The controller would normally have a current limiting system to protect itself and the motor under normal operation. The fuse/MCB is there to deal with abnormal situations. There should be a fuse in the battery feed to the controller (may be built into the controller).

If you are using one controller to feed two motors there may be some sense in protecting each motor with a fuse or MCB. If a drive belt breaks, for example,  the full current will go to the other motor.

If using a MCB make ure it is rated for DC operation. Many are but it’s often not on the “headline” specification. You need to use the detailed datasheet. MCBs certified to UL 1077 will work with DC. An example is the ABB S201 series.
If the MCB is not DC rated the speed of opening and gap when opening may not be enough to extinguish a DC arc. The arc can start a fire. AC arcs are easier because the voltage drops to zero every half cycle.

Robert.

Roger B,
Can you provide any references or evidence to back up the assertion that “The voltage needs to be above 50V before it is a real problem.”?
Arc welders are not a comparison.Even DC welders use a pulse output that causes the voltage to drop every half cycle. The exception is industrial welders with 3 phase input and some inverter welders.

Robert.

On 24 July 2024 at 08:26 Roger B Said:

Arc formation and extinguishing is a wide an complicated subject. Materials, geometry and polarity all have influences. There are various voltage thresholds for starting an arc without contact, sustaining an arc and extinguishing an arc.

To move to practical examples: Automotive systems have used 24V DC for decades with no special consideration given to arcing at the contacts or the design of fuses for arc suppression.

 

…

As always, design engineers always have to consider the context!    Though what Roger says is generally true, ‘problems are expected to start above 48V.’ is a practical rule of thumb, not appropriate in all circumstances.   For example, a capacitor is fitted across the contacts of the contact breaker in a traditional car ignition system.   Contact breakers open and close repeatedly at high-speed, and unless suppressed, the resulting sparking and arcing soon cause damage;  the capacitor also makes the energy that would have been wasted in sparking available to the ignition coil.  This is well below 48V.

Another example, what’s acceptable in a road vehicle may not be good enough in an aircraft.   An electrical fire in an aircraft is far more serious than one on the road because passengers are trapped in the plane until it can land safely.   Therefore, I hope that no-one in aviation follows Roger’s guideline, even though it has merit in other circumstances!

In my judgement, Duncan can apply Roger’s rule of thumb.  His application is a hobby locomotive, and the worst case scenario, a burning MCB, passes my ‘so what test’.    If Duncan is hauling passengers when an electrical fault causes the MCB to arc and catch fire, perhaps involving the motor and cables as well, then his train will coast to a stop with plenty of time for the driver to escape.   Passengers are at an even lower risk – they’re protected by being further away from the loco and they can also escape easily.

An MCB may not be the best answer.  In Duncan’s place, I’d have a think about what the MCB is for.  MCBs are good for protecting cabling if a temporary overload occurs and for breaking the circuit as soon at an exact rating is exceeded.  A standard 20A MCB will break at 20A, unlike an ordinary 20A fuse which will tolerate a very high current for short periods.    A fuse’s ability to tolerate temporary overloads makes them attractive for motors because motors typically draw a lot more than their rated current when starting:  not for long, but maybe enough to flip an MCB out unnecessarily.

In a loco, I guess Duncan’s purpose is to protect the cabling/battery, in which case a fuse is a better bet.  It balances reliability and protection.   I think in Duncan’s application that a blown fuse indicates a serious fault requiring investigation, not fixable by simply resetting an MCB.

An electronic motor controller might change everything. Fuses are slow acting and insensitive because they rely on a length of wire getting hot enough to melt, which can take a long time!   Circuit breakers typically use an electromagnet to trigger a spring-loaded switch. Far faster and more accurate than a fuse, but very slow compared with electronics.  Electronic fault management can also be remarkably clever, for example programmable to meet a motor’s starting current requirement, but treating the same current as a fault once the motor is running.   On the downside, the manual might be incomprehensible!

Dave

There is no reason not to use a component correctly rated for the application. Modern MCBs are available with DC ratings so use one. DC rated contacts operate faster and have wider gaps than similarly rated AC ones. Any fire is to be avoided. What if the vehicle is running near dry grass for example?

Robert.

On 24 July 2024 at 12:35 Roger B Said:

On 24 July 2024 at 11:00 SillyOldDuffer Said:

For example, a capacitor is fitted across the contacts of the contact breaker in a traditional car ignition system.   Contact breakers open and close repeatedly at high-speed, and unless suppressed, the resulting sparking and arcing soon cause damage;  the capacitor also makes the energy that would have been wasted in sparking available to the ignition coil.  This is well below 48V.

 

Dave

The Kettering ignition system is a bit of a non sequitur when talking about MCBs. It is designed to have a few hundred volts across the contact breaker from the coil back EMF and so it is well above my guide line voltage. Hence it uses a capacitor for spark suppression (as well as to produce a resonance to extend the spark period).

I gave it as an example of the need for the designer to do more than assume a rule of thumb like  ‘problems are expected to start above 48V‘.

A car ignition system’s contact breaker requires special attention even in a mere 6V automotive system.   It’s because the breaker rapidly switches an inductive load, and the inductance causes high-voltage spikes that extend how long each arc lasts.   The nature of the load makes a difference, and Duncan’s load is an inductive electric motor…

Dave