10, 2015 in Facilities, General Commentary
Tags: , ,

The recent press release announcing Philips, Cisco, et al,  joint venturing to deploy and build Power over Ethernet (PoE) networks in lighting is going to fuel this discussion and create a stir, without a doubt. In the press release, all the current hot buttons were pressed with vigor, from App controlled lighting using smart phones to ties to the Internet of Things (IoT). The picture painted by this release, presentations on this topic, and other articles floating about, indicate a future where lighting breaks its bonds of wiring to be free to serve us all in magical, never before realized new ways, using less energy through magic DC power, finally severing us from the drag of AC power. It’s certainly got folks talking.

At the recent LED Specifier Summit in Chicago, I was asked by no less than 8 people what I thought about PoE, and whether it was going to be the next big disruptive innovation to strike lighting. Concurrent to this were phone discussions with technology providers and fixture manufacturers, asking similar questions. It was hard not to think that something was going on, as everyone seems to be all quivery about it. The problem is… I am not so sure what all the fuss is about, and whether anyone is really thinking this through. I like the concept of a distributed network style, low voltage DC lighting infrastructure. It solves fixture design issue, and presents intriguing possibilities for integrating controls, lighting and the IT universes together in ways our current system of isolation-in-high-voltage simply cannot easily address.

Advantages Impossible to Ignore

As we move into more electronics integration into lighting, it is hard not to look at the IT universe, and its capacity to deliver data in both wired and wireless networks reliably and effectively throughout large areas. It would be phenomenal to have that same level of inter-connectivity between controls and controlled lighting, with each fixture set up with its own address, and simple software interfaces to allow any fixture to be controlled by any control or population of controls, in addition to responding to global data, like time of day, light conditions, even weather conditions. No more stupid controls circuits where lighting I don’t want on is left on, because its hard wired to another I need to remain lighted. The concepts of Human Centric Lighting, and even those of the far-reaching Semantic Light concept cannot be obtained without a layer of control sophistication the existing lighting market struggles to deliver. HA! I just injected two more hot-button topics the press release had not even mentioned….

The inter-connectivity of a lighting infrastructure that has close ties to the IT universe, Bluetooth utility, wireless network visibility and access also solves issues of controlling portable lighting, specifically at the task level, that are normally plug connected with no wired control connection. To be able to address these products as components of a larger lighting package and system would bring them finally into the picture as more than accessory add-ins. In this, energy code compliance could integrate all forms of light, from portable to daylight harvesting, into one unified total system, monitored, commissioned and controllable using a single controls layer.

While there is movement toward higher and higher voltages in LED packages, the fact remains, LEDs themselves are low voltage devices. Distribution of remotely controlled, current limited 24VDC, or even 48VDC makes more sense with LEDs than any other source, and resolves a great deal of the issues of packaging driver, power conversion and light source into luminaires. Portable, surface wall and ceiling mounted products could become much cleaner and slimmer, freed of housing chunks of non-luminous hardware. Dimming control within this proposed data biased infrastructure would be far more consistent, and tunable to match human visual response, with far fewer compromises to electronic interaction and proprietary hardware interference.

Further, the idea of the IoT is pretty fantastic as a concept. Having lights not only controlled locally, but to external data sources has some interesting implications. From daylight following to being able to send someone a message that includes a “lighting” message is intriguing. To have a web site not only present a video on-screen, but control the room lights in response to the presentation at various points has real potential for video conferencing, especially when tied to white light tuning – HA, HA! I just added yet another hot button topic to this discussion.

I’ll dump a couple more hot topics into this. The discussion of AC driven LEDs and the entire flicker issue (hot topic alert), is essentially eliminated with a DC infrastructure. The issue of lumen depreciated luminaire life ratings ends, as we could adopt one of my favorite concepts – Lumen Priority, where luminaires deliver steady state illuminance over their lifetime, with reactive control to modulate current to overcome lumen depreciation of sources and dirt accumulation. The IoT would create an opportunity for this type of control to be globally monitored, allowing real-time collection of light loss data for all to utilize in future product development.

Of all the advantages of a PoE foundation to lighting that intrigues and interests me most, is that it could finally end the mish-mash of proprietary controls, generic controls, 0-10V, Ecosystem, DALI, DMX, Zigbee, Enocean, Triac, MOSFET, leading edge, trailing edge, 1%, 10%, 20%, etc… etc… that makes creation of product and lighting system functionality a nightmare. One infrastructure, one controls foundation, all products connected to remote current control power, portable and building mounted, landscape to roadway…. end of story. I am IN!

Problems that Can’t be Ignored

First and foremost, distribution of low voltage power to lighting products is a bad idea. Edison lost the DC battle, because DC stinks in distributed power systems. Tesla and the AC power grid is efficient and can support long distribution distances with minimal losses. Anyone who believes any different needs to revisit the library on this topic. Voltage drop cannot be ignored, as it is a power robbing parasite. For example, a 24W luminaire connected @120VAC, 75 feet from its power source over #12 AWG wire, experiences only .09V drop (.07%), adding up to .02W load, or .08% power loss. Even if the driver losses 5%, the total is just 5.08% total per luminaire. That same 24W luminaire tied to a DC power supply, 75 Feet from its power source, over #18AWG wire, experiences 1.53V drop, (6.83%), adding up to 1.53W load, or 6.38% power loss. If the driver at the head end of that circuit is 98% efficient, the total loss to feed that luminaire is up to 8.38%… not exactly a loss anyone is going to embrace. To eliminate the cable loss, to get it back to the 120VAC level, would mean running #6AWG to that DC fixture… not exactly a savings in distributed power hardware, nor practical in any sense of the word. Oh yes, lets not forget that the existing infrastructure of commercial lighting is not 120VAC, it is 277VAC, where the voltage drop, number if circuits involved, etc… are less than half what most of the marketing materials for PoE show as examples and cost analysis. Marketers doing what marketers do I guess.

So, really, the idea of true DC is really dead before it is even born. However, all is not lost. DC LEDs are not actually DC, are they? DC power is usually attained using switching power supplies which can deliver AC at a frequency of >20KHz. Now, the issue of DC (which is not DC at all) simply evaporates. In the aforementioned 24W luminaire, at 24VDC, over 18AWG wire, operating at 20KHz, voltage drop is now only .31 (1.29%), adding up to just .06W, or 1.2% loss. That’s not hard to absorb in the grand scheme of things, so let’s just say forget the DC thing, and look to high frequency AC circuits, since LEDs really run just fine in this type of system – perhaps better that straight DC in many cases (another topic). This is also perfectly compatible with any Ethernet system concept, since the design of that entire infrastructure is around digital data frequencies at higher speeds than that. It also makes use of PWM or PAM control of current delivered to the luminaire simpler still.

Next, let’s talk hard wiring. If anyone believes that the ideal solution to wiring lighting is to had it over to the IT, low voltage cable tossers, they need to get on their coveralls and pop a few ceiling tiles. I have had many experiences with the absolute  mess these folks think is acceptable cabling practice. Look at the spider web of cables, unsupported, running this way and that over ceiling tiles and through walls. It is a disaster now. In fact, it is so bad, that when a computer drop, phone cable, or camera line fails, they just abandon it in place, throw another cable through the plenum and down the wall, and call it a day. Now, add a run for every 24W of connected lighting load, router boxes and hubs tossed around among for good measure, redundant cabling to cover failures and miss-wiring… it is going to be an absolute, without a doubt nightmare. This is also not going to be simply accepted by code authorities. About the time the PoE revolution is beginning to heat up, there are going to be meetings and codes re-written to resolve the emerging and expanding mess of unprotected, un-supported, and disorganized wires running this way and that. There are cities everywhere that demand low voltage cables be run in pipe, for a reason. This will apply to lighting run on PoE networks as much as it does fire alarms, clocks, camera, and data today, so claims that this is not an issue, are simply incorrect. The savings of running wire in pipe for a low voltage system over line voltage are only in the cost of the wire, and perhaps the pipe size – the labor and hardware remain the same.

About installer expertise. In virtually every instance I have been party to related to LED product failures in the field, it has come down to installer errors more than any other cause. Everything from wiring line to low voltage connections, cutting off connectors and miss-wiring controls (power to 0-10V control input), arcing from live wire connection, etc… While PoE appears to resolve some of these, as the connectors preclude anyone goofing up the circuits, the fact remains, almost all of the cabling that will be run will be made from raw cable, with terminations made in the field. Than means there are more than ample opportunities for bad cable connections, broken wires, broken connectors, cut cables, etc… The idea that the IT players already experienced in this type of connection will take over from union labor paid electricians is, well absurd. Take it from someone who has had more than a few run ins with the brothers, they are not going to just let go of the work and accept this new development without being heard. How they will be heard is through the local inspectors, who will be encouraged to make code adjustments to keep the peace. It has happened before… look at the example of how EMT pipe came to being, to take the pipe work away from the plumbers unions. While I am sure there are pundits who will insist that those days are over, that the contractors have less power now then they did back “in those days”, I suggest they revisit the transformation of the market from  design/contract to design build, to see that contractors have more power today than ever to influence what is and what is not acceptable in their universe.

More on the installer base. Forgetting issues of control, the real issue of moving toward PoE will come down to installers qualified to install these new systems. There is a lot of training to be done, information to be distributed, and buy-ins to be obtained between now and full implementation. I am seeing classes, certifications, and other programs to deploy this in a way we can all depend on. None of this is even on the agenda from what I am seeing, but the concept is also far from settled and in full swing… so we’ll see.

The issue of disruption and the perceived need for it

Do business owners perceive of a need to see a massive disruption to the way lighting is connected and distributed today? No they don’t. You can show them all the wonderful presentations about energy saved (most don’t care a great deal), you can espouse the wonders of flexibility (they don’t actually understand how things work now, so this has not point of reference to work from), and you can attempt to excite them with promises of great new capabilities (they are not demanding)… and the end result will be no change at all.

There will be a few example case studies of large profile projects displayed, usually enjoying their slice of the new frontier at deeply discounted prices just be used as window dressing. That is not a revolution. The revolution comes when the customer is demanding and absorbing the new thing at a rate greater than those producing it can keep up with. That requires widespread active interest, which is founded on a general consensus of perceived need. In a market that just barely accepts that there is a need (mostly to comply with the law) to change from T12 lamps to higher efficiency bulbs, who see LEDs only in the vein of Edison socket replacement lamps, who are not absorbing other much simpler, easy to apply controls products today… it is hard to believe there is any ground swell of unsatisfied demand pointing customers to the doors of those offering PoE as the next big thing in light.

For new construction, on projects that involve Net Zero (another hot button!) concepts, or LEED, or publicity related something or another, there will be opportunities for PoE to show its stuff and demonstrate its coolness. How much this inspires the rest of the market to get in line, vs. eliciting a blank stare and a yawn, will have to be seen. I can see it going either way. I personally like the concept from a design vantage point, but am skeptical it will ever really catch fire and make a big difference overall. In the end, I am afraid it will just become one more controls system in an overly messy, over-populated universe of controls approaches. Soon enough, there will be another hot button concept proposed to get excited about, putting us all further and further away from the solutions we really need… and the cycle will start once again.

I truly hope I am wrong.


Leave a Reply