Design help for fuse selection!

[Swapnil]

Hello!
I am designing a PCB for my project where I have 5 ESCs mounted on the board. The maximum current rated for the ESC (at 100% operation) is 30A at 16V. I am planning to add fuses to each of the ESCs in case of an overcurrent situation, and I have derated the current by 25% to 40A for each ESC. The problem I am facing with respect to designing is that I am planning to use 5 fuses(https://www.digikey.in/en/products/detail/littelfuse-inc/0299040-ZXNV/146669) for five individual ESC but instead of five individual fuses, I can go with a single fuse rated at 200A(https://www.digikey.in/en/products/detail/littelfuse-inc/153-5631-6202/2515960). So, what will be a good design choice keeping in mind the mounting of the fuse on the PCB and other criteria?

[wraper]

For 200A fuse to blow in a time of a few seconds you likely will need something like 400A current and to blow in a fraction of a second significantly more. Are you sure that faulty ESC circuit will not protect the fuse by blowing up first?

[moffy]

You are better off with the 5 individual fuses, a single failure of an ESC module might not be sufficient to blow your single 200A fuse. Also limiting the failure current with 40A fuses instead of a 200A fuse will limit damage to board and wiring due to current. 5 fuses means it is easier to identify a single failed module, as well as protecting the wiring going from the ESC modules.

[Swapnil]

For 200A fuse to blow in a time of a few seconds you likely will need something like 400A current and to blow in a fraction of a second significantly more. Are you sure that faulty ESC circuit will not protect the fuse by blowing up first?

I think the 200A fuse will be able to protect the faulty ESC but it's too much derating. I think I need to change the rating of the 200A fuse and make it smaller. I just considered 30Ax5=150A and derated that to 25% which is 200A. My doubt was regarding the individual fuses which are connected to individual ESCs in my circuit. Will it be a good choice to go with the configuration as per my schematic? I am worried that the first fuse will be affected when an overload will occur and the other fuses won't be affected.

[Siwastaja]

No, faulty ESC cannot be protected by a fuse. (Assuming you mean that ESC would survive being functional after the incident. No way.)

Fuse can only blow due to design failure in ESC*, so don't fail your design, OK? Then, fuse will never blow.

*) or serious abuse beyond the ESC ratings; for example, severe input overvoltage

Fuse is still important and used to protect against fire.

One fuse per ESC would be obviously the right choice given such large current.

On the other hand, you might really want to ADDITIONALLY consider one fuse right at the battery output.

Fuse is dimensioned so that it protects the downstream wiring from catching fire.

[Terry Bites]

A rule of thumb is to pick a fuse 25% higher than yout normal max current consumption.
Melt time, voltage and waveform will also affect your choices.
Choosing the right fuse is not as simple as you think it might be. "Fuseology" may help you here.

[Swapnil]

Thank you for your valuable suggestions. I had another doubt w.r.t the back emf that will be having due to five inductive loads. How can it be effectively suppressed?

[ejeffrey]

As Siwastaja said: fuses aren't to protect your circuit.  They are there to protect the rest of the world from your circuit should it fail by preventing a fire.

For a fuse to do anything useful, a potential fault has to be able to pull enough current to blow the fuse within a useful amount of time, considering both the source capacity and the wiring resistance.  For instance one ESC fails and the total short resistance is about 35 milliohm the fault current will be around 400 amps and the fuse may take many seconds or even a few minutes to blow -- you would need to check the data sheet for your particular fuse.  If the fault resistance were 50 milliohm, the fault current would be 300 amps and the fuse might never blow.  If your device is unsafe in that condition the fuse didn't help and if it is safe then the fuse might be unnecessary.  Dropping the fuse to 150 amp doesn't change this calculation much.

If you instead put a 40 amp fuse on a each ESC and the fault current for a single failure is still 300 amps now that is almost 10x the fuse rating and it should blow in a fraction of a second,  In general that's exactly the sort of situation where you use fuses: when you have a big beefy power supply feeding a lot of smaller loads in parallel.  You give each load a fuse or circuit breaker to keep it from having to handle the full short circuit current of the bigger power supply.

[trobbins]

Like any device, a fuse needs to be designed in to a circuit - just pulling a fuse value out of thin air is poor design.

You need to define your current supply and what it is capable of.  The schematic suggests a battery - you need to confirm how that is set up, and each cell/monoblocks short-circuit current capability, and then follow that through with connection resistance data to determine the worst case fault current rating at the fuse input - there is software and design info available to help if that is daunting.  You also need to work through the fault scenarios for your load sections, and decide what fault is the worst you want to try and cope with and what happens during such a fault as time progresses.  Then you need to define what normal current needs to flow through the fuse, and whether that has transient levels included (like in-rush at turn on).  Then you can look for fuse or other methods of over-current protection that may best achieve timely protection.

[Siwastaja]

Thank you for your valuable suggestions. I had another doubt w.r.t the back emf that will be having due to five inductive loads. How can it be effectively suppressed?

No need to "suppress" anything. Draw up the motor controller (the half bridges: 1 for unidirectional brushed DC, 2 for bidirectional brushed DC, or 3 for BLDC/PMSM) schematic on paper and work out where the current flows during the switching states. There always is a path for current.

Do you need to get this done? If yes, hire someone experienced in motor control design. Starting from complete scratch, it will be more difficult than you may understand. OTOH, it's a good learning experiment, but you need to invest some time and start reading instead of asking irrelevant questions.