Scrutable

Grumble wrote: ↑

Fri Jan 01, 2021 11:21 am

shpalman wrote: ↑

Thu Dec 31, 2020 9:53 am

Grumble wrote: ↑

Thu Dec 31, 2020 9:39 am

What the theoretical energy density of a lithium battery, and what’s the current state of the art in an as built battery pack? There’s room to at least double, and a lithium-air battery is a theoretical possibility for aeroplanes I’ve heard? Don’t know if anyone’s even got one working in a lab yet though.

https://www.theregister.com/2020/12/30/ ... batteries/

That’s a strange article, they do know that Quantumscape aren't claiming a new chemistry don’t they? It’s still Li-ion.

Unnoticed by me at the time it looks like Solid State Battery are further along than QuantumScape as they are already shipping samples for testing by others and are producing multi-layer cells not the single layers typical of initial lab testing.

https://cen.acs.org/energy/energy-stora ... wer/98/i40

... but the Solid Power ones are sulphide-based, and I can find recent articles which say QuantumScape has published data which says it "can be quickly charged to 80% of its capacity in 15 minutes" and that it's predicted to maintain a capacity of more than 80% after 800 cycles. But not where this data actually is, but can we have more money please?

shpalman wrote: ↑

Fri Jan 01, 2021 11:47 am

Grumble wrote: ↑

Fri Jan 01, 2021 11:21 am

shpalman wrote: ↑

Thu Dec 31, 2020 9:53 am

https://www.theregister.com/2020/12/30/ ... batteries/

That’s a strange article, they do know that Quantumscape aren't claiming a new chemistry don’t they? It’s still Li-ion.

Unnoticed by me at the time it looks like Solid State Battery are further along than QuantumScape as they are already shipping samples for testing by others and are producing multi-layer cells not the single layers typical of initial lab testing.

https://cen.acs.org/energy/energy-stora ... wer/98/i40

... but the Solid Power ones are sulphide-based, and I can find recent articles which say QuantumScape has published data which says it "can be quickly charged to 80% of its capacity in 15 minutes" and that it's predicted to maintain a capacity of more than 80% after 800 cycles. But not where this data actually is, but can we have more money please?

They “published” a presentation (which I linked to up thread) but I don’t know if they’ve published a paper. The slides in the presentation looked good, but the claims do need checking out and things they didn’t mention need examining. Some questions at the end were good - asking about operating pressure for example.

https://www.wind-watch.org/news/2021/01 ... ate-goals/

Bit worrying that our power contracts in certain circumstances can disincentivise wind production

I think this is good news...

According to a short piece in the January Modern Railways, DB Cargo UK is "experimenting with the use of Hydro-treated Vegetable Oil (HVO) as a locomotive fuel in a bid to decarbonise the company's operations." Of course, it does depend on their feed-stock - if they are using palm-oil, it may not be such good news.

They are currently working with a Class 67 loco and planning to 'move to live, on-track testing with their 67 and 66 fleets'. Judging by this 66 I saw at Feltham, got to be an improvement.


Filthy Class 66 by SteveH, on Flickr

There are also pieces on the rail Minister claiming the government are committed to electrification,some company getting funding for a 'rail solar farm', the chairman of Vivarail cautioning against enthusiasm for hydrogen technology and and Siemans Mobility is 'targeting the introduction of hydrogen trains in the UK'.

Electric car batteries with five-minute charging times produced

https://www.theguardian.com/environment ... ging-times

MartinDurkin wrote: ↑

Tue Jan 19, 2021 11:15 am

Electric car batteries with five-minute charging times produced

https://www.theguardian.com/environment ... ging-times

Yeah, although germanium sounds expensive and they haven’t made it work with silicon yet.

MartinDurkin wrote: ↑

Tue Jan 19, 2021 11:15 am

Electric car batteries with five-minute charging times produced

https://www.theguardian.com/environment ... ging-times

Somewhere near the end of this video they demostrate charging a battery for an electric scooter (range 50-70 km) from ~15% to 100% in 5 minutes using what seems to be a current of 130 A.

The battery was taken out of the scooter to charge it because as you can imagine it needs to be quite seriously cooled while that is going on.

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

Grumble wrote: ↑

Tue Jan 19, 2021 11:48 am

MartinDurkin wrote: ↑

Tue Jan 19, 2021 11:15 am

Electric car batteries with five-minute charging times produced

https://www.theguardian.com/environment ... ging-times

Yeah, although germanium sounds expensive and they haven’t made it work with silicon yet.

Germanium is a commonly used seminconductor, used all over industry for RF and power electronics. It's unlikely that the material costs for it exceed the manufacturing costs.

Reasons to be optimistic about these batteries:
Really fast charging
Real working battery
Made in a factory not a lab

Reasons to be cautious:
They claim will be ready for market in a few years - clearly still in development and they want to change the chemistry from this demo
Needs serious cooling, although it looked like only air cooling so maybe within scope of battery temperature management already in cars?
Another new charging standard required? Maybe not though, maybe they’re staying at the peak for longer than current chargers but aren’t that much higher peak.

The scooter in the demonstration video is one of these: https://torrot.com/en/motorbikes/5/133/muvi

It has a 3 kW motor and two 48 V 31.9 Ah batteries, with a range of the order of 100 km (although the presentation in the video says 50-70 km). For a 100 kW motor in a car with 500 km range we'd be talking about two orders of magnitude MOAR.

I note that for a car battery they're not saying "5 minutes" they're saying "5-10 minutes".

I also note than the StoreDot battery seems to be made of lots of cells with air gaps between them to allow cooling air flow.

Still, everyone has to start somewhere and at least this thing seems to be working. They've certainly done better than the few hours it officially takes to charge the ordinary batteries in that scooter.

shpalman wrote: ↑

Tue Jan 19, 2021 12:03 pm

MartinDurkin wrote: ↑

Tue Jan 19, 2021 11:15 am

Electric car batteries with five-minute charging times produced

https://www.theguardian.com/environment ... ging-times

Somewhere near the end of this video they demostrate charging a battery for an electric scooter (range 50-70 km) from ~15% to 100% in 5 minutes using what seems to be a current of 130 A.

The battery was taken out of the scooter to charge it because as you can imagine it needs to be quite seriously cooled while that is going on.

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

130A sounds tiny.

We make lots of devices that have higher DC loading - even in a 3.3x3.3 mm footprint

When talking about vehicles they usually use kW and kWh for batteries. The current fastest chargers (no pun intended) run at 350kW max. I don’t know what ampage that is though.

Grumble wrote: ↑

Tue Jan 19, 2021 4:44 pm

When talking about vehicles they usually use kW and kWh for batteries. The current fastest chargers (no pun intended) run at 350kW max. I don’t know what ampage that is though.

In North America DC level 2 charging is is 400kW at between 50-1000V, so 400A minimum. Which needs some pretty heavy copper/aluminum conductors.

shpalman wrote: ↑

Tue Jan 19, 2021 12:03 pm

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

That's the way they talk about charging LiPos for RC plane use.

dyqik wrote: ↑

Tue Jan 19, 2021 4:53 pm

Grumble wrote: ↑

Tue Jan 19, 2021 4:44 pm

When talking about vehicles they usually use kW and kWh for batteries. The current fastest chargers (no pun intended) run at 350kW max. I don’t know what ampage that is though.

In North America DC level 2 charging is is 400kW at between 50-1000V, so 400A minimum. Which needs some pretty heavy copper/aluminum conductors.

I can imagine. Our switch room at work has 2 1600A cables from the substation and we have 400A cables in a few places internally. Takes some bending.

jimbob wrote: ↑

Tue Jan 19, 2021 4:37 pm

shpalman wrote: ↑

Tue Jan 19, 2021 12:03 pm

MartinDurkin wrote: ↑

Tue Jan 19, 2021 11:15 am

https://www.theguardian.com/environment ... ging-times

Somewhere near the end of this video they demostrate charging a battery for an electric scooter (range 50-70 km) from ~15% to 100% in 5 minutes using what seems to be a current of 130 A.

The battery was taken out of the scooter to charge it because as you can imagine it needs to be quite seriously cooled while that is going on.

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

130A sounds tiny.

We make lots of devices that have higher DC loading - even in a 3.3x3.3 mm footprint

I know for a fact that 130 A is enough to make a short coil of tantalum wire with a 2mm diameter glow white hot and emit so many electrons that you can then pass 50 A through an argon plasma (at ~0.01 mbar) with a negative bias of only about 20 V on it because that's how our deposition system works.

We have another deposition system in which there are three Ta filaments and two of them are in parallel on the same supply; we couldn't put all three on the same supply because it was unstable and they'd immediately burn out.

Not that I'm joining in with the "my cable is thicker than yours" penis-waving.

You know you want to.

Come here for the science and technology, stay for the willy jokes.

Sadly that Class 66 is going to chuck out just as much nasty smoke running on HVO as it does on petroleum diesel, so it will actually be worse for your health because Spoiler:

shpalman wrote: ↑

Tue Jan 19, 2021 6:14 pm

jimbob wrote: ↑

Tue Jan 19, 2021 4:37 pm

shpalman wrote: ↑

Tue Jan 19, 2021 12:03 pm

Somewhere near the end of this video they demostrate charging a battery for an electric scooter (range 50-70 km) from ~15% to 100% in 5 minutes using what seems to be a current of 130 A.

The battery was taken out of the scooter to charge it because as you can imagine it needs to be quite seriously cooled while that is going on.

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

130A sounds tiny.

We make lots of devices that have higher DC loading - even in a 3.3x3.3 mm footprint

I know for a fact that 130 A is enough to make a short coil of tantalum wire with a 2mm diameter glow white hot and emit so many electrons that you can then pass 50 A through an argon plasma (at ~0.01 mbar) with a negative bias of only about 20 V on it because that's how our deposition system works.

We have another deposition system in which there are three Ta filaments and two of them are in parallel on the same supply; we couldn't put all three on the same supply because it was unstable and they'd immediately burn out.

Not that I'm joining in with the "my cable is thicker than yours" penis-waving.

A high performance CPU can draw about 80A at 100% load. e.g. an Intel i9 can draw 125W at 1.52 V.

dyqik wrote: ↑

Tue Jan 19, 2021 7:50 pm

A high performance CPU can draw about 80A at 100% load. e.g. an Intel i9 can draw 125W at 1.52 V.

That's why you need to attach something like this to it

shpalman wrote: ↑

Tue Jan 19, 2021 8:31 pm

dyqik wrote: ↑

Tue Jan 19, 2021 7:50 pm

A high performance CPU can draw about 80A at 100% load. e.g. an Intel i9 can draw 125W at 1.52 V.

That's why you need to attach something like this to it

I've got a 2.9mV, 50 μA device downstairs being cooled by a 150kg, 10 kW air conditioner.

I also use a cryostat,* although it only goes down to 1.6 K. I seem to remember being told that most of the heat is taken away by the first stage of the cold head which is somewhere around 40 K, while there's much less cooling power from the second stage which reaches 4 K.

The superconducting magnet in that takes about 100 A too.

(* - well, it's my master's student using it these days)

shpalman wrote: ↑

Tue Jan 19, 2021 6:14 pm

jimbob wrote: ↑

Tue Jan 19, 2021 4:37 pm

shpalman wrote: ↑

Tue Jan 19, 2021 12:03 pm

Somewhere near the end of this video they demostrate charging a battery for an electric scooter (range 50-70 km) from ~15% to 100% in 5 minutes using what seems to be a current of 130 A.

The battery was taken out of the scooter to charge it because as you can imagine it needs to be quite seriously cooled while that is going on.

They say they're charging it at a rate which is 10 times the capacity of the cells. I'm not sure how you'd make the dimensions of that equation balance, maybe it means for example charging a 10 Ah cell in 6 minutes at 100 A.

130A sounds tiny.

We make lots of devices that have higher DC loading - even in a 3.3x3.3 mm footprint

I know for a fact that 130 A is enough to make a short coil of tantalum wire with a 2mm diameter glow white hot and emit so many electrons that you can then pass 50 A through an argon plasma (at ~0.01 mbar) with a negative bias of only about 20 V on it because that's how our deposition system works.

We have another deposition system in which there are three Ta filaments and two of them are in parallel on the same supply; we couldn't put all three on the same supply because it was unstable and they'd immediately burn out.

Not that I'm joining in with the "my cable is thicker than yours" penis-waving.

Yes, I was specifically thinking in terms of currents for cars. Starter-motors draw hundreds of amps - although we tend to use 5x6mm footprint devices for those.

The 160A 3.3x3.3 is probably for high performance computing - tbh, I can't recall all the applications, especially as I'm now more interested in the GaN power devices, with EVs as a key market.

shpalman wrote: ↑

Tue Jan 19, 2021 8:47 pm

I also use a cryostat,* although it only goes down to 1.6 K. I seem to remember being told that most of the heat is taken away by the first stage of the cold head which is somewhere around 40 K, while there's much less cooling power from the second stage which reaches 4 K.

The superconducting magnet in that takes about 100 A too.

(* - well, it's my master's student using it these days)

Mine really is being cooled by a window AC unit, cooling a water chiller, cooling a compressor that's running a pulse tube cooler. Because we can't keep the building's cooling water supply running in winter in Massachusetts.

So we're probably running about twice as much electrical power as we should be.