When a working and tested ESC lets loose the magic smoke, it can be a number of reasons. A common failure, at least one that recently destroyed an ESC of mine, is a Mosfet failure. These failures are spectacular due to the high fault currents supplied as the Mosfets are given low-impedance access to the battery.

The Mosfet is the hardworking switch that every ESC relies on. While the design of an ESC hasn’t really changed through out the years the Mosfets used have improved dramatically. This has improved the power density that ESCs could be designed for. 

Your basic ESC is shown above. The mosfets are the switches Q1-Q6. There are differences between various ESCs, but the basic structure remains the same.

Just a warning to the reader; I do tend to use technical terms that not everyone may understand. Please inform me if something doesn’t make sense.

Theory

A Mosfet failure is a really spectacular sight. A mosfet often fails to a short-circuit. This is caused by the dopants in the semiconductors diffusing and creating more conductive paths that the p-n/n-p junctions. If the mosfet heats up enough it may result in the bond wires from the package to the silicone junction fusing or desoldering themselves, which can in rare occurances lead to an open-circuit failure.

So how does this influence the longevity of my ESC. This is the basis for understanding how a mosfet may fail. To this there are 3 common causes of failure. 

1. Overheating

This is one that everyone understands somewhat. The more current you push through an ESC the more heat it produces. Common causes of increased current in FPV ESCs include:

1. Flying aggressively. The harder you push your quad the more power it needs to deliver to the props to perform the manouvers.
2. Overpropping. Using a more agressive prop can result in some interesting characteristics. However the more aggressive the prop, the more energy is needed to turn it and therefore the more power needed from the ESC.
3. Noise in the controller. The incorporation of newer faster flight controllers and faster communication between the FC and ESC, noise can make it into the ESC control. In the worse case this means that between control cycles the ESC could be commanded by the flight controller to go from 0-100% throttle. This is very demanding of the ESC especially with active braking enabled (a must have for smooth and locked-in flying 5″ quads). While this will definitely result in hot motors it may lead to ESC damage.

Naturally using a higher current rating ESC can avoid some of these issues. As will using the mystical 6s, but that is a debate for a different post.

2. Instantaneous Over-current

While this may seem similar to overheating as in essence it leads to higher heating, over-current leads more specifically to a localised fault in a single ESC. For instance; failure of the bond-wires. This can cause a single mosfet to fuse to an open circuit, which means your ESC won’t necessarily explode.

1. Cascade Failure. If another MOSFET has failed by another failure mode to short-circuit, the other MOSFET on that phase will overcurrent itself by essentially shorting the battery straight through the MOSFETS. For example in the image above, if Q1 fails to a short circuit, then when Q2 is switched on it will act as a short circuit to to the battery. This is often seen when one or more mosfets will fail in an ESC.
2. Motor failures. This to me feels the worse, as after destroying a precious motor, you run the risk of destroying a ESC if you don’t notice the failure. An inter-turn short circuit could lead to the ESC delivering more current that normal during a switching cycle of the ESC (this will be discussed later in a new post).

Naturally using current limits available in the new BLHELI_32 ESC that support individual current limits will reduce the chance of a serious overcurrent event destroying your ESC.

3. Over voltage

This is where things become quite a bit more interesting. By overvoltage I do not only reference blowing your 4S ESC with a 6S pack. The spikes produced by switching noise or even plugging in a batter pack can lead to voltage damage to components. This may not seem like an issue we as hobbyists need to be concerned with, however I have seen a few failures which I believe can be attributed to an overvoltage condition. Overvoltage faults can affect other components on the ESC or lead to PCB level failure. This will be explored in a later post.

In fact a few of us may have seen a failure that can be attributed to an overvoltage. If you have seen an ESC failure following a high capacity capacitor coming off during flight, than chances are you have witnessed a high voltage failure. This is a reason why certain ESCs will not honor any form of warrantee if the capacitor was not soldered on.

In fact a recent ESC of mine blew due to an issue I believe can be attributed to an overvoltage condition as a result of a capacitor coming loose. I believe this is a common failure cause of the HobbyWing G2 ESC, as a new ESCs I just received come with this sticker

Conclusion

Hopefully this has explained some of the ways in which your ESC will give up the magic smoke. In understanding  how components fail we are able to take appropriate steps to avoid faults or identify potentially ESC destroying scenarios.