Hacker News Reader

Comment by cat_plus_plus | Hacker News Reader

Wouldn't the best type of ICE be a flexible fuel backup generator to extend range of otherwise electric vehicles? Which should call for altogether new designs, for example optimization for single RPM and torque to spin a dynamo rather than having to handle rapidly changing load of direct driving. Then over time we can find cleaner fuels and, if one day we have batteries that charge in 15 minutes and something to charge them from eveywhere, ditch it all together without having to revamp every car's design.

View on HN

This is sort of what Edison Motors[1] is doing. They have built what is essentially a diesel electric locomotive in truck form, where they have a CAT diesel generator charging two battery sets which then power two electric axles in the rear, two traction motors per axle. The generator runs at a constant efficient speed (when needed) and has no mechanical connection to anything other than the alternator.

[1]: https://www.edisonmotors.ca

I once came across an ancient genset based on the original Willy's jeep engine, it was an interesting little device, the cam rockers were very sharp because it only had to run at a very narrow band of RPMs and this made the valve timing very tight.

It's neat how given a chance to optimize instead of just sticking the standard camshaft in there the engineers immediately went for the efficiency gains.

Does it really use a dynamo? I would have expected an alternator with an electronic rectifier instead.

Oh yes you're probably correct. Their website doesn't say and I was, until now, not familiar with the difference.

That's called a plug-in hybrid (as opposed to normal hybrids which switch from electric to combustion operation depending on power output needs). They're quite good and clever vehicles, but their name has caused a gigantic amount of confusion and problems for them, and they seem to be in serious trouble because of that.

I think they are actually talking about a series hybrid (https://en.wikipedia.org/wiki/Hybrid_vehicle_drivetrain#Seri...), which is different than a plug-in hybrid (though it would be somewhat crazy to make a series hybrid that was not also a plug-in hybrid).

.. as to why those never caught on: I have no idea. Plug-in series hybrids seem like a brilliant solution to most issues with both ICE cars/parallel hybrids and electric cars. My honest theory is that Tesla started pushing all-electric cars much earlier than anyone else would have, and that caused most people to skip plug-in series hybrids.

The problem is that by going with that, you lose many of the benefits of an EV, like not lugging around several hundred pounds of cast iron, or having oil to change.

Totally fair. My thinking is that means you could lug around a lot less battery weight (ie, make the all-electric range 60 miles or so), basically just enough to complete the average amount of driving a car does in a day and operate on the assumption that most cars would get plugged in every night. In return, you'd get the ability to use the existing energy distribution infrastructure built around roadways (ie, gas stations) whenever you wanted to drive your car further than it could go on a charge. I'd also assume that the ICE/generator combo could be tuned to operate entirely within whatever window results in highest efficiency for the ICE engine, which should be a notable improvement on the way ICE efficiency varies by speed in traditional cars.

I'm not saying that there aren't tradeoffs, just saying that it feels like an interesting idea. Enough so that I'm not at all sure why it didn't take off.

If the engine is able to always operate at peak efficiency, there is little reason to bother with batteries beyond a bit for handling transient peaks.

I guess what it comes to down for me is that the hybird as you described, while it has advantages over pure ICE, it's sorta the worst of both worlds - all the moving parts, all the complexity, the fire risk from both li-ion AND gasoline, more expensive (since two drivetrains, essentially).

A constant RPM motor for an extended period isn't a super enjoyable sound, either.

> If the engine is able to always operate at peak efficiency, there is little reason to bother with batteries beyond a bit for handling transient peaks.

My thinking here is that most of these theoretical cars would spend a good part of their life only operating on battery power - basically, all of the home-to-work-to-supermarket-to-home trips would be pure electric in my mind. So much so that I've wondered if the ECU would need to occasionally run the ICE for a bit just to make sure it doesn't ever spend so much time without being turned on that entirely different problems arise.

> it's sorta the worst of both worlds

I get this. This idea is very much the best of both worlds/worst of both worlds depending on your point of view.

> more expensive (since two drivetrains, essentially)

I'm not sure about this. I think the right comparison here is to imagine a battery that has (say) 25% of the range of a Model 3 battery, but is (say) 30 or 35% of the cost. Based on EV battery replacement cost, could you buy a cheap 3-cyl ICE engine and generator for that? Intuitively this feels possible, though I'll admit I don't have access to the data to be able to say, for sure how these costs would compare at scale.

> A constant RPM motor for an extended period isn't a super enjoyable sound, either.

The sound should be similar to the sound a pure ICE car makes on the freeway when on cruse at a constant speed and not changing gear. Yeah, you'd need to tune the exhaust to make sure the sound is pleasing, but this feels like a solvable problem also.

(As an aside, sorry if it feels like I'm hassling you. I'm honestly not, this is just an idea that has lived in the back of my brain for a bit and I'm enjoying the discussion.)

An ICE places a lot more constraints on design and manufacter...since it has to be behind a firewall, it's large, it has all kinds of fluids that need routing out... it's a nightmare. For comparison a typical EV battery is just a bunch of bricks inside the skateboard base.

Ah good to know about the name "Series Hybrid", I've only ever heard people call a Series Hybrid a "Plug-In Hybrid" even in somewhat official and educational contexts, but I suspect that's just a symptom of the problem I was complaining about.

For sure. And, at the risk of being pedantic, a plug-in hybrid can still be a parallel hybrid - series vs. parallel is about how the ICE power gets to the ground, whereas plug-in vs not is about the existence of a way to charge the hybrid battery packs separate from the ICE engine or regenerative breaking (or whatever else that's not electric power from a wall outlet).

And most plug-in hybrids are (parallel). The "plug in" part is usually referring to (even if technically wrong) the electric subset being able to reasonably move the car alone, instead of being just a substitute to the ICE in near-idle cruise and an assistance at peak load. The combination of reasonable utility electric-only and serial is usually referred to as "electric with range extender" instead of hybrid.

Nissan make a series hybrid (which they call "e-power"), although with a very small (2.1 kWh) battery, so it doesn't work as a plug-in hybrid, and sadly it is apparently rubbish:

https://www.arenaev.com/nissan_qashqai_epower_confirms_what_...

I don't know what their engine is; if this is a normal car engine repurposed to drive a generator, this is not a great test of the idea.

I don't know what fuels it takes, but the "optimized for a single RPM and torque to spin a dynamo" is pretty much describing Mazda's MX-30 R-EV which does this with a rotary engine.

https://www.autocar.co.uk/car-review/mazda/mx-30-r-ev

Won't be coming to the US though, and the EV version is dropped here as well. Now that they have PHEVs in the CX-90 and upcoming CX-70. Previously the MX-30 was available in California as a compliance car.

Yeah, and the Rotary is actually really good for this job.

The problem with doing that in the purest sense is that mechanically connecting the engine to the wheels is always more efficient than turning a dynamo, then running the power through an inverter and a motor. Honda and Mitsubishi actually do pretty much what you're saying, but they both couple the engine to the front wheels at high speed. And at low speed, the efficiency in hybrid mode is not that great. It is a very simple, elegant design though, so I'm still fond of it.

I agree that the path from engine mechanical output to wheels certainly seems less efficient. We might be wrong about that, underestimating the losses in gear sets, but that's not my point and I also suspect that in that case we would hear more about gear set cooling solutions/issues.

But the path from fuel to engine output can be made far more efficient by designing the engine for a very narrow band of usage conditions instead of the crazy wide range from cruising at near idle to showing off what you paid extra for. Diesel trains have been diesel-electric for almost a century.

> if one day we have batteries that charge in 15 minutes

We're actually pretty close to that already. the Ioniq 6 Long Range adds 340km of range in 16 minutes. [0] The Porsche Taycan Plus adds 330km in 17 minutes [1]

[0] https://ev-database.org/car/1718/Hyundai-IONIQ-6-Long-Range-...

[1] https://ev-database.org/car/1924/Porsche-Taycan-Plus#chargin...

How much battery life do you lose if you always fast charge? I thought batteries strongly prefer the kind of slow overnight charging that needs a garage.

Tesla used to tell users to avoid overusing the fastchargers, but they no longer include such language on the site.

A recent study by Recurrent Auto which looked at data from 12,500 Tesla vehicles said that there is "no statistically significant difference in battery degradation between Tesla Model 3 vehicles that charged at least 90% of the time using Superchargers and Model 3 vehicles that charged no more than 10% of the time using Superchargers."

Other brands might have different technology and might degrade differently. Tesla's aren't the fastest chargers, so perhaps the data for Ioniq or Porshe would be different.

[0] https://cleantechnica.com/2023/10/01/tesla-supercharging-not...

[1] https://www.tesla.com/ownersmanual/model3/en_us/GUID-7FE78D7...

If a series hybrid were optimal in a sports car or performance car, then I would expect to see them used in the 24 Hours of Le Mans and 24 Hours of Rolex, where maximizing fuel economy without sacrificing speed is essential to winning the race. Instead, they mostly use the hybrid system for energy recovery, and it sounds to me like the new RX-9 might be more similar to those systems than what we are used to seeing on the street.

FWIW, I have endurance raced a hybrid vehicle -- a stock Toyota Prius with a roll cage. Under racing conditions, it mostly keeps the engine at a constant 4000 rpm as it struggles to keep the battery charged. Racing a Prius is a little odd, because you end up doing a few laps to charge the battery, then strategically executing a pass once the battery is charged enough to give you power. In a series hybrid, you would only get power from the electric motor(s) rather than getting the combined power of the MG and ICE.

> then I would expect to see them used in the 24 Hours of Le Mans and 24 Hours of Rolex

Powertrains in these leagues are heavily, heavily regulated. Outside of baja, there haven't been "unlimited" high budget racing leagues since like the 70s or 80s.

Hence crazy 80s Group C Le Mans cars (before they got banned) got similar lap times similar to modern Le Mans prototypes with much worse tyres, and the 70s Porsche 917 still holds the lap record at Talladega, by a wide margin, because such a nutty unregulated engine never lapped it again.

Also, the performance targets for a pure racecar are totally different than a sports car or even a luxury supercar. They are either at high RPM full throttle, full brake, or in a turn 24/7, with no concern for any other kind of load (like puttering down a highway at low RPM) or any kind of sane maintenance schedule. And weight is absolutely critical.

No (I think), unless your are just talking about hybrid vehicles which have been in use for a very long time. You are proposing a vehicle with redundant drive systems and two primary sources of fuel. This only makes sense in very specialized cases where some odd sort of asymmetrical failure mode is worth the extra complexity and weight.

The only reason these exist on the market today in the form of Plug-in Hybrids is that power generation in a hybrid vehicle requires some intermediary storage to be effective. Manufacturers can add a plug and allow the vehicle to run directly off the intermediary storage for a very short distance (usually less than the distance provided by single gallon of fuel), however as far as I know this is basically done as a form of greenwashing. Given the ranges allowed, you might as well ride your bicycle.

That sounds like the BMW i3 with range extender

https://en.wikipedia.org/wiki/BMW_i3

Isn't this what a locomotive engine is? A giant diesel generator to power the electric motors that actually drive the train?

I like this idea, especially the flex fuel aspect. Something purpose designed to easily convert to run gasoline, natural gas, pure ethanol, you name it. Though it might have to have some sort of cycle to ensure you don't have 2 year old gas in your tank when you finally need to charge the batteries on the road.

I have a plug-in hybrid (Toyota RAV4 Prime). It doesn't support flex fuel, but it's an awesome implementation- I typically drive for months before having to refuel the 10 gallon tank, with most of my driving on the battery. There's no dynamo, it switches from charging to discharging smoothly.

"Best" is a subjective term, especially when talking about a performance car.

However Mazda has started releasing PHEV designs as you design over the past few years, with the RX-9 being a performance optimized variant.

that's also would allow broader adoption in countries where charging infrastructure is not yet ready. eg. in my country we have very little charging stations because there is just not enough demand, and we do not have much electric cars because not many people want to pay that much for car that you have no way to charge (especially when you do not have garage)

FWIW, Mother Earth News offered plans for just such a car, back in the late 70s. It might have even been a lawn mower engine they used.

Aren't you describing hybrid cars?

He's describing a particular kind of hybrid, where the ICE acts purely as a power generator. Most hybrids don't work this way. They tie the output shaft of the ICE into the powertrain directly, so it can operate the wheels by itself if needed.

This setup allows them to make the electric motors much smaller and let the triple-rotor engine lay down power for acceleration, while the small motors handle the cruising.

The MX-30 R-EV operates the way you describe (https://www.caranddriver.com/news/a42485033/mazda-mx-30-r-ev...)

just wanted to add interesting detail: hybrid cars use slightly different engine construction that is more efficient but lacks power in low rpm making them useless in regular cars albecause you would not be able to move from full stop, hybrids just use electric motor to solve that issue

> He's describing a particular kind of hybrid, where the ICE acts purely as a power generator.

That's pretty much how diesel railway locomotives work, they just ditch the batteries and throttle the engine to control power. For regenerative braking (that isn't) they just have giant resistors with fans.

Reply on news.ycombinator.com