The idle heating issue of Ubox ALU 150a in float package mode

One of our customers reported to us, that he tested our Ubox ALU 85V 100A VESC controller, when he just installed the float package for one-wheel and keeping power on the controller, at 20S battery, on bench, without motor connected to the controller, the controller will be heating up quickly, will reaching up to 85C degree in a few minutes. This report was upheld by few users who have same controller with same float package mode.

When this symptom happened, the controller was at an idle but driving state: the motor is not being driving, but the MOSFETs are being driving in an idle state, each half bridge is just up-on/low off, then up off/low on, with a 50% duty cycle PWM.

This state can be reproduced with VESC tool desktop version, connect the controller at 85V voltage, click the keyboard control on right side bar to use key board control, then tap the right cursor key and release it, the green led of the controller will bright, the controller will idle-driving the MOSFETs, in real time data page, we can see the temperature of FETs will start to rising.

This symptom is only on our controller who adopted the JMSH1001 MOSFETs, mainly the Ubox ALU 85V 150A, Ubox ALU 85V 250A.

Increase the deadtime of PWM can not ease this heating.

Set lower switching frequency, for example 3Khz, will ease this heating.

The Ubox single 100V 100A and Ubox ALU 100V 100A lite don’t have this issue.

Other applications such as PPM, UART, ADC, etc., at current control modes, don't have this issue. Since the controller will not drive the FETs When input signal is idle.

This may be caused by the source-drain capacitance of the JMSH1001ATL FETs, on a single half bridge, a single PWM cycle, when high side FET on/low side FET off, the low side FET will be charged, when high side off/low side on, the low side FET will be discharged. This procedure made a current flow from batt+ to batt-, generate the heat.

JMSH working well enough at working load, compared to the other FETs.

1. Although the JMSH1001 FETs heating in this state, but JMSH is proven good at heating when full running, nearly two years of user’s fed back that the Ubox ALU 85V 150A is excellent at full load. So, make an adequate thermal dissipate for the controller will easy this issue, anyway, every controller heating at working load.

2. Modify the “float package”, let the script release the FETs when user not riding the one-wheel, this need the package author’s help.

Use our Ubox ALU 100V 100A Lite instead, this one uses the MDT10N023RH FETs, same as the Ubox single 100V 100A, but with compact size, lower price, and Aluminum power stage, the thermal conductivity is far better than the Ubox single 100V 100A, if we fix it on the metal frame of you vehicle, bottom sticking to the metal surface. This is the advantage of Aluminum power stage.