Hacker News Reader

Comment by cyberax | Hacker News Reader

Back in 90-s your only realistic choice for batteries was nickel-metalhydride chemistry. And that's around 50 Wh/kg of energy density for the assembled battery pack, compared to 150 Wh/kg for Li-Ion.

With a huge number of issues: you can't fast-charge NiMh batteries, and the self-discharge rate is horrendous.

Then there is the question of the control circuitry. The first truly reliable high-voltage IGBTs started appearing only in mid 90-s. Before that, you in practice had been limited to synchronous permanent-magnet DC motors.

So I think that EVs really started happening almost immediately after all the enabling technology got there in the early 2000-s. The first Tesla Roadster was released in 2008, so perhaps you could have shaved 3-4 years from that, but not much more.

View on HN

NiMH was far superior to lead acid, superior to nicad, and much cheaper than lithium ion.

The biggest issue with NiMH was that it was heavily encumbered by patents owned by Ovonics / Cobasys, who were acquired by Texaco and then Chevron...who did everything they could to prohibit the battery tech from getting into electric vehicles, with (among other things) Texaco/Chevron having universal right to "veto" any proposed sale/use.

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

Sure. Oil companies tried to delay EVs, but I just don't think it really mattered. Practical EVs happened just about as soon as they became feasible, once the enabling technologies fully matured.

Even the best NiMh batteries were inferior to the worst LiIon batteries from the early 2000-s.

RAV4 EV was pretty much the best example of what is possible with NiMh, and it was pretty weak: 70hp motor, top speed of 80mph, range of 90 miles.

I think it did and still does matter. Slowing battery development also slowed other aspects of EV tech. To be fair, most auto manufacturers weren't and to some degree still aren't fully on board with even their own EVs.

If it wasn't for companies like BYD, Tesla, etc, pushing EVs forward, I think we'd still be stuck with compliance cars.

NiMH wasn't great though. I think the best was the NiMH EV1 at 142 miles of range.

Was battery technology actually slowed? When Tesla began mass-producing EVs, they were using off-the-shelf vendor batteries. It seems like if anything the battery tech was ahead of the EV wave, having been driven by thin and light ultraportable computers.

The two guys who created Tesla were the same guys who created the e-reader, which would become the kindle.

They took their knowledge from the kindle batteries and put it into a car.

P.S. Tesla has always used off the shelf vendor batteries. Panasonic has their separate section within the Tesla factories where they are mass producing regular looking batteries that are then assembled in series.

> I think it did and still does matter. Slowing battery development also slowed other aspects of EV tech.

I just don't see it. The first LiIon EVs had very little technology that was not off-the shelf. Roadster was quite literally a bunch of laptop batteries bolted onto a car frame. It's not that hard to make an expensive EV.

Ignoring improvements in battery tech, look at what BYD/Tesla are doing now with packaging/production/aero/etc.

https://www.drive.com.au/news/toyota-tesla-model-y-work-of-a...

Most auto production is optimized for ICEs. Even 20+ years ago, Toyota's engineers knew there was plenty to do in the EV space.

https://content.time.com/time/specials/packages/article/0,28...

The problem was/is that optimizing EVs will undercut ICE sales, so Toyota just ignored what their engineers said for decades to maintain the status quo and, like most legacy auto manufacturers, fell behind newcomers in the EV space.

The batteries are still pretty expensive.

The EV1 probably cost like $150,000+ in 2023 dollars just to build, so they've made lots of progress, but electric vehicles don't beat comparable conventional vehicles across the board, they have better performance and lower operating costs, and that's most of it.

The 2011 Nissan Leaf wasn't much better than that RAV4 EV: a quick search showed that the 2011 Nissan Leaf had a 107-hp motor, 92 mph top speed, and a 92-mile range.

> NiMH was far superior to lead acid, superior to nicad, and much cheaper than lithium ion.

Except on a couple accounts - Memory and Lifetime.

NiMH batteries have almost as bad as a memory effect as NiCD, and have a low lifetime cycle count compared to lead acid. I dont think you could put NiMH batteries in a production car, at any point, without designing the battery pack to be cheaply and easily replaced. NiMH batteries are far more of a consumable than even Lithium batteries are. There are also real limits on how much current an NiMH battery can deliver in a given period of time.

NiCD batteries because of the memory effect an inability to deal well with high current applications, made them unsuitable for a mass market electric car.

> low lifetime cycle count

Top of the line NiMh chemistries had lifetime cycle count better than many Li-Ion batteries, like 10% degradation after 2000 cycles. They also don't have the memory effect, but can exhibit a somewhat similar but fully reversible effect.

Really, the energy density was the main killer.

NiMH works fine in cars - Toyota has been using NiMH in many many generations of their hybrid cars like the Prius and they have held up very well. The problem is just the energy density once you need to maximize range.

I stand corrected, thank you!

It's really counter-intuitive to those of us who grew up with the NiMH->Li-ion transition in consumer electronics! When I learned that Toyota used NiMH in the Prius it was kind of a "mind blown" moment.

I could swear that the Prius was Lead Acid AGM batteries.

But like, most of the issues I had with NiMH was because the devices that had the batteries in them had no charge controllers, so all the effects I spoke about were true for those - NiMH with a battery controller, and without might as well be completely different beasts.

Yeah proper battery controllers, and modern chemistries (see eneloops on the consumer end - they're also a bit of a game changer)

> I dont think you could put NiMH batteries in a production car, at any point, without designing the battery pack to be cheaply and easily replaced.

RAV4 EVs have gotten 150k miles out of their original battery pack.

IGBT: insulated-gate bipolar transistor.

Allows for fast and efficient switching of high current circuits and are thus used (from my understanding) to modulate the power that goes to the motor.

This comment doesn’t seem to make a lot of sense. IGBT-based variable frequency drive is an improvement over GTO Thyristor technology, not a fundamental enabler for a VFD or ACIM.

The main issue are thyristor based inverters that are needed to produce high-voltage AC from high-voltage DC of the battery. They sucked, efficiency was low and the power limitations made reasonable performance unachievable.

Honestly in this day and age of keeping your stuff plugged-in every night the self-discharge of nimmies seems less of a concern. It's only a real pain when you put rechargables into a remote or a flashlight and then find out in a week it's already dead just from sitting and you have to pull the batteries out (which involves screws if it's a children's toy) to recharge them.

But the lower energy to weight ratio would be a showstopper for a car, yeah.

What's wrong with synchronous permanent-magnet DC motors?

You need a lot of rare-earths for the really powerful motors. The rare-earth production is never going to scale to replace all the ICEs. Also, permanent motors are less effective at higher RPMs, so you'd still need a gearbox (though it probably needs just 2 speeds).

Control circuitry without IGBTs kinda sucks. You're stuck with switched serial/parallel circuits, like on locomotives or trams. Regenerative braking would be possible, but very limited because you can't control current/voltage well enough not to mess up the battery.

None of this presents an insurmountable barrier, but it definitely limits the design space.

Wouldn't series wound DC motors have been an option? They're awful for a lot of reasons (low efficiency, brush wear, hard to do regen or even reverse without complex switching mechanisms, etc...) but they're simple and don't use permanent magnets and can be quite powerful.

I don't know when SWDC motors were invented but I would have assumed it was long before the 90's.

Uhm, that's what I said.

Before IGBTs, the main way you controlled a powerful DC motor was by having multiple windings that can be connected in parallel or series configurations, modulating the resulting current. This type of control goes way back to 19-th century, it was used on the first trams.

elihu is talking about using a universal motor to mitigate the rare-earths aspect: https://en.wikipedia.org/wiki/Universal_motor

Sorry, I misunderstood that. I was still thinking about controlling the power, my bad.

You said: "Before that, you in practice had been limited to synchronous permanent-magnet DC motors."

Series-wound DC motors don't have permanent magnets at all -- rather, they have two sets of windings (connected in series) that push against each other, so there would be no concerns about rare earths.

I don't know what a "synchronous" DC motor is, unless you meant to say "synchronous AC and permanent magnet DC."

Reply on news.ycombinator.com