To Power the Future, Carmakers Flip on 48-Volt Systems | Hacker News Reader

To Power the Future, Carmakers Flip on 48-Volt Systems

nytimes.com

57 pointsdwynings62 comments

It's too bad that automakers didn't make this change a long time ago.

One thing the article didn't mention is that power losses in a wire are a function of the square of the current. So, let's say one needs to deliver 48 watts to a load. In a 48 V system, that's 1 A of current. Using 12 V, that's 4 A of current.

Power loss is 16x greater at 12 V for same diameter of wire, loss = (4 A * 4 A) * resistance of wire.[1] Therefore, wires are made thicker in 12 V systems to avoid this. That's a lot of extra weight in a car.

Before the nitpickers chime in, I know the above is simplified. But it also clearly explains why long distance power transmission lines run at such high voltages, 230 KV or higher.

[1] https://en.wikipedia.org/wiki/Copper_loss

You're not wrong, but you should probably tell the whole story if you're going to post on HN.

Primary Objection: Um, Ohms law gives V=IR. And P=IV. So, P=I^2R. Except that P ALSO can be calculated as P=V^2/R. So, we got squared losses anyway and R should go the other direction. WTF?

The answer is that a copper wire is a distributed system, not a lumped system. So, you derive the power loss from the generated heat across a differential slice of the line and then integrate over the whole length of the line--which gives you Power=RI^2 in the limit (and V->0 for any individual slice as the differential ->0).

It is one of those (un?)fortunate happenstances (like the Bohr hydrogen atom model) where the system in the limit sums to something we already know and leads to a bit of misunderstanding when you confuse two things which appear to be related but really aren't...

The V in V^2/R would be the voltage drop along the wire, not the total system voltage.

> It is diesel's decline, executives and analysts say, that has finally set the stage for mass electrification. While diesel pollution problems became notorious with the Volkswagen test-cheating scandal, the subsequent shift to gasoline is bloating carbon dioxide (CO2) emissions, making the next round of European Union goals harder to meet.

> PSA Group, which had previously seen no need for 48V hybrids, now plans to introduce them "across the board" in response to diesel's faster-than-expected decline, the Peugeot maker's programs chief Patrice Lucas told Reuters.

> "What automakers are finding is they need more than just advanced combustion engines to reach the fleet average reductions," said Mary Gustanski, Delphi's engineering boss.

quoted from http://europe.autonews.com/article/20170923/ANE/170929892/vw...

This article conflates two things. One: adoption of a 48v standard for powered systems because it's a sweet-spot for efficiency, current draw, systems design. Two: PHEV and the adoption of electric transmission and drive train, which permits a non throttled engine of any form to run at constant high efficiency or not at all and use batteries.

Two things. One story. Cars are becoming like common rail deisel with a unified power model for systems integration and cars are becoming like locomotives which use electric traction for systems efficiency.

The 48 V DC "sweet spot" is mostly safety / regulatory. As soon as you get to 50 V, the safety requirements (NEC) change. You can hold on to bare conductors with a 48 V potential across them just fine, even though 120 V AC will kill you and make it hurt the entire time you are dying.

48 V devices can require connections that are inconvenient and large, depending on what you want to power with them. Try running an air conditioner off 48 V and measuring the peak current draw.

https://electronics.stackexchange.com/questions/267789/how-s...

> You can hold on to bare conductors with a 48 V potential across them just fine

Even with wet hands? I was under the assumption the unconditionally safe upper limit was around 24V.

> "But skeptics may wonder why engineers are devoting so much effort to developing the mild hybrid systems that could be seen as stopgap measures on the path to a fully electric vehicle fleet."

Let's do some napkin math.

Assumptions:

1. Gasoline will never exceed $10/gallon. (assumes eventual plant/algae sourcing/some kind of magic). 1a. Vehicles achieve 50 mpg (2 gallons/ 100 miles) via tricks like 48-volt.

Thus future gasoline MHEV vehicles cost 20 cents/mile to power, so $20 per 100 miles

2. Batteries get stuck at about $100/kWh. (maybe lithium etc never scales) 2a. Consumers are satisfied with 60 kWh batteries.

Thus BEVs cost $6000 more than comparable MHEVs (napkin math)

4. Electricity costs $0.20/kWh (assume some surcharge for the charging station). 4a. Your BEV gets 34 kWh / 100 miles (99 MPGe)

Thus to power your BEV costs $6.8 / 100 miles.

5. The difference between 20 and 6.8 is 13.2 - so you are paying off your battery at $13.2 / 100 miles. 6000/13.2 is about 450

---

Conclusion: It will take 45,000 miles to pay off your battery, assuming the competition is a 50 mpg vehicle and gasoline is $10/gallon.

Additionally, the initial cost for a BEV is $6000 higher, which is quite significant for entry-level buyers.

---

With 33 mpg and $5/g you have $15/100 mile, thus a 80-90,000 mile payoff.

With 33 mpg and $3/g you have $9/100 mile, thus 270,000 mile payoff

With 33 mpg and $2/g you have $6/100 mile, and you will never pay off the battery investment.

With 20 mpg and $3/g you also have $15/100 mile = 90,000

---

Home Solar: add $10,000 to your battery = $16,000

50 mpg and $10/g you have $20/100 mile = 80,000 mile payoff

33 mpg and $ 5/g you have $15/100 mile = 107,000 mile payoff

33 mpg and $ 3/g you have $ 9/100 mile = 179,000 mile payoff

33 mpg and $ 2/g you have $ 6/100 mile = 267,000 mile payoff

20 mpg and $ 3/g you have $15/100 mile = 80,000 mile payoff again

Replying to myself: One thing the above does not take into account is the feeling of a BEV, which internal combustion and weak hybrids simply cannot match. I deeply feel that luxury customers will demand strong electrified powertrains after experiencing them.

Strong hybrids could match the feeling of a BEV, at a similar per-unit cost.

One issue is engineering and certification costs - manufacturers can only do so much at once or else shareholders get anxious.

Also not included are warranty reserves (how much to set aside per vehicle for warranty costs). More complicated vehicles call for more reserves.

> One thing the above does not take into account is the feeling of a BEV, which internal combustion and weak hybrids simply cannot match.

I absolutely agree about most luxury customers. However, there are certainly those of us in the minority who favor a more mechanical experience when driving a car, and most electric cars seem to move even further away from this in pursuit of efficiency, etc. Consequently, I don't think this market will exist for much longer once driving becomes more commoditized.

I'm certainly on a tangent by now, but there an interesting piece by Malcolm Gladwell in Car and Driver about a possible result of the rise of autonomous cars on car culture today: www.caranddriver.com/features/when-autonomous-cars-rule-what-will-happen-to-car-culture-feature

>Gasoline will never exceed $10/gallon.

The consequences and repercussions of global warming and climate change plus Nixon era doctrine of interventionalist policy in the interest of maintaining global US oil hegemony have easily put real barrel cost easily above this. Iraq didnt pay for itself twice, after all.

Algae derived oil, if it could be made to work at scale, has none of those problems. It is a not unreasonable upper limit for a totally renewable source of petrochemical fuel.

There's plenty, like a lot, of shale oil available at $80-100/bbl, which allows for significantly less than $10/g gas.

The only meaningful challenge to $10/g gas is policy (ie taxation).

Why everyone talks about switching to electric cars while nobody ever mentions electric heating of buildings as environmentally friendly alternative? Because it isn’t? Why bother with electricity-saving lightbulbs while EV owners burn the current like there’s no tomorrow...

In most of Europe, gasoline is already close to 10$/gallon (1.50€/l or more).

The big error in your calculation is, that you calculate only the cost of the battery, but not the cost reduction of removing the most complex parts of a combustion engined car: the engine itself, the gearbox and the whole exhaust system including cleaning. If you calculate this in, you should come up with very different numbers - a lot of experts think, that electic cars might be even cheaper to buy in the early 20ies years.

So far electric cars cost more than ICE including this cost reduction

That is correct. Most car manufacturers currently pay way more than $100 per kWh battery capacity. Also, the electrical vehicles are made in much smaller numbers than combustion engined ones, raising production costs considerably. But for a long-term calculation, you cannot ignore the cost savings of leaving off all the combustion-related technologies. And that there are significant cost savings should be obvious.

You are correct, it is napkin math after all. Assuming similar sales between mild hybrids and BEVs, the eventual cost of a BEV is probably only $2,000 to 3,000 more in the short term... but consumers make decisions based on differences $500 or even less.

I could do some more napkin math with engineering costs, unit costs, Bills of Material and amortization schedules... but you can be assured that automakers have done more than napkin math, and they are pursuing (as an aggregate group) both technologies.

The business cases for each of MHEV and BEV make different assumptions. I think it's great that different methods are being pursued. Eventually some automakers look very silly, but you can't tell which ahead of time.

This article is kind of weird. The push for 48 volt is mostly about secondary systems that need a lot of power and doesn't have much to do with "mild electric" or whatever it is implying.

The 21 HP direct boost mentioned in the article is going to be for performance rather than running on electric, the vehicle comes with a 1 Kw-h battery (~5 minutes at full output).

Whether it's for performance or economy depends on how it's tuned and how the engine is sized relative to the equivalent pure gas car. The attractive things about these mild hybrid systems is that they're cheap, they don't weigh much (so you're not hurting efficiency by hauling around a bunch of batteries and electric motors), and they give you a lot of what a full hybrid (non-plug in) offers.

If you're not plugging in, you don't need a big battery, because the hybrid only needs to store the energy from braking, and then deliver it when the car speeds up again. Even though the motor is only 21hp, it provides a lot of low-end torque, which lets you have good acceleration from a stop. This is something modern efficient gas engines are generally bad at, and the immediate acceleration in a lot of modern cars is even worse because of aggressive upshifting to save fuel. The electric motor papers over this deficiency, so you can be more aggressive about using a smaller engine with less low-end torque. The electric motor is also much more powerful than a normal starter motor, so it allows the car to more aggressively shut off the engine at stop lights and even when coasting, and turn it back on very quickly.

I buy what you are saying but that particular gas engine is 330 HP.

At one particular RPM.

A basic lack of fundamental understanding has caused the author to quickly fall into an expertly laid trap by marketers that seek to equate 48 volt systems with some sort of "green, hybrid, clean" spirit, leading to such bizarre statements like "The advantages of the new 48-volt systems are considerable. They offer fuel economy improvements of up to 15 percent"

Of course it's all about saving a few dollars on copper.

48 volt is what the industry calls mild hybrid. It allows for much better stop-start systems, which automakers get points for on EPA and European emissions testing.

It also allows for 48 volt super-charging and e-assist turbo, as well as a small amount of actual propulsion power (for instance to both start the engine and start the vehicle moving from a complete stop after the stop/start system has been engaged).

All of these systems fall under the term "mild hybrid".

The main advantage of a hybrid system is that the engine can run at a more efficient speed, which is independent of the power being delivered to the wheels. Say you are driving through a town, accelerating from stopped to 30mph then cruising in traffic, the engine will start shortly after you start to move, run at 3000rpm the whole time (or whatever is the most efficient speed), then cut off when you begin the cruise.

This is why automatic transmissions now often have 7 or more gears, as it allows the engine to run at a more efficient speed. A hybrid system in comparison is mechanically a lot simpler though: In a Prius the clutches, transmission and alternator have all been replaced by a simple mechanical device with fixed gears and two electric motors.

As for the battery the Prius has a 1.3kWh battery, but due to battery longevity (the system tries to keep it charged around 60%) only around 400Wh is actually available for use.

Some RV makers (both commercial [0] and DIY [1]) are already on 48V systems for their house batteries. Advantages: lighter wiring, can use a higher-voltage alternator that needs less heat dissipation. Note that the system in the Sprinter van is able to run the air conditioner at 100% duty for 8 hours, then allow the alternator to recharge the battery while continuing to run the A/C unit. Which is really impressive (and should be, for the $200k their RVs cost)

[0] https://www.youtube.com/watch?v=ofCcnxJxzeE

[1] https://www.youtube.com/watch?v=LV8fyzOTjsk

48 VDC will be good for amateur radio and car audio. Specifically the power amplifiers for both.

Noise from high power motors will be the bane of all operators, even when the car just drives by. However the junk yard will soon be a good place to source 48v generators.

I have a Nissan Leaf, but don’t operate mobile with it. The AM reception on its radio is fine, which I’ve heard is a problem in other EVs. Proper EMI mitigation can be done, but these companies may not spend the expense to do it.

Certainly a nice technology, but I think it is 10 years late. This kind of modernized eletrical design of combustion engined cars which enables mild-hybrids should have been standard 5 years ago. Going forward I cannot see a large future for combustion-engined cars at all - for a few few years, plugin hybrids migth be popular for those, who need a lot of range. But pure electric cars are now coming to the market in numbers. I don't think as stated in the article, that in 2025 only 5% of the vehicles sold will be electric, I would assume that number rather for 2020.

2020 is only 2 years from now, we’re nowhere close to that 5% number.

We are already at over 1% EVs in several important markets in 2017, and the EV sales numbers almost double every year, so 2020 probably won't see 5%, but should be pretty close to that - at least closer than 2025.

About a decade ago, wasn't there some interest in moving to 24 volt systems; nothing came of it due to switch failures, IIRC?

It's always been 48V. The problem has been a chicken and egg sort of thing. Its expensive to have both 12 and 48, but you can't go full 48 unless you can source all components for that voltage. Suppliers didn't really commit because there were no real customers.

How hard is it to put a DC-DC converter between the systems?

That's technically and conceptually easy but it costs a lot of money. One approach would be to put a large DC-DC converter on the car and run a 48V bus next to the 12V one. That's a very expensive device, needing to be perhaps a few kilowatts depending which devices are on 48V. It's also a lot of extra wiring, and the need to manage the extra variations in vehicle systems.

A second approach would be to put a DC-DC converter only in the devices that need 12V. But that's asking every supplier to change their design to accommodate 48V and the vehicle manufacturers will bitch about the cost increases. It'll be a moderate increase across a LOT of devices.

I'm working on a product that uses a highly integrated SoC that has all the regulators needed to connect it directly to the 12V line (it'll run on 6-20V or so). It even has an on-board 5V regulator for additional off-chip devices. The best thing for us would be if the same chip existed but could be connected directly to 48V (not sure what the range is on that) but from a silicon standpoint there is a big difference as well. Adding an external regulator would add cost, and invalidate some of the wake-up features and low current standby capabilities of the chip.

There are always "reasons". The only way to make the transition is for someone to make it happen, but again the suppliers who participate may be seeking premium pricing in the beginning because there aren't options.

http://archive.is/PYZ6G

There is a TLDR for this article, but as far as I can tell it cannot be derived from the article itself.

Will someone please explain why 48-volt systems are newly desirable in cars?

The higher voltage means lower amperage, which means smaller components and so on.

You can use thinner cables or have the same cables with less loss. As for actual component size it’s not that simple the higher the voltage the more clearance you need between conductors. So that is bigger airgaps in PCBs and higher rated insulation where gaps aren’t possible.

True, however, this can also be achieved by different insulating materials. In general, this is not really much of a problem; I agree, it needs to be considered.

You want a higher voltage to allow more power without huge cables due to high current. You want low voltage to avoid high-voltage safety issues which increase engineering, assembly, maintenance, and emergency costs.

The 48V system sits right at the boundary where you can deliver enough power for more auxilliary systems and some mild-hybrid techniques including frequent start/stop systems, brief acceleration boost, and regenerative braking. Yet, it's not high enough to require switching to a different electrical wiring safety paradigm.

This is similar to the reason power-over-ethernet runs at the voltage it does, or "low voltage" lighting exists for home applications where consumers are moving fixtures around and interacting with the wiring.

Thanks. I had always understood PoE's relatively high voltage as helping deal with the attenuation from long runs of thin copper wire.

I couldn't find a succinct explanation in the article why cars would benefit from this spot in between merely sending information (digital levels like 1.8, 2.2, 3.3, 5.0, etc.) and using electricity to do "real" work (120V, 240V, 480V, etc.). As I understand your answer, yes, they'd like to use it for work, but not so much as to stray into the range of orange wires. I got the second part of that from the article, but the first part was very hard to extract from it.

The specific number of 48V is presumably inherited from telecommunications:

https://electronics.stackexchange.com/questions/10851/why-do...

There's lots of telecom equipment that works at 48V, so it's presumably cheap and easy for carmakers to find suppliers.

There is practically no overlap between automotive and telecom. When I was at Tesla, we couldn't find even one manufacturer for common use subsystems (e.g. window motors). It's a chicken and egg problem for now. That's why 'everything' is still at 12V.

Canot say in the US, but in EU most trucks have (since I can remember) 24 V circuits, so there is anyway a long list of parts at 24 V manufactured.

All US military vehicles (even COTS ones with civilian counterparts like the CUCV and newer) use 24v. So do most heavy trucks.

I would have assumed it's just because it's a multiple of the current 12V (so you could use shortcuts like wiring 4 existing 12V lamps in series)

> Yet, it's not high enough to require switching to a different electrical wiring safety paradigm.

This is the tldr. Also once you ditch lead acid batteries for higher voltage hybrid batteries there no reason to keep using 12VDC.

The math bit is that the required wire size is related to the inverse square of the voltage. At the same time reduced current means smaller switches.

> Also once you ditch lead acid batteries for higher voltage hybrid batteries there no reason to keep using 12VDC.

Not really. The operating voltages for most of the actual electronic circuits will still be the usual 3V3/5V/12V - and converting from 12V to 5V/3V3 is easier than from 48V.

The switched loads I was thinking of are lights and motors. So 48V headlights, 48v blower motors, 48v door solenoids, etc. All these draw a fair amount of power and thus a lot of current at 12V.

It's not a total win especially for mechanical switches and relays which tend to be rated by their ability to interrupt power. And you do need higher voltage transistors.

Hmm. Definitely an interesting idea. Downside: manage to e.g. swap a 12V and 48V line and while an 48V component shouldn't be damaged by 12V, a 12V component which does not have a voltage regulator in front will get fried. It's nasty enough with a truck with dual voltage and trailers - a 24V trailer will at least be recognizable as the lights will simply have a dimmer light but a 12V trailer on a 24V truck means hours of swapping every goddamn tiny light bulb. Add a 3rd voltage into the mess and the potential for chaos multiplies...

From my reading of the article, 48V has been chosen to comfortably fit in under some regulation that defines 60V as “high voltage” which required special (read “more expensive”) components.

Keep in mind there’s no such thing as a “48V battery” or a “60V battery” if you have hard requirements on those voltages. If they’re using lithium polymer batteries, they’re rated at 3.7V per cell nominally, but will be at 4.2 or 4.3V when freshly charged - so a 48V lipo battery will probably be 13 cells and 48.1V nominal, but possibly as high as 55.9V fully charged (and one more cell in a 14 cell pack would hit 60.2V fully charged). Even lead acid batteries do the same thing, your 12V battery today is six 2V nominal cells in series, but you’ll commonly see 14V when it’s fully charged, so four current tech car batteries with a nominal 48V rating will have 56V fully charged.

I don’t think there’s anything “optimal” about 48V apart from it being as high as you can risk going using available battery technologies without bumping into the 60V “high voltage” regulations when your battery is fully charged.

I think it’s 50 VDC that is considered HV. Any equipment we buy at work 50 VDC and over requires NRTL (i.e. UL) certification. Also interlocks and safety boxes when working on circuit.

I have no actual knowledge here beyond the article itself:

“Why 48 volts? Settling on this level holds the electrical system under the 60-volt safety threshold of what is considered high voltage, where power cables must be orange and special connectors, costing 10 times as much, are required, said Mary Gustanski, chief technical officer of Delphi Technologies.”

I wonder if your 50VDC limit is driven by the same “typical peak voltages for common battery chemistry” margin of safety I’m speculating exists?

My understanding is that 50-60VDC is approximately the voltage at which electricity will arc through a person's chest (i.e. pass through their heart) if they hold a wire in each hand. That makes it unsafe to handle without special insulating equipment. That doesn't mean that 48V is safe necessarily though. If your hand is wet or even just a little sweaty, you can arc from pinky to thumb at ~48V.

When I was studying electrical engineering, we were taught that at 30V, current may be able to pass through the skin (depending on a number of conditions presumably). Once it's under the skin, you only need a tiny current through the heart to kill you.

At 48 volts you start running engine accessories off of the electrical bus instead of belts and pulleys. That means the engine isn't necessarily responsible for keeping everything under the hood running at all times. You might not need a water pump or the AC when the engine is off, except to keep the cabin comfortable. But you still need things like power steering and power brakes, stability control and other safety systems.

Also, you start getting economies of scale with all-electric car components that pull down the price of EVs over time, because they can use the same components.

> You might not need [...] the AC when the engine is off

Or you might! Driving a car with a start-stop system in the summer, it doesn't stay stopped at a red light for long until the AC wants to kick in and the engine comes back on.

Other people talk about the power, but there is also the fact that lots of actuators want more voltage to work with. So, any motor, servo, solenoid, relay, etc. works much better at higher voltages.

In addition, it is also much easier to convert between high voltages because your transformer/inductor connected to your semiconductor switches is so much smaller.

I thought the article did a good job of providing examples such as the VW group using 48V to provide 'sportscar' handling in their high end SUV offerings. Having 48V means that these things can be done.

Incidentally VW started out with 6V. Early Beetle and bus models have the joys of 6V electrics, usually replaced for 12V. The advantages are clear. And so it is with the move to 48V.

With turbocharging there is the problem of the wait, by simply using an electric motor then this problem is solved and a smaller engine can be used to deliver the same power and torque. This is not possible on 12V. But it is on 48V.

Stop-start is another area where 48V makes sense. Plus the HVAC.

Why don't Americans have kettles? Because of 110V, you need 240V to boil a kettle properly. Clearly 240V needs proper British Style plugs with convoluted earth arrangements and different wiring. So 48V is the happy compromise where there is no need for excessively regulated voltages. The 48V number is a multiple of 12V therefore four 12V batteries in series will get the desired potential difference.

Much like how we still have older USB ports for keyboards as well as the fancier ones that do power and multiple monitors, there are plans to run the two power levels in the wiring loom. All the lights on the dashboard will run at 12v for now as will the motors in the doors and seats. However, in time those will possibly become 48V too. Sportscars demand theatre with doors that open in cool ways, so imaginably there will be application here for 48V door locks. Imaginably regular VW hatchbacks will have similar 48V based door mechanisms soon with older models still using the 12V cheaper components.

Before these hatchbacks get 48V windows they will probably get the 48V ABS and the 48V power steering along with the 48V stop start.

None of this is a waste of time for companies such as Volvo and how they are engineering their platforms to take any one of many power supplies, hybrid or full electric. Volvo have no future plans for making ICE only vehicles and clearly the market is going towards this smart 48V tech whether the future is full electric or not.

Reply on news.ycombinator.com