ULN2003A driver board inverted?

[panflash]

I almost hate to ask such a basic question but I've gone cross-eyed searching for the answer. I'm working with a common ULN2003A driver board to drive a 12V 28byj-48 motor. I'm working with a 5V Digispark microcontroller which has a voltage regulator on the VIN pin, therefore I patch the positive and negative pins of the driver board to the VIN and GND pins of the microcontroller respectively. A simple program was able to run the motor in either direction. After turning I want to power-down the motor to reduce consumption, so I set the logic pins to low. At that point the ULN2003A gets very hot. Seems like current is still running to the motor.

My setup is like this or this except using a 12V supply. Also like this.

When voltage is low on the input pin I expect the output pin also to be low. I want to turn on the 12V (less the drop) current to the motor by providing 5V on the input pin. But I find the output pin voltage is high when the input pin voltage is low.

Supply is good:

Input logic is low:

I expect this to be 0V:

 

[betwixt]

The ULN2003 is a transistor array, the output pin conducts to ground when the input is high so effectively it does invert but with an open collector output.
Think of it this way: if the motor common connection has supply on it, the motor coil will be energised when 5V is applied to the ULN input pin.

Brian.