Well, the general answer to this is that as the motor needs to overcome the inertia from its state of rest, it requires high starting torque and aka high current is consumed to do that work.
On the other hand, Speed is also inversely proportional to Torque. Torque is directly proportional to the current and Voltage is directly proportional to the Speed.
As a result, it would seem only natural for a dc motor to consume high current at the start and then gradually consume less as the speed increases and torque requirement goes down.
However in this particular case, where I'm trying to understand the Speed Current relationship of a Geared DC motor. Which operates at 12V with max RPM 900 as read on the label. (These were the motors salvaged from a drill machine), I see this
The above is a current vs time plot, where I am independently increasing the speed from 0 to its max rpm via PWM.
The Input is Given via PWM with a range of 0-255 8Bit resolution.
The Peak current is achieved around half the Dutycycle aka Half the max RPM. The downward slope makes sense from the above assertion since speed is increasing current should decrease.
What confuses me is the increasing slope observed in the beginning. Speed is increasing, current is also increasing. But until it's half the RPM. Whereas it should have been a decreasing plot since the start of its operation.
My opinion about this is that, since the motor is self-locking it starts from a low start current as it has to overcome its own torque offered by the gears of the motor. And only after that point ie. Half the rated RPM is able to overcome the torque offered by the gears.
PS: The current sensor used is a shunt-based sensor if that helps.