Cell Level Fusing: Discussion

No mine was a enclosure, so it might be different for your trucks.

His last response was 25d ago.

The “recommended current rating” you find on most wire is based on the requirement of low voltage drop and low heat generation over a long distance - many tens or hundreds of feet. Doesn’t really apply to a 12mm long piece of wire acting as a fuse.

Same reason wire for RC applications is used with far higher current than the same size wire used for houses - 20A for 12AWG house wiring, vs 60-80A for 12AWG wires used in EVs. Our 12AWG isn’t 3-4x better than house wiring, we just don’t care as much about voltage drop or heating, because our wires are a couple feet long at most, vs tens or hundreds. (Also, we use silicone insulation, which is good for up to 200 degrees C.)

1 Like

Yeah same for me, was told week or so wait time, its been almost a month.

Here’s my take on cell-level fusing.

As p-group size increases cell fusing makes more sense. Below 4p fusing is not worth doing. Above 4p fusing is absolutely worth doing. At 4p its a coin toss, depending on your esc settings.

The above statements are in relation to 30q batteries only.

I made a whole calculator for it. Its on the github.

12 Likes

Could someone explain to me the benefits of using narrow nickel VS copper wire for cell level fusing? I know @Kaly uses nickel, whereas most others use copper.

One guess I had was that copper, although having less resistance and thus producing less heat, might be worse at transferring heat away, therefore being a better fuse. But then I remembered copper is used for heat sinks so probably not :stuck_out_tongue:

I believe that the benefit is that the nickel can be easily spot welded

Edit: if anyone has a link to a chart showing nickel size to fuse rating that would be awesome

2 Likes

I would love a chart like that as well. For reference, here is the chart but for copper wire:
https://www.powerstream.com/wire-fusing-currents.htm

1 Like

I honestly believe “cell level fusing” using narrow gauge wire just adds excessive unnecessary internal resistance, i.e. tripling the internal resistance of a battery pack. The odds of an individual cell failing short circuit are incredibly low, unless they are reused/refurbished cells. In the event a short between layers of a battery pack occurs, you will have a fire anyway due to the extreme arcing that occurs.

Whats wrong with using fuses dedicated for short circuit protection instead of narrow gauge wire? Those are designed to have very low resistance while also providing arc supression.

Proper fuses are not simply designed to open at a specific current, they are designed to supress arcs and withstand well specified standoff voltages while minimally impacting the circuit during normal operation. Attempting to create a fuse using narrow gauge wire just feels dangerous.

5 Likes

@Gamer43 That is an excellent argument in an ideal world.

Riddle me this: Why does Tesla, the world’s largest electric car manufacturer, which uses 18650 and 2170 cells very similar to the ones in our batteries by the thousands per car, use cell level fusing on every cell in every single pack they manufacture?

I postulate that they have a very good reason, otherwise they wouldn’t be doing it on an industrial scale.

2 Likes

Are they doing it with proper well designed fuses? Or are they doing with ballmark estimated narrow gauge wire?

Tesla has had issue with quality control and battery fires recently…

They are using narrow gauge aluminum wire.

Well then, just like their rockets, they WILL blow up at some point, or the narrow aluminum won’t do the job a proper a fuse would do. Unless an engineer specifically tested and designed the wire in such way that it acts like a proper fuse.

You’d think the failure rate would be higher, but then you’d also have to remember just how damn lucky NASA got each time.

Let me be clear, narrow gauge wire CAN be designed as a proper fuse, BUT significant engineering processes must be done to ensure it behaves within the specifications of the application.

By engineering I mean more than just ballpark estimating the onderdonk fusing current.
The ends of the fuse need to be well secured and the way the fuse opens needs to be well defined.

Some intuition tells me Tesla has done their homework.

Don’t get me wrong, you can totally substitute actual fuses for aluminum wires. But if aluminum wires are cheaper and good enough, why spend more money on something that does the same job?

Tesla batteries also have a big “proper” fuse on the output of the battery. But that serves a distinctly different function than cell-level fuses.

1 Like

The wires are attached at very specific points and bent a specific way, they probably went through the processes to qualify and ensure they act as fuses.

The losses in other parts of an electric car probably make the additional losses here negligible. Unfortunately for ebikes, escoot, and esk8, this not the case. In my scooter, half of my electrical losses are from the battery’s internal resistance.

Yup.

My fuses (ahem wires) are also attached to specific points and bent a specific way. Probably not as good as theirs because I did mine by hand, but then my battery, while bigger than most esk8 packs, only contains 140 cells, rather than ~6000, so the cost of failure is slightly lower.

I did do my math and test my setup before starting.

I dont think you realize just how rigorous the entire process of qualifying a fuse is…

Not to mention the precision in the mass assembled battery is critical.

Oh, I do. But I’m not designing a commercial automobile, so I can happily not give a damn.

1 Like

Well, if you choose to sacrifice 20% of your range for minimal benefit, that is your choice :).