Ask Hackaday: Who Wants An All DC House?

logo if acdc band

Sometimes when working on a righteous hack, we get goosebumps while watching our code execute faster than we could ever possibly comprehend. Seeing the pixels of the LCD come alive, hearing the chatter of relays and the hum of fans…it’s an amazing thing what electricity can do. And it is equally amazing when you realize that it all started one hundred and thirty five years ago, when [Thomas Edison] changed the world forever with the first practical electric light bulb.

That bulb was lit by a Direct Current – the same thing that runs the computer you’re reading this article on. The same thing that runs many of the hacks you read about here on Hack a Day, and almost all electronic devices in your house. But somewhere in the mix must exist a device that changes the Alternating Current from your wall outlet to the needed DC. Why? Why is it that we transport electricity as AC only to convert it to DC in our homes? You might answer:

“This argument was played out in the War of Currents back in the 1880’s.”

Indeed, it was. But that was a long time ago. Technology has changed. Changed so much to the point that the arguments in the War of Currents might no longer be valid. Join us after the break, where we rehash these arguments, and explore the feasibility of an all DC environment.

a cluster of wall warts

Let’s see…110 AC in, 5V DC out, 1000ma…this should work. Quick check with the meter to make sure it’s actually 5V and not 50 and you’re up and running. Each and every one of us has done this at some point in our lives. But why do we have to? Is there any reason we can’t have DC outlets? We’ve seen USB ports built into outlets while strolling the isles of our favorite hardware stores, but most are unlikely to be switch mode supplies.

You would still need AC for kitchen appliances and such. But consider changing these over to DC. Imagine a house where everything ran on DC!

Let’s take it further and imagine running DC from the power station to your house. This brings us back to the War of Currents. We all know that it’s relatively easy to step AC voltage up and down. You just need a transformer. But it’s not that easy with DC, so running DC over long distances is just not practical. Indeed, this was true in the late 1800’s. But is it true today? The technology exists to step up DC to higher voltages. But can we do so efficiently?

Lots of questions still remain. Be sure to sound off in the comments on the idea of an all DC house. Good idea? Or not so good idea?

268 thoughts on “Ask Hackaday: Who Wants An All DC House?

    1. I think that’s the only way it will be practical. As the article said, it’s easier and more efficient (energy wise and cost wise) to step AC up and down for transmission over long distances. That same cost-effectiveness is critical to large appliances, particularly ranges, ovens, water heaters, and central heating systems.

      All that said, I don’t see why we can’t have a hybrid AC/DC house wiring setup. Wire each standard receptacle with an AC plug, a 12v DC plug, and a set of 5v 2A USB plugs.

      1. What about the voltage drop? Wires are resistors. Wiring 5V in a ring setup in a room of 5×5 meter will give your different Voltages on each outlet. You would be better off wiring 24V and have a voltage reulator on each outlet.

        1. And then most of the devices you then plug into that outlet have their own regulators inside too! Maybe there could be two sockets to each outlet. One would be a regulated 5V, the other would be a direct connection to the unregulated “24” volt line. Devices with their own regulators could just plug directly into the unregulated line eliminating the wall wart.

        2. >You would be better off wiring 24V and have a voltage reulator on each outlet.

          Excellent point. And there are some devices (laptops for example) that would require more than 12 volts; I didn’t think about that.

          1. I don’t understand what this gets you. Now instead of a single AC to DC switcher for each device, you want a bulk AC to DC switcher followed by a separate DC to DC switcher for each device. What do you gain?

        3. As long as you start with a higher voltage (maybe 48 volts). Since there is almost zero standardization in the electronics industry about what DC voltage devices should be designed to run on, you’re going to need a wall of various voltage level plugs in every room anyway. This is, in reality, what kills this idea straight out of the box.

          1. Not sure if we can wire the solar cells in a way so that it provides a wide range of voltages, like a battery in series and then by tapping into terminals in the middle to get lower range of voltages.
            But the power delivered will reduced

          2. Im thinking a each desk would just use a repurposed atx psu it would have most of the voltages needed and can easily be hooked up to a voltage regulator for any weird ones. If efficiency is key then maybe a newer more efficient one… my 750w psu can do upwards of 60 amps on its 12v rail so maybe just buy a decent computer psu for your computer and run your stuff off that…

          3. I personally think that the BIGGEST problem would be the need for a standard connector for your 48-volt DC. Also, it should use the AC ground, and if you try it you’ll soon be teaching consumers about ground-loops and other forms of voltage offset.

            Still, switching-PSU designs answer [Nickson]’s concern, and [jadoo]’s actually makes this seem practical (caveat: I would probably install the ATX PSU into the ceiling in a large construction wiring box, but that’s just me). Of couse, [jadoo]’s also points to the correct way to do USB outlets…

        4. Your scheme is how I approach larger electronics projects today. I distribute a higher voltage, then regulate it for each sub circuit. I learned a long time ago that one regulated bus just doesn’t cut it. One regulated bus may seem like the simplest, and most elegant solution, but in practice, well it is about as realistic as trying to juggle so many balls at once. Sure, with an amazing amount of effort it is possible, but there are easier ways of putting X number of balls where you want them all to be. Or voltages as the case may be.

          But supplying more voltage, and regulating off that bus down to what each circuit needs is a relatively trivial matter. Professionally designed systems even employ the same technique. That is why the 12V rail on a modern PC is so loaded down today. Back in the Dark Ages it used to be the 5V rail that took the load. Components used their current from it directly too. But it is simply asking too much of a PSU to regulate everything tightly all of the time in a complex system with a variable load. Well, too much if you expect the regulation to remain precise.

      2. One reason DC was dropped because of arcing problems. Go look at a switch, note the different current rating for AC & DC.

        DC will sustain an arc better than AC, and that causes fires. I read a story about a fire caused by a broken light bulb, and the arc went up the wire into the roof (back in the days of 100VDC mains & cloth cover wires, etc).

        Things have improved, but probably too entrenched to change.

        On the other hand, I look forward to each country having it’s own style of DC plug & socket. Progress!

          1. @Lwatcdr – good idea, incorporate a highspeed data connection with the DC supply. Leave the AC sockets anything > ~25W or so.
            Where I live electrical rules state you cannot have any other sockets in the same faceplate as an AC socket.

          2. @bootstrap some network equipment manufacturers already have not-completely-802.3af/at compliant PoE PSUs and PDs that work with 50W. They use all four pairs for power, while the IEEE only allows two pairs to be used.

            Problem is, that’s enough for most small appliances, like phones, small radios, alarm clocks etc. Maybe even some tiny energy efficient notebooks. But most consumer hardware that requires a wired ethernet connection uses more power than PoE can supply.

            But on the other hand…it’s time for a new network plug standard anyways, rj45 is cool and all, but the whole thing with the easily breakable locking tabs is annoying. SO why not just make a cable that also incorporates a pair of wires for power?

            Hm, i could refire my research on that mad idea i had a year back or so.

      3. A transformer is conceptually easy. Whether that means it is still cheaper than a “DCDC converter of the same efficiency” remains to be seen. At least for transporting large amounts of energy over long distances, it has become cheaper to do so using DC. There is a High Voltage DC link between norway and The Netherlands.

        1. There is a break even point for cost differences between AC vs DC transmission. I believe it is around 400 miles. If you need to go further than 400 miles it is more cost effective to do so with DC transmission.

    2. I think running lights (especially LED) and built-in USB ports around (plus whatever else) off a dedicated, solar/ battery driven 24v DC system would be cool. I’m not an EE, but I feel as though the drop from 24v DC to 5v DC is going to have a higher efficiency than 120v AC to 5v DC. Maybe create a new type of converter for laptops (which usually run around 19.5v.) A 12v DC laptop power adapter is much smaller than a 120v AC one!

      1. I understand transformers to be fairly efficient at voltage-level conversions, it’s just that 50-60 Hz transformers are physically big, and therefor both heavy, and financially expensive. Higher frequency transformers are better on those, but apparently the higher the frequency, the larger the difficulties (also, the higher the frequency, the shorter the distance between maximum-power nodes and zero-power nodes: 60 Hz apparently puts the zero-power nodes somewhere out over the oceans).

        Incidentally, the highest-efficiency DC converters actually create an internal AC signal: BUT it’s at a high frequency, so the inductors can be small. You could probably build them to run off of line power too, but you might have to greatly increase the EMI shielding for the powerline: high-frequency noise absolutely kills generators (as Tesla accidentally demonstrated).

    3. Just to leave a comment near to the top so people don’t waste their time. DC transmission lines and DC lines in a house, is more than stupid, its down right wtf? AC travels along the outside of a wire while DC occupys the intire space, this allows High-Voltage AC lines to be stranded wire, significantly increasing the surface area without increase the diameter. On top of the fact that a DC line at 1kW would 6 times or more larger than AC, be way harder to control, almost impossible to step up DC to the 400kV required, and no way to convert easily to other voltages, or isolate from the mains. This is downright stupid, who the hell thinks this is good?

      1. I think theres a conceptual problem here. AC doesnt travel on the “outside” of the wire, it travels close to the surface (but still inside). It’s actually a bit more complicated than this but that doesn’t matter so much. What this means is that you have to have more complicated wire (stranded) to try to get the AC to use all the copper in the wire, and even then its not perfect. DC uses all the copper in any wire, so it can just be a nice efficient solid wire. Stepping DC up to 400KV is not super difficult, and im not sure why you think the wire would need to be 6 times larger, nor why DC is “harder to control”. Could you please elaborate? Usually if a lot of smart people have spent a lot of money on something (like the ~20 HVDC links in europe), it means that they’ve thought about it for at least 10 minutes, and probably have already thought of all the issues (or non-issues, in this case) that a random internet commenter with very little actual knowledge would bring up.

  1. My Sister once had a 100% DC House. All the lights and stuff were DC. Then she grew older and she didnt liked to play with her dollhouse anymore. My Dad never build a DC House again.

  2. In fact, for long distance high power transfers, we use direct current by today. However, for smaller devices, such as transformators that run smaller house blocks, efficiency of DC-DC conversion is so much lower than that of simple transformers that the there can be no efficiency gain. Also, you have to realize that any DC-DC converter (ok, these milliwatt chopping inductorless aside) has to have significant amounts of coils — up to the point where you wonder if saving a bit of transformator wire is worth of introducing possibly failing semiconductor-based converters.

    This is an article that the market has already an answer to: No. We still use AC because it’s practical. Your PCs, your iPhones, and all of your consumer devices that rely on DC aren’t making as big of an impact to justify having a DC network. Your stove, your refridgerator, your air conditioner, your electrical heaters, your lamps, electric trains all run as efficiently (if not more efficiently) off AC — so stick with that.

        1. You would still want to have the rectifiers in the devices even if the house has a DC bus.
          e.g reverse polarity protection and to prevent the charged caps inside the device to give you electrical shock if you were to touch the plug after disconnecting it.

      1. A brushless DC motor is a alternating (no, usually even a three-phase) motor with an internal “alternator circuit”. Your argument is sadly the opposite of what you intended it to be :)

  3. I live off grid. I have a battery bank. I could have a ‘DC’ house if I chose. Why don’t I?
    1) I like to be able to switch stuff on and off.
    2) I like to be able to buy household appliances from normal shops
    3) I like to minimise shock risks

    Having said that, there is room for compromise. A small number of essential/emergency items – communications equipment, for example, which is powered 24/7 is marginally more efficient powered direct from the battery bank – and are required even in the event inverter failure. I have a small number of low voltage outlets to support these, but if I lived somewhere ‘normal’, or on-grid, even that would be overkill. Switchmode supplies are very efficient these days…

    1. I don’t understand 1 and 3. DC is switchable. You just need a bigger switch for the same voltage than you do for AC, but presumably a DC installation would be lower voltage anyway. DC is safer for shocks than AC, because AC triggers muscle contractions so that you can’t let go (and again, the DC would presumably be at a lower voltage).

      1. Which is really good, because the only apparently standard DC jack & plug system (the common 12-volt car cigarette lighter socket) is physically easy to short. It takes a little thought to accidentally short out an AC socket during everyday life (spraying water all over your walls is NOT an everyday event), but the common DC jack? A baby could do that easy.

        tl;dr: DC jacks are dangerous.

        1. Anderson powerpoles have been the standard for dc connectors in ham radio for some time now. They won’t short out or wear out easily and are resistant to reverse wiring.

      2. you dont understand arcs…
        the DC supplied open circuit voltage does NOT matter!
        the cables in your house form an inductor!

        when you switch off a switch the inductor(in-wall-cabling) causes a HUGE voltage at the switch, arcing accross it!

        it does not matter if your voltage is 10v or 100v … THE ARC WILL STILL BE OVER ONE THOUSAND VOLTS!!!

        check out what happens when you use a 50A/120vAC switch to shut off a DC motor pulling ONLY 25 amps at 12v … the motor does NOT stop and THE METAL CONTACTS OF THE SWITCH GET SO HOT THEY MELT THE SWITCH HOUSING!

        PS: in this situation, the switch wont get hot while it is on (25A out of 50A), assuming you have not yet tried to switch it off yet.

        1. A DC switch would have to have a fast acting zener diode, that’s all. Then only the short-circuit arc is the only problem. Well, in what you mentioned. It’s true DC has problems but your post seems alarmist for a reality that wouldn’t exist if you had properly wired a DC house (not using AC switches for DC inductive loads.) But then of course there would be a cost associated with the DC devices, and sadly you’d have to pay it for all switches assuming they *might* be used for inductive loads.

  4. Heat guns would get smaller, as the amount of heat produced by the nihrome wire is determined by ohms law. So with DC you could make them 1/10th the size for the same heat output. Besides charging batteries, I’m sticking with AC.

          1. You’re both right. Power transmission lines need the added strength of a steel core, AND take advantage of the fact that most of the current flows in the outer portion of the conductors. Plain copper or aluminum would break under the tension and plain steel would melt in the heat it would produce.

          2. Steel core is for strength, but using steel on the outside or a cable made entirely out of steel would be bad. The skin depth (where 68% of current flows) for copper wire at 60Hz is around 8mm. Skin depth is a function of many things one of them being magnetic permeability of the material. Ferrous materials have a huge magnetic permeability (that’s one of the reasons they are good cores for electromagnets) therefore the skin depth is much much smaller hence the effective resistance of the conductor goes up. That’s the beauty of ACSR practically no current flows in the steel core. (I’m an EE and used to work for a power co-op)

      1. no, it’s due to an assumption I made. I assumed the DC in the house would be 12V (which is really stupid for anything with more than 20 amps and more than 10 feet from the power source)

        Nichrome is linearly resistive, the 120vac heat guns have a good length of it, normally many feet worth of a 26-32guage wire, wound up in a coil still several feet long, in order to make the proper amount of resistence for the right amount of heat generation. in a 12v DC house, assumed, it would be 1/10th the amount of said nichrome wire, about 6 inches, which can be wound into a 1/2″ coil, to produce the same heat (same wattage)

        1. Did you ever wonder how they could make 100-watt light incandescent bulbs in 240V, 120V and even 12V versions? The answer is, different thicknesses of filament wire. The factor that limits how small you can make a heater is how fast you can transfer the heat to its desired task, not how much wire it takes to dissipate the required power.

    1. ” So with DC you could make them 1/10th the size for the same heat output ”
      Care to elaborate on the math ?

      The amount of heat generated by a nichrome wire is a combination of current and resistance, hence ” P = R x I² ” .
      Lowering the voltage as no influence over the system.
      Want a smaller heat gun ? Less resistive nichrome wire that uses more current.

    2. That doesn’t make sense, does it? The AC voltage supplied to our homes is measured in RMS, which is equivalent to DC voltage for such calculations. If you use Ohm’s law to calculate the current which is pushed by a known AC or DC voltage through a given resistance of nichrome wire, It will be the same. The heat generated will be proportional to that current, and it should also be the same. I don’t know where you could have gotten the “1/10 of the size” idea.

  5. The reason for AC still persists and remains valid. Long distance transport is nor practical or more efficient if done on DC and as previously stated, introduces potential points of failure totally avoidable with AC.
    So, no, do not convert my hole house to DC.
    Now, if you say we should have some sort of DC sockets as well in the house (not in transport) in order to have a single AC-DC conversion solution for all DC equipment we use, therefore reducing conversion losses of having to convert in every PSU, I would agree.
    Compromise usually yields the best results, and in this, for the moment it does.

    1. “Now, if you say we should have some sort of DC sockets as well in the house…”

      I’m picturing a big high-wattage 5V power supply in the furnace room and USB jacks in ever wall. It sounds good but I’m not sure it would work. What would the voltage drop be between rooms?

      1. R=r(L/A) where r is resistivity, L is length, A is the cross sectional area. Copper’s resistivity is 1.68E-8 ohm-meters. Figure out how fat a wire or bus bar you’re willing to buy and the total round trip length for +5V and ground, and solve.

        1. Why? Ohm’s law is still the same with AC as it is with DC. Mains lines have lower loss because they are higher voltage, not because they are AC. AC comes in to play because those high voltages need to be stepped down for use. This is easier to do with AC.

          1. Ohms Law is NOT the same for DC and AC calculations actually… It is close enough to work between the two for the most part.

            I squared R is exactly why they don’t transmit DC as mains voltage. AC current doesn’t have to travel as far as DC current does.

            Induction is only a side benefit, and a good one at that.

            I think the guys of hack a day are more writers than engineers. And if they are engineers, they are probably writing for a living for a reason…

            Until super conductors at room temperature are a “thing” DC household current will never be a “thing” -and thats just one reason, there are hundreds of others that are just as valid!

    1. This. This is the most valid argument against a fully DC household, in the sense of a standard being introduced.

      DC is easier to store, making it the best way to go off-grid. If there was suddenly a standard for DC storage and distribution in a household, energy companies would not make money by generating your power and sending it to you. Why would the big names allow this?

      Its the reason there are some areas that enforce local power supply.. as in you are welcome to not use the state power grid, but your still gonna pay for it regardless.

  6. How about just taking the death risk out of it all kinda like Tesla wanted? That has always been my problem with all electric power in a home environment. I still wake up when I hear the cat rustling in the other room thinking she has finally drooled into the power strip she is sleeping on lol. It just seems like something that could be done but would need some retooling efforts along the way which would be costly. Anyhoo, that would be my request. Nearly lost my grandma a couple of years ago when her finger slipped trying to plug in an adapter and she touched the tine and got a nice jolt. It was enough to trigger her arrhythmia again and put her down for a couple of weeks. My grandma shouldn’t die over an answering machine (that she ironically unplugged because of a T-storm in the area). Just seems like it we have progressed a bit and like the blurb says a local conversion could remedy. But oh well. We are all electron junkies lol

    1. Well maybe if the design of sockets was good then people would have no risk of electric shock… I find it unbelievable that such flimsy sockets are used in the US & around the world. Here in the UK it is virtually impossible to accidentally touch a live prong or even stick something metal into a live socket!

        1. Little heaters? You mean the fuse? The one that saves your house from catching fire?

          If you see / hear a UK fuse blow, you’ll be glad the explosion happened in an easily-removable plug and nowhere else.

          Our plugs, also, since the 1980s, have had insulation running halfway down the prongs. It’s very difficult to touch the live metal part with your finger, while the prong is connected to the socket’s live power. So much less chance of electrocution. Any plug would benefit from that, assuming it’s big enough to support the insulation.

          The hugeness relative to other plugs is no bother. They’re a few inches square and weigh a few grams. I’m 5’10” and weigh more kilograms than I care to admit to. There’s nothing inconvenient about them for an ordinary-sized human.

          1. We shouldn’t need that different of a design, just a change to the non-ground prongs, and a modification to matching receptacles (basically, disconnect them before the insulating bit fully leaves the socket). That should cover even most incorrectly wired houses. Changing the sockets would probably take a few decades, though, and a complete prong conversion might only happen with the death of the last piece of still-active equipment (one of my great-grandmothers still had an old-looking vacuum in the 90’s).

          2. We don’t have to change ANYTHING at the residential level. What we have, WORKS. Almost nothing any more actually wants 120VAC, so practically everything ALREADY has an efficient converter to go from 120VAC to what it actually prefers.

            Guys? The problem has already been solved, and it’s called a wall-wart. All you need is a better power strip so you can plug eight of them in right next to each other.

          3. I’ve often wondered why our mains plugs aren’t just coaxial… hot center pin, neutral inner shield, grounded outer shield. Surely this would be (while a little bulky) much safer all around…

    2. Here in the UK we might have more cumbersome mains plugs than anywhere else in the world but they avoid many of your issues. Live and Neutral have covers over the socket holes which slide open when the longer earth pin is inserted, so children and animals can’t electrocute themselves. Only the ends of the Live and Neutral pins of the plugs are uninsulated making it impossible to touch a live part even when the plug is partially inserted. The cable enters the plug from the bottom so if the cord gets wet water does not run into the plug.And each plug has its own fuse so the thin appliance cord is not relying on the 32A fuse in the distribution unit to protect it from getting hot in the event of a partial short. Actually quite a clever design for such an everyday object!

      1. Do you use 110 or 220 in the UK? Ive never seen issues with the US plug design, I think to get a serious shock you’d have to have a paper clip in each hand and insert both. The biggest danger for animals is wire chewing. Plus 110 isn’t super dangerous in most sutuations, 110 across you finger isn’t even likely to burn you unless you were stuck there for some reason.
        220 on the other hand I think does justify a much more safety minded design.

        1. 240v in the UK. Of the very few shocks i have had, the don’t hurt much at all compared to the reflex action of pulling your name or arm away and banging your elbow on a table, wall, radiator etc. UK plugs really are good.
          Having lived in a house, in KSA, with both 110 and 220, there is a definite need to differentiate between supplys. :)

          1. Yeah, 230/240VAC leaves you very alive, shocked so to say, and sometimes with bleeding arm after banging it to something sharp. I think all modern European plug standards are good enough to prevent accidental touches as most of them have insulated prongs, child locks, deep concaved sockets and what not.

          2. My experience and those of my colleagues, yes a very fuffy statement lacking hard science and statistics, indicates 240 volt shocks are not often fatal. Often implies is also an imprecise term.

            Voltage doesn’t kill, current does and so does stupidity. I am happy working on this an higher voltages, but I take precautions and care as I work. Turning off the circuit is not always an option.

            The advantage of AC compared to DC is the reflex action is to flinch away from and not grabbing to form part of the circuit.

      2. ” making it impossible to touch a live part”
        Hah, impossible is such an absolute term. Never underestimate the power of stupid! I’m sure someone has found a way.

        I wonder what a UK socket costs relative to a US one…

        1. Its way more fun when you can touch the prongs anyway… especially if you have a quarter or a screwdriver to do it with… =p

          If you’ve never been privileged enough to see this, I cannot recommend you try this without EMS standing by… =p

          When I was much younger, we had alot of fun dropping metalic items between the prongs, which usually ends up singing the wall socket and whatever plug you are using, but its kind-of fun to see enough electricity arc through the metal to melt it… =D
          This would often enough trip the circuit breaker though… =p

          Electricity is dangerous kids! Whatever your friends say, just say no! =p

      3. Current US code requires that all 15A sockets have shuttered neutral and line. It also requires arc-fault interrupters on most circuits (usually all outlets are included). Everything that plugs in is supposed to be UL certified, which means it was tested to fail-safe in extreme circumstances, either by internal fuse or otherwise.

        That doesn’t prevent some idiot from sticking their finger on a partially inserted plug. But the average 120VAC shock is brisk but not deadly.

      4. Shutters on the live and neutral prongs (and maybe ground as well? I forget) actually exist here in the US as well (I’ve seen and played with them in person), they just aren’t required, so they usually aren’t used by construction contractors. Most (or maybe all) of the other parts that you mentioned are the same: we don’t have that regulated, so it simply isn’t STANDARD.

        1. They are absolutely required in the US for new builds. They’re in the code now. So are AFI’s, though they’re a royal PITA (time to go out to the breaker box again to turn the living room back on…).

    3. I don’t know where you live, but in Europe we have a kind of power outlet that protects from what you just mentioned. Safe for toddlers, grannys and also cats ;)
      So AC is not the culprit you think it is.

  7. This article should have had more in it. A flame war about this topic was set off the other week in the comments about the google prize for a compact inverter, and the “discussion” lasted for pages. All this article did was skim the slightest surface of the issue and invite a repeat of the same comments from the other thread. When you say, “Join us after the break, where we rehash these arguments, and explore the feasibility of an all DC environment.” You should actually do that and not just basically repeat what you said before the break. It should at least be longer than before the break.

  8. “We’ve seen USB ports built into outlets while strolling the isles of our favorite hardware stores, but most are unlikely to be switch mode supplies.”

    Seriously, who writes this stuff?
    Of course they are switch mode otherwise they wouldn’t fit behind the outlet and cost a fortune to ship around the world due to the huge wight of a transformer.

    1. “Seriously, who writes this stuff?”

      Someone young enough never to have seen 5V, 2A done with traditional components :)

      I have an eastern europe sinclair spectrum clone with a 5v 1.5A psu built with communist era components literally the size of a brick (probably size comes also from efforts done to have it as stable as possible – to the day it still outputs 5.05V under load)

        1. Hey they all have their place –

          SMPSU –
          Compact
          Cheep
          Low thermal output under load
          Light weight

          Linear –
          Goes forever and never needs fixing except the occasional fuse replacement
          Practically immune to spikes that kill SMPSU’s

          I am building a soldering station and it will be a linear supply so that I know it will always work.

          1. I understand that you can build high-power SMPS. That having been said, I’ve ALSO read that some of the BEST supplies actually use an SMPS to provide the raw power for a Linear stage.

            Personally though, if I ever finish building my resistance soldering station, I don’t plan to have any more regulation than a transformer, fuse, resistors, and maybe some diodes, and a few capacitors (I’ll see if it survives it’s first test run DURING the first test run ;) ). Somethings just don’t call for high accuracy.

    2. “We’ve seen USB ports built into outlets while strolling the isles of our favorite hardware stores, but most are unlikely to be switch mode supplies.”

      Seriously, who writes this stuff?

      People who can’t spell “aisles”.

  9. An interesting question! Some commenters have said that long distance DC transsimission is not efficient but actually it is at high voltages and is being used more and more for long distance high power links. http://en.wikipedia.org/wiki/High-voltage_direct_current.
    So it’s better on the high power transmission and it’s better for some things in the home but converting between the 2 is not easy. And anyway, what voltage would you wire your house at? 5V usb? 17.5V or what ever your current laptop runs on?

    1. Just an additional note, bulk DC power transfer is usually used when going between “zones” in the grid. Inside of one AC connected zone, all the generators need to be synchronized to very tight tolerances so that one generator doesn’t back-feed another. Converting to DC and then back to AC on the other side allows power to be fed from one “zone” of the grid to another without needing to maintain a phase lock which gets increasingly difficult with the geographical distance between the generators and variations in load across a large region. HVDC over long distances suffers because there aren’t things like reasonable auto-transformers to boost sagging voltage on long runs. That said, if you want to ship power from one region to another with minimum fuss, HVDC is fantastic. (I used to work for a power co-op)

        1. First, AC transmission systems don’t have ground capacitance issues. They’re transmission lines, which means that as long as the load is matched, they act purely resistive. In fact, virtually all AC transmission systems have to ADD capacitive loading in order to get the power factor close to 1. Second, DC transmission systems DO have capacitance to deal with – it affects the transient response of the system. No real-world system is purely DC.

          I’m not saying HVDC transmission isn’t a great thing – it’s probably saved us from many a widespread power failure here in the U.S.. I’m just saying that just because it’s “DC” doesn’t mean it’s simple.

          1. Good reply and thanks for clarifying :). Do you know how in HVDC systems loads are matched? Or how big is the issue dealing with transient response? Just asking if I could learn something.

            You’re right. There is always some inductance and capacitance in real-world – no matter how “DC” the system is.

      1. +1 all the way- Edison was a total douche. Ever heard of Joseph Swan? Edison’s lab did come up with some interesting things, but the “Edison base” screw-in bulbs were an idea of one of his lab assistants. And Edison insisted on carbon filaments, the tungsten filament was another lab assistant. Tesla was the man.

  10. I, like Will Richey, was expecting a nicely written, argumentative or persuasive article when I read “Join us after the break, where we rehash these arguments…” but all I read was a poor attempt to get the readers to say exactly what they said last week over that Google prize inverter article. I am disappointed.

    @gravatarnonsense: I agree, that statement was also somewhat rubbish.

    Try harder next time, Hackaday.

  11. dc in the whole house is a terrible idea, my uncle had one due to a dc generator in thje basement, and the powerlines needed way thicker dimensions since they heat up more.
    and you had to use an inverter to power things that need ac

    1. I like the idea of building a “tiny house” and have it run (mostly) on Wind/Solar, but then I’d have to get rid of 95% of my “stuff”.(inre George Carlin) If I could just convince SWMBO (She Who Must Be Obeyed) to let me build one in our back yard.

      1. Interestingly enough, downsizing/minimalizing goes hand in hand with living in a tiny house, and extends to the digital life as well. My peak power usage is starting up an electric kettle at 1300 watts; other than the kettle, my peak power usage hits about 75-80 watts, and averages at 22 watts. This includes the LED lighting in my tiny house. Living entirely off-grid on solar/wind is entirely possible, but now that I’m there, the ‘desire’ for many devices, and the ‘need’ for high power just isn’t there anymore, and diminishes over time.

    1. To paraphrase the wikipedia article, the situations where it has advantages:
      -undersea/underground cables, where the high capacitance and dielectric losses of the cable generate greate losses when used with AC power
      -really *long* high voltage above ground systems, where with DC you only use two wires instead of three wire (three phase AC) for the transmission line and the extra savings per km compensate for the cost of inverter/rectifier stations at the ends
      -transferring power between different power grids that are not synchronized, an example is Japan where half of the country is 50Hz and half is 60Hz, but even same frequency grids can be unschronized.

      On a domestic scale, for powering gadgets and lights sure…. but being able to plug a space heater/washing machine/hammer drill in the same place where we charge our cellphones without having to think too much sure is too convenient to just give up.
      Maybe there is some merit to data centers with huge racks of servers, where there may be somehting to gain with economy/efficiency of scale and making one big rugged DC power supply for an entire bay of computers.

      1. This is how modern big iron works. Usually a cabinet has an array of hot swap 208VAC switchers balanced across 480VAC 3 phase, with big bus bars running to the backplanes. The bus voltage is often 48V or so to take advantage of off-the-shelf telecom power bricks for more usable rails on the blades themselves.

  12. Indeed main problem is more copper wire is needed to transfer same energy in Watts= V*A. You want V to be as high as possible as when Amps get higher you need thicker wires everywhere in your house and indeed the voltage drop is also an issue… So 220V dc might work but you’d need to step that down to 5v at the wall which might as well include a rectifier circuit to do ac -> dc as well… Doing 24v dc or 5v dc from a central point down the street up to houses is never a good idea as you’ll be spending more on copper than you’d win in buying some transformers… Now if you have solar panels then this is another debate you could use it to charge some 12v batteries and have a separate 12v circuit dc throughout the house… Then again the hassle of making new wiring in your walls is going to be more work than just doing 12v or 24v dc->220 ac at a central place and then connect it to the existing wiring system in your house. Today’s system is actually pretty much here to stay IMHO

  13. A bit of interesting education for you Americans. Don’t take it too seriously please.

    Regarding Edison, please Google Joseph Swan.
    The cross channel link (England to France) is DC

    I am old enough to actually remember the DC appliances we had as a kid before they introduced AC supply to our house. I am not as old as one would think, that was in the late 1950s.

  14. The power distribution problems are just as relevant in house as in the grid. It is impractical to run you house at a low voltage AC or DC. A standard 20A AC circuit can theoretically supply 110VRMS * 20A = 2200W. The same gauge wire circuit wired for low voltage has the same current limitations; 5VDC*20A = 100W or 12VDC*20A = 240W. This implies any solution will be using a high voltage at least 100VRMS.
    So the question we are asking is “Are transformers better than DC/DC converters?”.

      1. Because of the much higher losses in the transformer cores (core losses increase with frequency). Airlines use 400Hz power because for them, weight is more important than electrical power efficiency (the efficiency of the aircraft’s prime mover is much, MUCH more important than that of the on-board electrical systems)

  15. Wouldn’t long stretches of DC wiring slowly corrode themselves from the difference in potential?
    We manufacture so many of our electronics in China, expecting it all to work with DC would be a stretch.
    Or at least all my Chinese LED flashlights oxidized their battery contacts.

    1. Wow. That was a stretch for a xenophobic remark. Sorry to hear about your flashlights. All the US-made Maglights I’ve owned have suffered switch or seal failure within a year or two of light use. The micros, which before LEDs were popular were my favorite EDC, would run through bulbs as fast as batteries. Meanwhile my no-name DealExtreme LED lights are all working after many years except for the one that was corroded beyond repair from a leaking (Energizer) battery.

      Assuming by “we” you mean USAnians, we invented dangerous wiring standards. Ever hear of aluminum house wiring? Do you know what forms on the surface of aluminum nearly instantly when it reacts with the atmosphere?

      1. The sapphire film problem apparently goes away after the second or third regularly scheduled entire-building round of terminal tightening, the problem is that almost no one was willing to do it. That could have been fixed by professionally welding (or maybe soldering, with the right solder) on copper terminations to the aluminum wire, but no one wanted to do that either.

    2. Virtually every low-power device manufactured today will run on 90-250V, AC or DC, 50 or 60 Hz. That’s possible because everything that doesn’t produce hundreds of watts of heat uses a power supply that first converts the line voltage to DC, then uses a high-efficiency switch-mode DC-DC converter to convert that to what it needs. This has all happened in the background without needing our attention at all. So when you say the present system is “here to stay”, be advised, it’s almost gone already, because almost everything has to be designed now to operate anywhere in the world.

      The biggest problem with DC is trying to turn it OFF using mechanical switches. AC switches can be very simple because the current is interrupted 100-120 times/second, and each of these interruptions will extinguish an arc. With DC, switches have to be designed to break the arc when heavy loads are interrupted. I found this out the hard way when I tried to operate an AC iron on 120 VDC, as a test load for a big power supply. It heated up just fine, but when its thermostat tried to turn it off, the contacts turned into a single mass of metal in a matter of a few seconds.

  16. For transmission, DC is more effective. We can throw out reactance and skin effect which is where a lot of losses come from. We use DC transmission lines for cross grid ties. From a home standpoint, DC would also be a good choice. Most equipment in our home runs off of DC anyway. The problem lies in distribution. While we do have the technology, it would be much more expensive to replace transformers with DC-DC converters that could cater to the same amount of power. Secondly, as many have sated, breakers and switch are much less effective on a DC current. If an arc develops in AC, it can be easily broken when the sine wave momentarily hits zero. There is no such condition for DC, therefore arcs are harder to break. It would be interesting to see people convert their homes to DC and install one single large switching PSU to go from mains to home. However, I don’t see very many advantages in this as you’ll still need DC-DC converters at each device to convert the higher voltage necessary to be distributed in your home to the lower voltages used by most equipment.

      1. Yes! That is exactly one of the reasons they do it. It is also beneficial in these situations because of the typical long length required for cross grid ties. The magic number is around 500 miles. At this point, negated losses pay for the increase in equipment cost.

  17. Most of the people posting here seem to have no idea of the problems that come with AC power transmission over long distances, such as grid instability, lack of energy storage or buffering, high losses for underground lines, etc. A HVDC grid has advantages, as well as the disadvantages more often cited – and HVDC definitely integrates better with wind and solar sources.

  18. My parents house has a 3KW water heater. Say we have a 24v DC supply to the house. That’s a staging 125A. I’m guessing the walls would need to be thicker to support the cable…..

  19. This is not a question of the voltage level, this must be the same, someting around 230VDC up to 400VDC.
    The question is what other difference DC would make. Nearly every device in a house runs already on DC, every switching mode power supply rectify the input voltage. One step further, these power supplys would life longer because the AC kills the capacitors after a few years. -> LED and CFL lightbulbs, computers, every usb wall plug, TVs …
    Most devices with a motor have brushed dc motors which are desinged to run on AC as well as on DC. Only very few motors relly need AC, as the ventilator in an oven or the compressor in a fridge (sometimes they already run on DC) or pumps for water (which would now need a frequency converter).
    The other devices are oven, toaster, classic lightbulbs – which only want the same voltage, 230V AC or DC.
    So I say, 90% of the devices in your house run out of the box on 230VDC and it will be more and more.
    About safety, touching DC is not as dangerous as touching AC. DC stop your heart and it will start again after a few seconds (maybe), AC will disturb your heart in a crazy way an will kill you (maybe).
    We need new main fuses and maybe ark detectors and maybe maybe isolation monitoring if everything is ungrounded. (Maybe it would be nice to isolate mains from ground, but this is another question, why not?)

    Okay, I would say everything in a house should be DC. If it goes over big distance, we need 10kV, 20kV, 110kV, 330kV …, this should be AC (for now). Maybe we make a HVDC grid in the future, or we have high-temperature superconductors, or anything else, but not now.
    In the first step we need a big rectifyer in every house or one in every town (instead of a 20kV – 400V transformer)

    And please, don’t make a 5V grid in a house, maybe 24V or better 48V to be in SELV, but this only additional. For power we need 230V, everything else is useless. Maybe a combination of both, for safety.

    1. AC will go through zero volts 100 times a second so you get a chance to let go of the wire (maybe), instead of being stuck until you fry. in the case of an arc between wires DC will extinguish itself (maybe), instead of just keep going until the house burns down

      1. You need a much greater air gap to extinguish a DC arc because the voltage does not vary. AC only requires a much smaller air gap to extinguish an arc because the voltage drops through zero twice per second as you mentioned. The difference is not as great as you expect as the air ionizes and becomes more conductive.

        Things are much different the case of humans. You don’t get 100 chances per second to ‘let go’ because it takes much longer than one fiftieth of a second for muscles to relax.

        Here’s a video that shows the effect of ionization. It is AC and you will see that arc path is much greater than the distance between the contacts as the ionized air is also heated and rises.

        http://www.youtube.com/watch?v=hIkNY5xjy5k

        1. Right, but we already have enought fuses which are able to cut DC current and some electronic for arc-detection too. But in a normal house installation we don’t have enough current that arcs would be such a problem, most standart fuses work here fine.
          And you mean “Things are NOT much different in case of humans.” You say muscles need more then 1/50th of a second to relay, right, AC or DC is the same, it’s only a thing of current, more than 50mA an you can’t let go, AC or DC. Maybe if we have 5Hz AC but not 50Hz.

  20. A more practical solution…

    I’m thinking a whole house low voltage supply would be a bad idea due to loss. A high voltage DC supply wouldn’t make much sense because in would need to be converted to a lower voltage at most of the devices. AC to DC is easier than DC to DC.

    But… What might be a great idea is to build large, multi-device power supplies to place at specific outlets. I think we probably all have certain outlets now with power strips that are filled with wall warts. Those wall warts may come in a variety of voltages but I bet many of the devices that use them have their own internal regulators and probably don’t need that specific input voltage. (do your homework and be careful with that idea) Determining the range of voltages each device can happily work with would probably show enough overlap that a couple well chosen voltages could supply everything.

    Then you can just build one power supply which outputs those voltages to replace a whole strip full of wallwarts. I would expect that to be more efficient than a bunch of individual supplies. Also, they could be built with better filtering and better regulation than wallwarts normally provide. That might even help your devices to last longer.

    This is actually something I have had it in mind to do at a few locations in my house for quite a while now. Someday I’ll get around to it! Maybe….

    1. Computers (and their peripherals) would be an obvious candidate for such a local DC converter approach.
      e.g. The original IBM PC had a special outlet on the back of the PSU for the monochrome adapter. That was just a regular outlet with a different connector (advantage being could turn it all on/off with one switch).

      Point is – around a computer one has a large DC converter (the PC PSU) and a lot of DC devices. (LCD monitors, modems, network gear, speakers, external hard drives, KVM switches, some printers, battery chargers (for mice, cameras, …).)

      A set of standard outlets that provided 12v and 5v from the computer PSU (with option of switched vs. unswitched) would get rid of a lot of wall warts (and make the computer more convenient to turn on/off).

      Additional benefit would be integrating this with battery backup. Typical inexpensive UPS are kind of silly because when operating on battery power they convert it to AC, which is then immediately converted back to DC in the computer/monitor/etc. Could be more efficient and convenient if the computer PSU had a special input connection where you could supply it with 12v DC from a UPS.

      Since typical PC power supply operates least efficiently when it is lightly loaded, adding the other loads to it could boost energy efficiency.

        1. I thought it was because in a datacenter you have many PCs. By running them off DC you are elminating X number of internal PSUs by replacing them with one external PSU. In a home, typically with one PC or at least one PC per location you would be elminating only one internal PSU to replace it with one external one. Electrically nothing would be changed!

      1. As [jadoo] suggested above, you could run this sort of thing off of a literal ATX PSU. I think that I would run at least a full bedroom off of it, though. Also, 12-volt is a no-go because the standard connector is big enough to be dangerous, I would go for a 24-volt “raw DC” connection instead (with ground shared with the AC line).

        1. Are you suggesting binding posts? Or maybe Fahnestock clips (google it)? Or how about running bus bars through your walls? The Molex connectors used in computers are okay, but NOT made for frequent plugging and unplugging. And as you say, the 12V cigar lighter plug is completely inadequate in several ways. Just keep in mind that if you decrease the voltage by a factor of one fifth, you increase the current by a factor of five, and require five times as much copper to avoid increasing losses. I forget now, what is it we were trying to improve?

          1. Bus bars? That’s some expensive power wire! How conductive is your tap water? Do you have copper pipes? Maybe you could install water heater dielectrics on the hot water lines right at each sink. Then the hot water lines could be your +5V bus, your cold water lines of course would be ground.

            No, don’t actually do this. I am not serious. I am not liable if you do!

          2. No, I’m suggesting a new standard DC power connector that you connect to a PSU inside a household distribution box. Probably a round connector (but not a normal barrel connector, purely because they’re already used so randomly), which has small holes that a finger can’t fit into.

            And fairly low amp too, probably a maximum of 10-ish amps if you feel the need for high-power stuff. If ~240 watts isn’t enough, probably a 48 volt system.

        2. You must mean “cigarette lighter plug” when you say “standard”. Yeah, those do suck. Don’t use them. Lacking a decent industry standard you could just chose your own. just pick a connector than can handle the power, is safe, will not be confused with other things and which you can get a plentiful, inexpensive supply of.

          Or… use Power Poles. It’s ALMOST a standard among ham radio operators to use Anderson power poles for 12V. http://en.wikipedia.org/wiki/Anderson_Powerpole Not everyone here is a ham but at least going with somebody’s standard means you are more likely to be compatible with something else.

          Powerpoles are great, they are generless so you won’t reach into your junkbox and find a handful of males when you need a female or vice versa. Also, they come as single conductors. They have specially shaped sides that allow you to slide them together in any configuration, side by side/ top and bottom, etc.. so again, you don’t reach into your junk box needing a 3 conductor plug but only finding 2 conductor plugs, etc… Just having a pile of power poles means you can make whatever connector you need.

          You could use side by side power poles for 12V, following the ham standard. Then use top/bottom powerpoles for 5V or some other voltage. That way you could keep only a supply of one kind of connector on hand and yet you couldn’t accidentally plug into the wrong port.

          1. 5-volt is already USB, but yeah, that Power Pole style connector could work. I’ve actually had some experience with large versions on an electric forklift. They do strike me as a bit more industrial than residential, though that might just be from the forklift connector requiring quite a bit of force (some people haven’t been able to get it plugged in until that understood how much force it required).

      2. I used to have a power supply that had an external 4-pin molex connector, the same as used by old hard drives, 12V, 5V and two grounds. I only ever used it to power up CD-ROM drives that I removed from PCs, forgetting to remove the CDs first. When that supply died I saved the connector to add to a new supply but never did.

  21. The only good DC energy sources are solar panels or chemical. But for large scale production it is all AC generators as they are more efficient than DC. The DC grid has its problems and upsides and so dose the AC. I like the idea of using small HVAC grids interconnected with HVDC, would use the good features of DC and AC. Even though the technology is there and there are a couple HVDC line in use if i remember correctly I don’t think that financially it is viable to switch to DC for the household. Aluminum plants would be really happy though.
    Its all comes down to physics and finance.

  22. This article is a total facepalm. The reason this issue was settled in the 19th century is because they settled upon the safest and most practical solution. Technology has changed, but electricity has not. The same disadvantages to DC that existed in the 1800s still exist today: safety issues, the difficulty of stepping voltages up and down, efficiency, and the fact that it is nearly impossible to move DC over large distances.

    1. Electricity has not changed since the 19th century!!!!???? Dave Berry was right, the power companies don’t MAKE electricity, it returns back to them through the other wire and they turn around and sell it again! B^)

      1. Semiconductors won’t change the fact that a DC power grid requires a generating substation every 1000 feet or so. Power transmission over distance has not changed because the laws of physics have not.
        Just because you can bump voltages up and down easily with low current applications does not mean that a DC grid is viable.

        Oh, and good luck finding an economical semiconductor solution for real power. I once sold a JUNK motor controller for 250A @ 480V and got $1200 for it. For a new one expect to spend as much as you would on a decent used car.

  23. Most of those things that run off DC internally will run off ~170VDC input also. One of the first steps in a switching power supply is a bridge rectifier. That rectifier converts the incoming AC to pulsed DC which is then smoothed by a capacitor. If you feed it DC, the rectifier will just pass the DC straight through and you get the same end result.

    1. @ab0tj, when you say “Most of those things…” do you mean “Most of those things that were designed for 220 VAC,” or “Most of those things that were designed for 110 VAC,” or something else? My EE skills are weak, I’m more mechanical.

  24. I would say this article is a rehash from the utility perspective. We are not going to change how power is transmitted or consumed from utility companies. What will/is/may change is if more homes become net generators vs consumers of electricity. Solar creates DC. Wind creates variable AC converted to DC. Is it more efficient to consume the DC internally, or convert it to AC. I don’t think a smart grid will get fully implemented in the US. The lobby groups are too strong/effective and politics has become a rich man’s game meaning the status quo will be very hard to change.

      1. Not how it works.
        The government states whether the utility companies are required to compensate for tie-in inverters. Whether you get credit for feeding energy back into the grid. The utility companies (lobbying against this government intrusion) would then need to build a smart grid and the system to balance the demand/supply. That means their electric plants would need to be more responsive. This is infrastructure. If we all don’t build it, it won’t come. This one capability is key to sustainable energy. The electric utilities will push those charges back to the consumer and increase lobbying efforts to minimize how much they must compensate for tie-in energy. If they keep that number low, there won’t be an incentive for residential consumers to do it and the utilities will have higher rates in place while building out the grid by demand. Keep the demand for a smart grid low, problem solved for the utilities.

    1. The biggest force pushing the “smart grid” is the reliably long-term utility-focus of power companies, not the “slam, bam, thank you ma’am” sound-bite obsession of politicians. Now, does that mean that all power companies will be happy to buy power from consumers? No, but the smart grid is about much more than that (e.g. electric meters that automatically report your usage).

      Also, consumers are often under the misconception that they should be paid the same amount for providing power as they get charged for consuming it. The persistent power grid problems of California back in the 90’s were caused (in part) by falling for that sort of nonsensical blather (spoiler alert: the blatherers bought the generators & took them elsewhere, while the power company that got accused of perjury turned out to be both honest and right), so this blind version of communism has a direct example to show that it’s falsehood. Consumers should be paid exactly the same for providing power as any other power station would: they deserve no more.

      They do, however, deserve every cent THAT ANY OTHER power supplier would be paid for providing that power, they just don’t deserve MORE, particularly since the power companies aren’t likely to get a choice on whether to accept it (also, more sources increases the chances for malfunctions, and complications resulting from grid damage, though it does also increase the reliability bell curve OUTSIDE of the 100% zone).

  25. I think the best solution is a hybrid. In the USA, Europe, Japan – everywhere; the best solution for regional / local mains distribution is AC. It’s less expensive for the initial capital expense, the infrastructure already exists, and solutions for protection such as fuses & circuit breakers are much simpler. But inside the house, it would be more convenient to have at least room-wise DC outlets – for computers & small electronics, lighting, and the like.

    But it is NOT true that ‘most’ appliances with electric motors would work as well on DC as AC. My oil furnace, ventilation, and well pump will not run on DC. The thermostats on my electric stove & oven would quickly burn out due to arcing on DC. My woodworking tools all have squirrel cage AC motors that will not run on DC.

    I would propose that someone could make a mint by developing a universal DC power supply that plugs into an AC wall outlet and ALSO includes a current sensor and power management circuit that shuts down the switching power supplies when no device is attached to them, thereby eliminating the small – but significant – energy loss in a switching power supply at idle. Made in huge numbers, such a device would be inexpensive and can be made reliably. It could have 5VDC USB outlets for phones and other small devices, 12 or 24 VDC for lights and other small electronics. And a SW configurable ~18 VDC outlet for laptops. This could be accomplished by keying the cord for a laptop such that the device outputs the correct voltage & current based on the laptop cord that is plugged into it. Yes, this device would draw a small current to run a controller to monitor the draws on the DC power outlets, but this could be minimized. And a unit could be made that would fit within standard electrical box sizes to manage lighting – right in the light switch panel, for example.

    As mentioned elsewhere in the comments, High Voltage / High Current DC is a very economical solution for connecting regional grids. The USA has 3 major grids, and Europe has more. Not only are England/France connected by HVDC systems, so is Norway & Germany, and many other regions. But that kind of economics does not translate to local distribution or in-house distribution.

    The final argument relates to homes with solar or wind power. There is a significant economic advantage to dumping excess electricity into the grid and getting paid for it. Unless a house is not within range of a power grid, the grid-intertied solution is better and AC is the way to do that for the foreseeable future. Safe grid intertie inverters are readily available and affordable. So if you already need to make AC using an inverter for the grid intertie, it still makes sense to use AC for traditional AC devices within the house.

    1. 7 volt DC (to be regulated locally to 5v, thus accounting for losses) and 24 or 48 volt “raw” (unregulated or poorly regulated) DC (because 12 volt has a dangerous standard connector) should provide for all of the half-way normal DC needs in most rooms, and be possible to provide from a single per-room converter. Your programmable-voltage system would certainly belong in individual outlets, since higher frequencies are supposedly more efficient for switching, but cause problems sooner when dragged out over a distance (we don’t want to use BNC connections to run power, and we don’t want a single-point converter to have to run a bajillion cables since it would be a nuisance to run them). Of course, electricians would complain even about the two extra “live” (assuming that “neutral” is shared with AC ground or AC neutral) anyways :P .

  26. tl;dr – Troll post, lol

    1) AC power is 120V people not 110V… Elec code calls 120V “nominal” as its 5% regulated (115-125V). Check the ratings on wall plugs and switches. This misconception was caused by pissy motor winders who did not want to accept AC and thats why AC motors are rated 110/220V. This gets annoying at higher voltages as a 480V suppy runs 460V motors.

    2) DC wired in a house is a waste of money. Even if you have $50k in a solar setup the cost of install is more than an inverter, and you would still need batteries. That inverter efficiently prize is what we need, not this concept.

    1. The capital costs to change over ac to dc or from 115 to 240 V across the country would be staggering. We have a standard and everything in the country works with it, it is no small thing to change that.

    2. To the best of my knowledge, the majority of US ovens & at least one other common major appliance (window-unit air conditioners? water heaters? clothes washers or dryers? I forget which) run off of 220. I know that most of the heavier-duty window-unit air conditioners run off of 220 (a relative once had to make do with a 110 unit, I think for wiring reasons). If we had a reason to run most of our stuff off of 220, we would ALREADY be doing it.

  27. AC vs DC, The war is over, LONG LIVE THE WAR!

    OK so for the argument that long distance power transmission is only the realm of AC, look over at the Pacific DC intertie.
    http://en.wikipedia.org/wiki/Pacific_DC_Intertie 500kv lines x2 for a 1 million volt potential, 3.1 Gigawatts, ~%40-50 of LA’s power needs. Its a big boy when it comes to DC transmission. Lines are small, support is small and light, only two conductors. Earth grounds are MASSIVE! miles in diameter etc

    Ohm’s law, I don’t recall see much work by Ohm on AC systems. Others sure, a modified version has to be used to fully describe AC systems.

    Can you have an all DC house, sure get an RV. However if you have a lot of digital gizmos having a dedicated DC loop in your home would make a LOT of sense from a power supply and standardization stand point. Still not practical today but it merits thought.

  28. What if instead of having all of those power bricks and wall warts throughout the house, would having only one AC-to-DC converter in your entire house make it more efficient? and would that be a lot?

    (So AC to your house, then AC-to-DC, then DC to the outlets, lets say.)

    1. Instead of trying to route DC bus through an entire house, it might be better served by having multiple AC/DC converters that serves small island of devices near by. e.g. I have a cluster of network stuff around my cable modem which has a DC bus with battery backed.

      http://tekkieneet.wordpress.com/2013/08/10/ive-got-your-back/

      Since the power spec of each of those devices change over time, I made pluggable DC/DC switch mode modules for them and a break out to accommodate for the power plugs. I had to make an isolated DC/DC module for my phone as there are grounding issues with my ATA. For devices that worked over the entire range, that is simply a jumper. At one point made a boost module for the DSL modem that requires 26V DC.

      There is a charger, OR’ing diodes for battery backup, bar graph for bus voltage display and undervoltage shutdown circuit. It is currently running 5 devices, so saving 4 bricks.

      There is another cluster around my computer for Ethernet switches and computer speakers. That latter is served by 9V input brick with a 5V switch mode converter.

  29. I don’t want a DC house. DC has an annoying tendency to spot-weld contacts. I’d prefer high-efficiency switching rectifiers with very low standby current over long stretches of low-voltage wiring.

    I have lived in older houses/apartments missing 3-prong plugs and having miswired grounds. To have a second system that adds a chance of a DIY homeowner swapping the AC and DC is asking for trouble.

  30. All DC has its advantages, to an extent that indeed it is already used for some data centers exclusively:
    http://www.heise.de/newsticker/meldung/Rechenzentrum-mit-Gleichstrom-1605295.html
    And I suppose Hangman is right: Most appliances, including big ones like washing machines, rectify their input and internally use DC. Even more so for electronic devices like smartphone chargers, PCs, TVs, etc.
    What I personally would expect is, that in a few years time, the manufacterers will produce “DC-compatible” equipment, with the difference with regard to today being officially supporting plugging it into, lets say, 220V DC. After that point, newly build houses / villages / whatevers could, if the local energy supplier decides so, be supplied directly with said voltage. So when the time is ready, those parts could easily be integrated in an DC-only grid.
    With regard to safety, there are possibilties to detect faults in DC-Systems available today. If they were used more, they would probably soon be reasonably priced.

    As a sidenote: High temperature superconductors are used in few grids already. The suprising part is that at least two of them use a 3-phase-AC system, for instance
    http://www.rwe.com/web/cms/de/1301026/rwe-deutschland-ag/energiewende/intelligente-netze/ampacity/ or the Holbrook Superconductor Project run by the Long Island Power Authority. I would love to see the combination of HVDS & superconductors!

  31. Shame on you. I expect BETTER from hackaday than this. Edison did not invent the “first practical electric bulb”. He stole it from Joseph Wilson Swan, who patented the same design significantly prior to Edison’s patent. There was then a lawsuit. Which was settled, because it could be proven that Swan’s design existed prior to Edison’s.

    Edison is not an inventor. He is a thief. I cannot think of a single “invention” of Edison’s that was not either stolen outright, (the light bulb, via swan), contracted for but never paid for, (his “improvements” on power systems, which he never paid Tesla for). Or the work of a sub-contractor in his thinktank.

    1. Edison really did invent the first practical light bulb. It was not the first light bulb by a long shot, but it was the first _practical_ one! Swan’s bulb while similar, lacked the all important screw base.

      Edison also paid Tesla’s salary. What Edison did not do was pay a bonus he allegedly verbally promised Tesla. As legend has it when confronted about it Edison claimed it was a figure of speech he was using, not an actual agreement.

      Tesla was no subcontractor for Edison, he was an employee. Where do you get the nonsense you spew?

      1. Because, of course, Swan’s bulbs were not in use before Edison took the design an “improved it” and, of COURSE every single lightbulb out there utilizes the Edison base today! Wait…no. Both of those are false.

        And, sorry, no, “I will pay you X amount of money for completing said project” is not a joke. It’s a verbal contract, that Tesla should have gotten in writing.

        And fine, if you want to split hairs, he had employes in a think tank. Not contractors in a think tank. A significant difference, I’m sure.

        1. Tell me, where can you buy a Swan socket today? Can I get one at my big box store?

          As far as what Edison actually said to Tesla we’ll never know. Based on what Edison himself said it was more like, why if you can do that I’ll pay you a million dollars! Which is just an expression of speech meaning it would be something impossible to do. Doing it one should never expect to collect on hyperbole as if it was meant literally.

          Yes, the dynamic is significantly different being a salaried employee, as opposed to a private contractor when dealing with people who might be considered your superior in one instance, and just a customer in another.

  32. To me, AC had two big advantages during the war of the currents: power conversion was easy to do with transformers, and you could drive a large motor without needing brushes. I don’t know if DC converters are as efficient as transformers at large power scales, but I do know that modern motors are increasingly driven in a way very friendly to DC mains power: Variable-Frequency Drives, or VFDs.

    Essentially, a VFD is an inverter that can vary its frequency and voltage as required. This allows it to adjust the speed and torque of a motor in a way that wasn’t possible with systems driven directly by AC mains power. Some even measure the current flowing to the motor as a feedback mechanism. As you can imagine, this is much more efficient than the alternatives of using brushes in the motor or supplying mains power directly to the windings. This is why Tesla uses a VFD as the drive in their cars, although theirs are complicated by the fact that they also do regenerative braking.

    So, VFDs are more efficient, and are increasingly common in household devices (ex. Dyson’s “Digital Motor” is just a VFD driving an AC motor). This pretty soundly kills the idea that household appliances can’t easily start using DC instead of AC, as in many cases you can practically rip out the rectifier circuit in a VFD and supply DC directly.

    Thus, while a DC grid might still be a long ways off, a DC household can be reaslistically done today.

    Note: As is common in engineering, the core concept of switching DC power to drive a motor with a waveform goes by many names. Look for BLDC (Brushless DC), VFD, or electronic commutation. You can argue all day about the *exact* implementation details, but they pretty much all come out to being the same circuit.

  33. Actually, most in-wall NEMA 5-15 / 5-20 outlets that have built-in USB charging ports are switch mode supplies. SMPS are smaller and more efficient. Not sure why the author would suspect most are linear….

  34. I didn’t notice if anybody mentioned corrosion. DC is much more prone to causing corrosion than AC as well. Moisture will make connections deteriorate over much less time on a DC circuit than on an AC.

  35. It may have been said, but just in case, and it is worth repeating:

    You cannot just convert DC to another voltage unless you are dropping it by an energy wasting device like a resistor or regulator. Inside DC to DC converters there are oscillators, rectifiers and most ofter a transformer.

    Also keep in mind no converter is 100% efficient so… loss, loss, loss…

  36. Look to datacenters and telecoms for the answer to this. There are telecom equipment rooms that run on 48VDC.
    The Open Rack part of the Open Compute project talks a bit about leaving the option open for DC power to be used. I think Google talked about this as a possible strategy for reducing the AC-DC switch mode PSU losses by centralizing that conversion as much as load and wiring will allow. But I can’t find that article (so maybe I’m remembering it wrong).

    1. The 48VDC standard is more due to custom than any actual difference in efficiency. Telecom equipment has operated on 48VDC forever, because early telephone systems ran on dry cell batteries at the central office (probably mainly because many customers didn’t have electric service), and 48V was the sweet spot for powering telephones through mile-long twisted pair of a reasonable gauge. Everything past that was just a matter of using the existing standard, because of the huge installed base, and the corresponding easy availability of 48V DC-DC power supply modules.

      1. There’s also the increased reliability of providing the power in conjunction with the data connection. Back in the 90s, the only half-way likely ways to lose your phone connection were via damage (rare, particularly since most of the copper is apparently buried for precisely this reason), or failure to pay your bills. Thus, phone connections were more trustworthy than power, which even in many cities and towns is carried on poles that can be taken out by both drunk drivers and various forms of storms.

  37. come on people, if you want to go DC you have to look no further than truck stops and mobile home supply shops. They have just about anything you would need appliance wise for DC current usage from washers and dryers to refridgerators and cooking equipment, not to mention Televisions that can use DC currents. granted it’s not big stuff but for the basics for one person cooking, works just fine. so no 55 inch tvs (without modifications) more like 13 inch with DVD players built in. From there it’s just recepticals from the mobile home supply shops and a DC power supply. it’s feasable in a house like you would see on the tumbleweed house website, but with the limitations of length on DC not as easy in a full size house.

    1. I had to Google tumbleweed house website. Now that I’m looking at them I’ve seen similar before. People sure are scaling back expectations these days aren’t they?

      Let’s all sing the I’m Living in a Cardboard Box song together shall we?

      1. Those really little houses are mostly for retired/retiring couples who want to be able to move around on a sort of permanent vacation, hence the trailer bases that they often get built on. As far as per-square-foot is concerned they’re probably hideously expensive (disclaimer: I was once trained as a professional draftsman, and at least at that point in time the per-square-foot costs actually rose as houses shrank, especially if you didn’t count stuff like hot tubs into your square-foot costs), but when you consider the total size, as well as the cost of some trailer park locations, they’re honestly quite cheap, potentially as low as a cheap apartment.

        They’re also nice if you’re moving around a lot (e.g. professional welder, any other profession that involves on-site work in widely dispersed locations), or if you expect to move soon for work. They are not so good for families with children (someone probably thinks otherwise, but I’m content to say that they’re wrong… and that if they don’t stop moving their children will never be socialized), or if you have a lot of stuff (e.g. a home machine shop).

        That general SIZE, however, would probably be good for most new college graduates: it should be much cheaper to rent or buy than any full-sized house (you do remember that most people don’t start off with 100-room gold-edged mansions, right ;) ? ).

  38. TLDR:

    P=V*I
    where
    V = Potential Difference as Volts
    and
    I = Current as Amps

    Increasing the Voltage capacity of wire means adding more cheep plastic as an insulator to prevent arcing.

    Increasing the Current capacity of wire means adding expensive copper as a conductor to prevent melt down.

    Plastic is cheaper than copper. To run a stove from 12 Volts DC would take wires an inch thick and they would cost over $400 per meter. (Ball park figures).

    The mains coming to your home will always be AC as it is the most economical solution.

    DC in the house would only be the result of an alternate energy source such as solar.

    Low voltage DC is only economical for low power needs such as lighting, entertainment, computers etc.

    A stove or air conditioner is never going to rum from DC unless it is within inches of the power source.

    Existing 12-48 VDC homes use DC for low power devices like lights, inverters for mid power devices like fridges and gas of other means for high power devices like stoves.

    I used to work in a telephone exchange where the electronics ran on 48 Volts. Power distribution was not wires, but bus bars. They had a cross-sectional area of over 20 square inches each and would have cost a fortune. Like $300+ per inch of length.

  39. I live on a boat and actually have 220VAC, 110VAC, 24DC, and 12DC. The hvac runs off 220 with the range, the appliances and tv use 110 with a few convenience outlets spread around. the rest is either 12v or 24v for LED lighting it’s a lot more efficient. There’s a ton of resources in the marine/rv arena for living with only 12v

  40. Motors run best on AC.
    Voltages must be high enough to have low enough current to be practical to transmit. I think around 100 volts is minimum. I wish my whole house was 220.
    Its not that hard to add a small rectifier, but a lot harder to add an inverter.
    Welders would be way to expensive to run on DC.
    Its easier to isolate AC than DC.
    And as said, AC arcs break much easier in switches than DC does.

    1. Motors run best on AC, but the frequency of AC they require depends on the speed they need to run. American motorized appliances are designed to run at 3450 or 1750 RPM, because these are the speeds that 60 Hz two-pole and four-pole induction motors run, under load. If you want any other speed, or heaven forbid, variable speed, you somehow have to generate the appropriate frequency. In the old days (i.e., the day before yesterday) this was done using mechanical commutators on what were called “universal” (meaning AC/DC) motors. These days it’s done with brushless DC motors, which are essentially three-phase synchronous AC motors with integrated variable-frequency inverters. So while you’re technically correct, it would be more accurate to say that motors run best on variable-frequency AC. And most of our houses aren’t wired for variable-frequency AC.

    2. Your whole house is 240VAC. Standard outlets in the USA just use one hot, and a neutral that get you 120VAC though. But between the hots it is single phase 240VAC

      Look at this picture
      http://www.icreatables.com/images/electricalimgs/electrical-breaker-wire-routing.jpg

      This is what a breaker service panel looks like with the cover panel off of it. See how the silver conductor bars in the middle look kind of like interlaced fingers? Each side is a hot of the 240VAC feed. A single pole breaker only connects to one finger. A double pole breaker hooks up to a finger from each side. Single pole 120VAC, double 240VAC

      All the breakers on the left side are double pole 240VAC What we see on the right is more of a mixed bag though. From the top there is a single, then a double, then a double single (duplex), or two breakers running off one finger.

      The other interesting feature to note is the two vertical bus bars on the outermost part of the panel. On the left we have the ground, and on the right the neutral return. Physically they are the same thing. But practically neutral return is the power path, and ground is the safety return. You should never run power through the ground on purpose either. There is a lot of code, and crap that make it how it is. Some things are better taken on faith. Let’s just say that neutral, and ground are one of them too (I don’t feel like explaining it here).

      Now the mysteries of your house’s electrical system have been revealed to you! As long as you get a staple within 6 inches of a box entrance you might even pass inspection if you do any DIY electrical projects. I perform all of my own electrical stunts at home :)

    1. For transmitting power, it’s the voltage that’s more important than AC vs DC. High voltage allows the very-high-voltage pylons to transmit large amounts across countries. For a system where power is generated at (I forget, a few thousand volts), then transmitted at around 100,000, brought down to a few thousand for local distribution, then down again to 220 / 110, you need a lot of transformers. And they only work on AC.

      On a ship, I would guess the electrics are all at the same voltage, from the generator to the lightbulbs. So they don’t need transformers. Because the cable runs are much shorter, the voltage drop isn’t as much.

      The advantage of using DC over AC, is that many ships have backup batteries. Particularly pre-nuclear submarines, of course. Running DC makes it easy to charge the batteries, if you have 220V worth in series. Then in an emergency, just switch over to batteries and everything carries on working as it was. It’s nice to have everything working even if the engine room is hit by a torpedo or whatever.

  41. As a middle ground, what about 170V peak trapezoid wave AC? It allows inverters to be very simple yet very efficient, solves the arcing issue with DC, works great with rectifier loads such as existing electronics, and doesn’t cause EMI the way square waves do.

      1. Depends how the motor is designed, it is possible to make an induction machine with trapezoidal back-emf. This type of motor will run fine with a square wave and perfect with a trapezoidal one. Many BLDC motors are made for this type of operation, since they are often driven with square-wave inverters.

Leave a Reply to iceboxCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.