DIY 3D-printed electric waterjet — hitting duty cycle limit at 89% / ~9,600 RPM, FOC cutout (Fault 18) — anyone dealt with this?

[Antares87]

Been developing a DIY electric waterjet propulsion unit for surfboard/foilboard use. 85mm impeller, rewound outrunner on 14S, Flipsky 75350 ESC on VESC firmware.


Latest test session hit 49.8 kg thrust before cutout. The limiting factor isn't current (capped at 250A — tested up to 450A limit with same result) or temperature — it's duty cycle saturation at ~89%. ESC throws Fault 18 (ABS_OVER_CURRENT) right at that point, but I suspect FOC observer instability at high ERPM rather than a true overcurrent event.


Enabling Slow ABS Current Limit pushed the cutout threshold from ~9,100 to ~9,500 RPM, which supports the observer instability theory over actual current spike.


Questions:

• Anyone experienced duty cycle as the hard ceiling rather than current/temp?
• Is there a reliable way to stabilize FOC observer at high ERPM/high duty?

VESC log CSV and test footage attached below.

 

[scianiac]

From my understanding you are just running into a basic issue with how ESCs work in sensorless operation. At 90% duty cycle that is pretty much all the motor will give and the VESC's observer is having a hard time tracking the rotor position with that remaining 10% of time. Observer instability causing current spike is a real overcurrent event and turning on slow ABS to just tell the controller to ignore massive current spikes is a how you blow up a controller.

Honestly though this is really easy to solve, you just need a higher KV motor or more voltage. Yes you can do some things to maybe get to 95% duty cycle but that isn't going to give much more power. Usually I've seen inrunners used for these applications and that is probably because they are better for higher RPM. Did you rewind the outrunner with a much much higher KV? At some point you will run into massive magnetic losses running a outrunner faster and faster, then again with that much current maybe you balance those with high resistive losses and saturation so it works out.

 

[Antares87]

Thanks for the reply! To clarify, I’m not necessarily looking to push the settings any further than this; I just want to hit the 10,000 RPM mark—it’s right within reach, after all. Haha. With my current setup, the most efficient operating range is around 9,300 RPM with a motor limit of 250A, keeping temperatures in the low 40s. I posted this because I personally want to take on the challenge of reaching 10,000 RPM. ^^