Fix DC motor Vmax limiting in voltage input mode

[giusenso]

This PR addresses the issue #3489.

Summary

gainprm[8] stores the input mode:

• 0: voltage
• 1: position
• 2: velocity

Previously:

if ((int)gainprm[8] > 0) {
  ctrl[uadr] = dcmotorVoltage(...);
}

The condition treated dcmotorVoltage() as only a PID conversion function. However, that helper has two responsibilities:

1. Convert position/velocity commands into voltage.
2. Apply Vmax to every voltage, including direct voltage input.

For voltage mode, the helper explicitly does:

voltage = ctrl;
voltage = clip(voltage, -Vmax, Vmax);

But because mode 0 failed the > 0 condition, this code was never reached. Raw ctrl went directly into:

torque = K/R * ctrl - K²/R * velocity

Removing the guard makes the helper the canonical terminal-voltage calculation:

• Voltage mode: V = clip(ctrl)
• Position mode: V = clip(PID_position(ctrl))
• Velocity mode: V = clip(PI_velocity(ctrl))

It is safe because when Vmax == 0, the helper returns voltage-mode ctrl unchanged. Position and velocity behavior is unchanged.

The old if conflated “requires PID conversion” with “requires voltage processing”; only the first is mode-dependent.

Testing

• Added VoltageInputVoltageLimit to verify positive and negative voltage clipping and unchanged in-range inputs.
• Added VoltageInputVoltageLimitWithThermalState to verify torque and thermal dynamics use the same clamped voltage.