Design a better inverter- win a million bucks!

[mikeselectricstuff]

https://www.littleboxchallenge.com/

I kinda think they're trying to solve the wrong problem though.

Why spend all that effort converting HVDC to AC, only for a large proportion of appliances just rectify it back to HVDC again.

I wonder if they'd have been better off looking at ways to make appliances work safely & efficiently form variable HVDC supplies

[HackedFridgeMagnet]

that's thinking outside the box. (sorry)

Totally agree, we can convert DC voltages very efficiently now so why not use it for most low-medium power appliances and especially lighting.

I guess you still need a grid tied inverter/rectifier though. That wont change soon.

[AG6QR]

I also think they might be solving the wrong problem, by concentrating on size, instead of on efficiency and cost.  Granted, if you make it small, you may have to make it reasonably efficient, because any waste energy has to be dumped as heat, and small boxes have trouble getting rid of heat without causing damage.  But size is much less closely correlated with cost.  I think the world needs cheap and efficient inverters more than it needs tiny ones.

I can agree with mikeselectricstuff's logic, as well.  It's probably better to figure out how to avoid multiple conversion steps entirely, instead of figuring out how to do the conversions better.

On the other hand, if they do prompt people to solve the problem as they've stated it, even though it's not exactly the problem I would have proposed, it still may stretch the state of the art, and that's not such a bad thing.

[T3sl4co1l]

General purpose, I suppose...?

Maybe Elon Musk dropped by Google and said, "Ya know boys, your self driving cars are great, but my customers are going to need to run their coffee maker, while they're on the road charging their iPads.  Wouldn't it be nice if there were a compact solution for that?"


The ripple requirement is ridiculous.  Presumably they're thinking of electric car batteries or something like that, which will have much less than 10 ohms ESR, and won't care about ripple.  (..Do they?)

They're also asking for a 30C temp rise ("15-30C ambient", 60C max at any point on the outer surface), which in under 40 in^3 x 0.5" = under 160 in^2 (two sided) and at typical convection rates of 150 C.in^2/W, that's about 30W tops.  The minimum efficiency is 95%, so 100W max dissipation, and that's if you're willing to sacrifice internal volume for heatsinks and fans.

I'll have to run some numbers and see how far from theoretical the volume is.

Tim

[SirNick]

Totally agree, we can convert DC voltages very efficiently now so why not use it for most low-medium power appliances and especially lighting.

Don't most of the techniques used to convert DC-to-DC efficiently rely on an intermediate stage of AC?  (Pulsed DC at least...)  Transformers and inductors are used for just about everything but linear regulation.  If AC is necessary anyway, why not stick to the tried and true?  (Although sine waves are maybe an unnecessary burden a lot of the time.)

Now, I do see the benefit of relatively low-voltage DC buses in-home.  I have considered setting up a large UPS-like system for stuff you really want working through (possibly extended) outages.  E.g., lights, smoke detectors, alarm systems, boiler-room water pumps...  We had a wind storm here a year ago that knocked out power to some neighborhoods for upwards of a week before the utilities could catch up.  It was during the fall.  A couple weeks later, and things would've gotten really bad -- pipes freezing, pets dying, etc.  Assuming residents themselves could just go somewhere habitable, that is.

Anyway, I wonder how successful this is going to be.  It seems efficiency is already getting pretty good.  A few more percent would certainly be nice, but what's wrong with a cooler-sized device?  Is that really a limiting factor for homes?  (For vehicles, sure, I get it -- but then, do we need the inverter at all?  Or is the DC rail voltage good enough as-is?)  Curiouser and curiouser.

[mtdoc]

Not sure I see the point in this.  I've got 4500 watts of PV, a 436 AH 48V battery bank and a 3600 watt inverter for back up power of our modern 3000 square foot home.   The inverter is not "cooler sized"  (more like the size of a 12 pack of beer) - and it's size is really not an issue.  Direct grid tie solar system inverters are even smaller.

BTW - DC only systems for off grid or back up lighting and appliances have been used for years. Modern inverters are so efficient that there is little to be gained. In addition, the cost of the additional copper needed to minimize losses with DC wiring and the extra cost and hassle of using DC only appliances has made this less and less popular.

[johansen]

about the only thing you need 60hz ac for is single phase induction motors and florescent lighting. and cheap wall warts.

i'm not aware of a problem that needs fixing here.

off grid folks have always been left with three options: cheap msw inverters, cheap sine wave inverters, and expensive sine wave inverters.
expensive sine wave inverters don't blow up when you start a 5 hp induction motor, or accidentally short the output. some of the 6 kw sized units will have no problem running stick welders, which are about 50% lagging power factor.. worse if its SCR controlled. Typical "ratings" for a good quality inverter is something like double the rated output for 5 minutes.

usually the topology of sine wave inverters are up to dozens of phases of forward or flyback converters to take the 12,24,36,48 volts nominal battery voltage and convert it to a floating 300+vdc or +/- 180+vdc, then sent to a full wave or half wave bridge, then filter the output.

the input side forward or flyback converters can be turned off to save switching losses at low power, but the output side half or full bridge cannot.. unless they have multiphase output inverters.. which they might start doing.

it might be interesting to see if a three level inverter could be built from 200 volt mosfets and diodes with lower losses than a 2 level inverter from 600 volt rated IGBTs or 500 volt mosfets. to see such would be about the only thing i would find interesting resulting from this competition.

input ripple tolerances are rather interesting.. it looks like you could be forced to use a PFC boost converter to take the 450 volts and boost it to 550 in order to have sufficient energy storage in 600 volt electrolytics to keep the input ripple down.. that way you could let the voltage flop around from 550 volts to 450 volts on your energy storage caps to keep the input ripple down. --such a design would seriously eat into the minimum efficiency specification because its hard to get beyond 98% when dealing with boost converters, diodes and 650 volt mosfets. unless of course you triple the cost with resonant turn on and turn off circuitry.

[SgtRock]

Greeting EEVBees:

--$1,000,000!!!! Just now in the register, a Household Inverter Design Contest.

--I was thinking that Dave could probably easily design one of these, perhaps by kickstarting the R&D budget. If only he knew some group that might be interested in his project.

http://www.theregister.co.uk/2014/07/22/google_offers_one_milleeon_dollars_to_inverting_acdc_boffins/

"If Mr. Einstein doesn't like the natural laws of the universe, let him go back to where he came from."
Robert Benchley 1889  -  1945
 
Best Regards
Clear Ether

[Skimask]

https://www.google.com/search?q=upside+down+house&espv=2&tbm=isch&tbo=u&source=univ&sa=X&ei=BTbPU9zTKoaOyATRoIHYAw&ved=0CBwQsAQ&biw=1920&bih=955

Do I win?

[David Hess]

The switching power supplies commonly found in consumer and computer gear will often work without problems with DC inputs from 170 to 340 volts.

[calexanian]

They just want somebody else to do their design for them. If I was going to go through the trouble to design it, I would sell the damn thing myself! The part 15 FCC bit will be the hard part. Handling that much power in 40 cubic inches starting from such a high voltage is easy enough. Meeting the noise requirements would be the tough part. I should make a big 10 square inch hydrogen thyratron, or better yet a mercury pool ignitron and a chopper circuit. Watch it blow up all their test gear when they go to try it out! Or a micro tesla coil and have it send arcs all over their lab!

[johansen]

mercury pool ignitron and a chopper circuit. Watch it blow up all their test gear when they go to try it out!

how many volts per nanosecond do i need?

[ejeffrey]

I am not sure household HVDC distribution actually saves you that much.  HVDC-DC power supplies don't really have any advantage over AC-DC.  DC-AC inverters are more troublesome, but I don't think that is fundamental: they are just much less common and haven't been optimized greatly.  The obvious benefit of AC is the ability to use induction motors and magnetic fluorescent ballasts, which is great, but those can be replaced with BLDC motor controllers and electronic ballasts.  However, AC has superior arc clearing and switch rating/contact life.  Between those advantages and market inertia, I don't see HVDC replacing AC any time soon.

A key point is that you don't actually get rid of inverters by going to HVDC distribution, you just replace them with DC-DC converters.  Your solar panel or battery or whatever is likely to not put out the standard household voltage you decide on.  In the case of solar panels you want to be doing power point tracking if possible, while some battery chemistries change substantially with state of charge.  So you still need a converter.

Now DC for grid distribution is great.  Right now it is mostly only used for long distance links or when connecting multiple grids, but I wouldn't be surprised if eventually we have DC distribution all the way to the neighborhood distribution point where the power company had an inverter that supplied people with household AC.

[rob77]

HVDC in the household instead of AC ? back to Edison ?

[HackedFridgeMagnet]

Would corrosion also be more of a problem with DC?
Such as with wiring joins subject to moisture.

[David Hess]

Would corrosion also be more of a problem with DC?
Such as with wiring joins subject to moisture.

I do not think it would be any worse than AC.  Either will cause electrolysis.  The power level is high so "high open" type failures like you find with telephone wiring are not going to happen.

[rob77]

i think changing to DC would be a big pain in the ass for the end customer - how would the generic user know whether or not his device is suitable for DC ? for new products it's easy - the manufacturer could indicate on the product that it's suitable for both AC and DC input... but what about the billions of devices already manufactured ? even if the older equipment is suitable for DC - it's not indicated on the device.

[EEVblog]

Why spend all that effort converting HVDC to AC, only for a large proportion of appliances just rectify it back to HVDC again.

Because almost everything on the plant needs that AC. Unless you plan on throwing everything away and starting again.
There is a need for both solutions.

[Precipice]

I do not think it would be any worse than AC.  Either will cause electrolysis.  The power level is high so "high open" type failures like you find with telephone wiring are not going to happen.

Balanced AC pushes those ions back on the next half cycle - no effective corrosion (unless you manage to engineer some weird copper-oxide semiconductor).

[KedasProbe]

first prototype drawing, they are working on making it a bit smaller.

[tszaboo]

They are asking for power density, which is about 10 times the currently available inverters. I'm sure that the challange is possible, maybe it would be worth to make an asic for the control. But.
It wouldnt be cheap, it wouldnt be ready in time, and most likely it would be loud as hell. You still need to dissipate a lot of power and that needs cooling.

[David Hess]

I do not think it would be any worse than AC.  Either will cause electrolysis.  The power level is high so "high open" type failures like you find with telephone wiring are not going to happen.

Balanced AC pushes those ions back on the next half cycle - no effective corrosion (unless you manage to engineer some weird copper-oxide semiconductor).

I have run across electrolytic corrosion caused by wet conditions in power line AC wiring before.  It may take a longer but the result is the same; it makes a big mess.

[David Hess]

i think changing to DC would be a big pain in the ass for the end customer - how would the generic user know whether or not his device is suitable for DC ? for new products it's easy - the manufacturer could indicate on the product that it's suitable for both AC and DC input... but what about the billions of devices already manufactured ? even if the older equipment is suitable for DC - it's not indicated on the device.

The only way this could safely work is with a deliberately incompatible power plug and socket as well as IEC connector.  This might be a good idea anyway because of the non-extinguishing arc characteristics of high power DC.

[HackedFridgeMagnet]

From the contest site:

One promising set of new technologies which may allow for the achievement of higher power densities are wide bandgap (WBG) semiconductors, such as Gallium Nitride (GaN) and Silicon Carbide (SiC).

Maybe they think we might be able use these to get higher switching rates. I can't see how else these semi conductors would help power densities.

If you could achieve a switching rate of 10x does anyone have a rough calc on what power densities you could then achieve?

[ejeffrey]

Maybe they think we might be able use these to get higher switching rates. I can't see how else these semi conductors would help power densities.

Possibly smaller heat sinks if they can operate at higher temperatures.