I got myself an Digital-Multimeter now. Measuring the current between the power supply + and the drivers PWR +, it only pulls 1A max. About 1A when accelerating and afterwards 0.7A.
The supply voltage stays at 24V, even when the failure occurs and afterwards.
The 7-8A I haven't measured, we have used a bigger stationary power supply in school before, because we thought we wouldn't have enough A available. The stationary power supply has an display showing the current current pulled. It showed us the 7A. But this was with the full 3.000rpm, after that the motor got the mentioned 120°.
When using it with 600rpm now, it pulls 1A and we still have the failure. We measured the temperature while the motor was spinning and it has 54°.
When the failure occurs the light blinks red and green endless. So more than 2 times.
The orange cable comes from an adjustable power supply now, just to give it the 0-10V speed level for testing.
The blue & yellow cable off the left side, are the vcc +5V & GND for the hall sensors, because our driver model doesn't have the vcc pin for the hall sensors. The 5V come also of the adjustable power supply.
All grounds are connected together & the rest of the circuit (arduino, display, valves, pressure sensors,...) aren't connected.
I really don't know what to try else anymore
So thank you very much for your help!