Hi! Long time lurker, current TSLA holder, potential Tesla owner.In order to drain 1500 amps from the batteries, Tesla invented an active fuse that monitors the current from the batteries with computers and cuts the wires by pyrotechnics if things gets out of control. I love the inventive step of the design, but the engineer in me gets concerned. In my mind, a fuse is a low-tech last resort that saves the day when the high-tech gizmos fail. Now this negative use case is depending on dedicated batteries, firmware and microcomputers to work properly. Maybe I am old-fashioned as I started driving when cars were all analog (no ABS or ATC), but replacing the fuse, the most primitive lifeline on earth, with electronic brains, got me worried.I am curious to hear your take on this.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
My question is what is the primary protection, and what is the secondary protection?
As viewed from the Cell stack all the way to the Power Inverter. The battery pack still has it's own protection, and doesn't rely soley on this "fuse".
I believe this "fuse" is simply secondary protection to make sure the Power Inverter doesn't pull more than 1500 amps from the battery. Where the Battery itself it has it's own protection to make sure nothing pulls more than X amount of current.
What they were probably running into is traditional means of protecting from over-current weren't fast enough or we're controllable enough to get what they wanted.
I do share you're concern about what I call e-Fuses. I wouldn't call them true fuses because their IC's and rely on some kind of coding. But, manufactures are forced to use them because to my knowledge there really isn't any precise way of doing a completely passive fuse. It's all somewhere around X amount of current/temperature the thing will open. Whether it's a PTC or a regular fuse.
As viewed from the Cell stack all the way to the Power Inverter. The battery pack still has it's own protection, and doesn't rely soley on this "fuse".
I believe this "fuse" is simply secondary protection to make sure the Power Inverter doesn't pull more than 1500 amps from the battery. Where the Battery itself it has it's own protection to make sure nothing pulls more than X amount of current.
What they were probably running into is traditional means of protecting from over-current weren't fast enough or we're controllable enough to get what they wanted.
I do share you're concern about what I call e-Fuses. I wouldn't call them true fuses because their IC's and rely on some kind of coding. But, manufactures are forced to use them because to my knowledge there really isn't any precise way of doing a completely passive fuse. It's all somewhere around X amount of current/temperature the thing will open. Whether it's a PTC or a regular fuse.
Also don't forget individuals cells have cell level fuses based on how the individuals cells are "wired" to one of the several battery packs in the car.
The only thing I find concerning here is that if the car pulls more than 1,500 amps and the wires are cut are you pretty much stranded? It seems there would be no way to reset that fuse and the car would need to be towed...
The only thing I find concerning here is that if the car pulls more than 1,500 amps and the wires are cut are you pretty much stranded? It seems there would be no way to reset that fuse and the car would need to be towed...
Thankfully along with the "Fuse" that $10K also buys you "free for 8 years and unlimited miles" towing.
It hasn't been announced yet, but we can totally see it coming.
I should stop making jokes about the "fuse" because most of you probably don't share my complete disdain, and feelings of how ridiculous it is.
mikeash
Active Member
I imagine the computer hardware and software in this active fuse is pretty simple. It's basically going to be running a tiny infinite loop going: while(forever) { if(current > 1500A) explode(); }.
And heck, they can probably even test the actual production items by running a ton of current through them without the pyro charge hooked up, and making sure it fires as it's supposed to. Try that with a normal fuse!
It's not like fuses are totally foolproof. Oops, the machine had a bad day and your fuse was manufactured a little too thick, and nobody noticed because we can't do non-destructive testing.... And the whole reason for this fuse is because once you pass a certain level, a normal fuse doesn't even work anymore, because the massive amount of power being passed through it is enough to vaporize bits, ionize them, and use them to keep on carrying current even after the fuse tries to open. Complication in one form or another is unavoidable.
EarlyAdopter
Active Member
No different than the current fuses. When they blow (burn through), you're stranded until the pack can get opened up and a new one installed. They are not resettable.
I'd mentioned elsewhere, but my bet is that the new smart fuse is in series with a much larger "dumb" fuse. The 85kWh D pack's I disassembled from an S85 and P85 both had 630A main fuses.
I'm actually just wondering where this new fuse will get the power to recharge it's small battery. There is no accessible data or power connection in the fuse section of the battery pack, so, it would have to get it's power somehow from the power flowing through the fuse, which I guess is possible when we're talking about hundreds of kW flowing through at times.
I'm actually just wondering where this new fuse will get the power to recharge it's small battery. There is no accessible data or power connection in the fuse section of the battery pack, so, it would have to get it's power somehow from the power flowing through the fuse, which I guess is possible when we're talking about hundreds of kW flowing through at times.
EarlyAdopter
Active Member
It's important to note that traditional fuses aren't perfect either and can fail to blow during overcurrent when they are needed or blow prematurely when they are not needed.
Also, traditional fuses have response times (time from fault to time to trip) measured in seconds. E-fuses are in milliseconds. That's a huge amount of dangerous current that a traditional fuse will let through before blowing that an e-fuse won't.
E-fuses don't have to be perfect, just better than traditional fuses. And believe me, that's a pretty low bar to beat.
- - - Updated - - -
The main pack has a 12v feed-in, does it not, to power the BMS? The smartfuse could draw from that. Part of the upgrade could be running a line to that section of the pack.
SebastianR
Active Member
First, I'm certainly not an electrical engineer.
My understanding of the fuse is the following: individual cells in the pack have simple thermal fuses that prevent at cell-level that failing cells in the pack cause a thermal runaway in neighboring cells (and thus a fire). The fuse they changed is not for that purpose. Quite frankly, at these currents I'm amazed they still used thermal fuses: my (admittedly very limited) understanding of power transportation / power network technology (i.e. a situation where you are dealing with high currents and where you need fast acting fuses at very precise loads + where you care about transmission losses) is that they are using similar active fuses for quite a while.
I think it is an interesting topic and would like to learn more - so please add to this (and let me know where I'm inaccurate).
My understanding of the fuse is the following: individual cells in the pack have simple thermal fuses that prevent at cell-level that failing cells in the pack cause a thermal runaway in neighboring cells (and thus a fire). The fuse they changed is not for that purpose. Quite frankly, at these currents I'm amazed they still used thermal fuses: my (admittedly very limited) understanding of power transportation / power network technology (i.e. a situation where you are dealing with high currents and where you need fast acting fuses at very precise loads + where you care about transmission losses) is that they are using similar active fuses for quite a while.
I think it is an interesting topic and would like to learn more - so please add to this (and let me know where I'm inaccurate).
Just to be clear: I am not voicing an opinion, I am just sharing my awe in the way electronics are influncening the way stuff works, down to such (or at least it used to be) basic items like fuses. When I got my license in the 70's, power windows and electronic door locks were cool, and fuel injection was gradually introduced by brands like VW and BMW. Power steering was around, but it was still a mechanically connection between the steering wheel and the wheeld. The worst thing that could happen was that you got locked out of your car or that it would not run. Then came the ABS, a system with the authoraty to undo the brakes induced by the driver. In theory that could spell disaster, but I hear of few ABS related incidents. The gas pedal is now often connected to the engine by electric wires rather than mechanical ditto, witch has caused some accidents to occur (I believe Toyota has been suffering from bugs in their control software). Infinity has steering by wire as an option on some models, although it can be mechanically reconnected as a back-up. When I first heard of e-fuses, I thought that it was the last frontier of mechanics (or physics if you like) to give way to electronics.
Since the new smart fuse would be replacing the main pack fuse, which is in the front of the battery pack, there is no 12V wiring there for it to utilize. The BMS is at the back of the pack near the contactors. The fuse is in it's own little cubby that just has two large bus bars going into it for the fuse connection. No 12V there. Basically it's inline with the one leg of the battery. It can't even access the HV power either.
Jeff P
Member
Andyw2100
Well-Known Member
I started a thread suggesting that the upgrade to Ludicrous mode could result in improved reliability through reduced catastrophic contactor failure due to the new fuse being discussed in this thread, as well as the different material being used for the contactor. I thought perhaps some of you posting about the intricacies of the fuse may be interested in sharing your thoughts there as well.
Sounds like Ludicrous upgrade may reduce incidence of catastrophic contactor failure
Sounds like Ludicrous upgrade may reduce incidence of catastrophic contactor failure
spaceballs
Member
Actually in that review they might get dinged as if the micro-controller has interrupts enabled, then tight loops doesn't guarantee real-time results. Best place would be to place that in an interrupt (trigger when Amp > X), and then another timer interrupt that would periodically reset an watch dog timer in case the micro-controller hangs.
Also little related programming humor is_computer_on_fire() - Everything2.com
Last edited:
If Tesla just needed to increase from 1300A to 1500A, they would surely have just changed the fuse to one with a bigger rating.
Here is what they wrote:
So, it is all about precision. They're taking the old simple dumb fuse that blows at approximately 1300A within a few hundred ms of that limit being hit, and replacing it with a smart fuse that blows at exactly 1500A, within a defined +/- Amperage and time tolerance. With that precision in place, they can then push the limits more in other parts of the system.
The downside is, as pointed out, that the complexity of the device has gone up considerably. A simple fuse is based on the known properties of the material it is made from, and is randomly tested on a batch-by-batch basis. It would be interesting to see, in detail, the design of this new fuse. In particular if it is designed to fail open or fail closed (from a pyro point of view).
Here is what they wrote:
So, it is all about precision. They're taking the old simple dumb fuse that blows at approximately 1300A within a few hundred ms of that limit being hit, and replacing it with a smart fuse that blows at exactly 1500A, within a defined +/- Amperage and time tolerance. With that precision in place, they can then push the limits more in other parts of the system.
The downside is, as pointed out, that the complexity of the device has gone up considerably. A simple fuse is based on the known properties of the material it is made from, and is randomly tested on a batch-by-batch basis. It would be interesting to see, in detail, the design of this new fuse. In particular if it is designed to fail open or fail closed (from a pyro point of view).
Similar threads
