After all, there would not be any load going through the motor and not having the controller powered up, it won't be able to switch to regenerate.
Without a battery connected, a single wheel can power a LED (40 led) headlight at full power, so I am also curious what power it generates!
The voltage generated by the wheel at speed would still power up the controller, even with the battery switched off
(or removed).
If you disconnect the battery and pedal fast enough
(with the brake lever pulled in enough so that the regen is on) the voltage being output does not go above ~60V which is fine, but if you do not apply the brakes, the voltage continues to rise with the wheel speed and can go well above this level.
I once tried this out by hanging my bike up so the wheels were off the ground and then I wound the pedals as fast as I could by hand, and I had a voltage reading on the battery connector of ~77V with no battery connected!
I would still be concerned about damaging the controller with such a high voltage going through it, how long would the 63V capacitors in the controller be able to cope with this excessive voltage before they finally failed?
I'm not sure if it would be OK to leave the regen permanently switched on to prevent the voltage in the controller from going too high while riding without the battery, because I don't know what the controller actually does in order to prevent the generated voltage from exceeding 60V at speed.
My maximum measured regen current was
23.35 Amps (681.5 Watts) using a 7 cell LiPo pack (
25.9V 10Ah) and
9.65 Amps (525.8 Watts) using 14 cell LiPo pack (
51.8V 5Ah), and these readings which were obtained at approximately the same speed. So at the 60V limit I reckon at least
8 Amps (480 Watts) would be available if a load was put on the controller while using regen at a similar speed.
Alan