Pump Stalled? How Fast Can Your BLDC Controller Respond?

Have you ever had a pump suddenly stall because a bit of debris got stuck in the impeller? It happens more often than you think. And if your BLDC controller relies only on software overcurrent protection, that brief moment of delay – typically around 100 to 500 microseconds – might already be too late.

So what is the solution?

A more responsive approach is combining hardware-based **cycle-by-cycle (CBC)** protection with MCU supervision. Here’s how it works: a comparator watches the voltage across the sense resistor in real time. When that voltage crosses a pre-set threshold (say, 200mV for a certain current limit), the PWM output turns off in **less than 1 microsecond**. That’s noticeably faster than waiting for software to react.

After the shutdown, the MCU logs the fault and can attempt a restart – for example, a light-current try to see if the impeller has freed up. In repeated stall tests, this kind of controller has handled up to 100 sequential jams without failing.

Does this mean it will never have problem? Not at all – no design is bulletproof. But using hardware-level protection gives your pump a much better chance of surviving unexpected stalls, especially in dirty water or abrasive environments.

At our end, we’ve made this feature as part of our BLDC PCBA design. If you’re working on a pump project and want to know how our controllers behave during a locked-rotor condition, feel free to ask – we’re happy to discuss and talk through real-world performance.