A follow-up on the DC-powered home

-January 30, 2013

Late last year I posed the question, “How do we get to a DC-powered home?”  Many of you responded and I really appreciate it. I was impressed by the range of responses. They ranged from “I have a DC-powered home,” to “it makes sense for lighting and other small home appliances,” and, finally, “I just can’t make this work.”

I mentioned in the first article that much of our home appliances are DC-powered and that we have many AC-to-DC converters because of it. Many of you acknowledged the ever-increasing number of AC-to-DC converters. However, the question remains whether we continue to do what we are doing now, or look for a point where we can start the change for a different future.

There certainly are challenges or problems to be addressed on the road to a DC-powered home. Table 1 lists some of the ones discussed in the responses to my original question. The one most often mentioned problem goes all the way back to Edison’s time, DC disconnects. How do you disconnect without drawing an arc? How do you deal with a partial disconnect when an arc is being sustained.


Table 1. Issues for the DC-powered home


The problems

The solutions?

DC disconnect

Special plugs and sockets

Multiple low-voltage DC

POL converters, multiple distribution, distribute AC and DC

Low-voltage distribution

Larger wire, distribute AC and DC

Retrofitting older homes

Limit to practical use

DC power expansion

Modular converters

High-power DC

Specialized connections and breakers


P. K. Sen presented a very good overview of the arcing phenomena from both a historical and present day perspective.1 Arcing voltage is heavily dependent on the current. In reference 1, much of the presentation addresses very high current. However, there are a few graphs that show arc voltage at one amp of current. The arc voltage for the one amp range is shown for a minimum gap of 5 mm. The author states that at very low current, arc voltages and resistance are hard to predict.

My only experience with very low-current arc is at high voltages. In my much younger years, I would make Jacob ladders with coat hangers and a neon sign transformer. The transformer had a 15 kV and 30 mA output. The gap at the base was about a quarter of an inch and the top gap was about two inches. The spark would travel the full ladder. I learned a lot about arcing by watching this device. I learned even more when I accidently made contact with one of the wires and my heart tried to sync up to 60 Hz! 

Probably, the main problem for the low-voltage DC is a lose connection that could have a very small gap where the gap resistance could maintain the current, and the load voltage high enough to keep the appliance working. Eventually, this may cause a thermal problem. At a much higher voltage like 400V DC, a traditional arc problem will occur. NTT and Fijitsu announced in 2010 that they developed a plug-and-socket set for high-voltage DC distribution aimed at DC-powered data centers. This technology uses a high-density magnetic field and an electric switch to guarantee safety.

Another point raised is the variety of different low-voltage DC requirements.  There were several different suggestions. One method is to create a more finite number of low-voltage rails as standards, then distribute them at each socket. Another approach is to distribute a very small set of DC rails along with an AC rail. The last suggestion distributes only one very common rail and has another DC point-of-load (PoL) converter do the final conversion. None of these sound all that attractive. Maybe a compromise is to have one DC rail with one AC rail at each socket. 

Distributing low-voltage DC is a problem due to the line loss, particularly at the higher current level. Having to increase the gauge of the wire is not attractive due to the cost. Having a combination, AC and DC, as mentioned above may be a solution. Perhaps choosing a higher DC voltage, 24V+, to help reduce the DC line current, would be beneficial. 

Retrofitting older homes would be a challenge. It may be that the DC-powered home would be a DC-powered room. Another approach might be to just address the high voltage appliances in the older homes. This could have the greatest payback for the level of investment. 

Another problem is the need for expanding the number of appliances or their loads. Having a central low-voltage DC or high-voltage DC source would require that it be sized to the future needs. If this is done, then efficiencies would suffer while operating at the initial lower power level. An alternative would be to have a very modular solution so the units could load-share and shed in accordance to the power needs. 

The breaker system would need to be changed to work with high-voltage DC and high current. These systems are being developed for data centers, so the technology could transfer to the home. It just may be that the combination of data centers and homes could be enough volume to reduce the cost of implementation.  

The more I write about it, the more work seems to be needed to get to a DC-powered home. We shouldn’t forget the idea just due to the level of effort. If the benefit outweighs the cost, then it can be worth doing. Many of you responded with the suggestion to focus first on the low-hanging fruit, like lighting power and low-power stationary appliances, like home entertainment. Several suggested that by doing this, creating an uninterruptable power source (UPS) for these systems could be simple.

Please let me know of any other thoughts or opinions you have on the DC-powered home. We may slowly get there, but we have to start somewhere first. 

For more information about this and other power topics, visit TI’s Power House blog: www.ti.com/powerhouse-ca.


  1. P.K. Sen, “Understanding DC Arc Phenomena & Incident Energy Calculations,” Department of Energy 2008 Electrical Safety Workshop Golden, Colorado, August 1, 2008.
  2. NTT FACILITIES, INC, FUJITSU COMPONENT LIMITED. “Development of DC plugs and socket outlets for smart grid and data center,” November 9, 2010.

Loading comments...

Write a Comment

To comment please Log In