Based on measured data, a typical stepper motor used in a cash transport belt will experience load torque change at a rate of about 300 mNm every 40 ms.

In lab, one such stepper motor was subjected to load torque transients between 176 mNm and 1.46 Nm at a rate of 300 mNm/40ms. The on time for the peak load was 500 ms, and the duration between peak load events was 4.5 s, corresponding to a 10% duty cycle for the peak load.

For this motor, the auto torque learning routine was run at no load with the following parameter values:

• ATQ_LRN_MIN_CURRENT = 00110b
• ATQ_LRN_STEP = 11b
• ATQ_LRN_CYCLE_SELECT = 11b

Figure 3-1 shows the snapshot of the learning routine for this motor.

Resulting values for the ATQ_LRN parameters are -

• ATQ_LRN_CONST1 = 83
• ATQ_LRN_CONST2 = 120

The parameters for current control and PD loop control were selected as:

• ATQ_TRQ_MAX = 120
• ATQ_TRQ_MIN = 35, corresponding to the current needed to support the motor when no cash is transported
• ATQ_UL = 45
• ATQ_LL = 43
• ATQ_LL = 43
• ATQ_FRZ = 1
• ATQ_AVG = 0

Figure 3-2 to Figure 3-5 showcases the output current and supply current waveforms with and without auto-torque in the event of a load torque change.

As seen from the scopeshots, supply current consumption is significantly lower with auto-torque. Based on lab measurements -

• Power consumed without auto-torque = 24 V * 885 mA = 21.24 W
• Power consumed with auto-torque = 24 V * 402 mA = 9.65 W
• This represents a power saving of 55% for this motor with the specified load profile.

As for the power dissipation in the motor -

• Power loss in the motor without auto-torque = (2.8 A * 2.8 A * 1.5 Ω) = 11.76 W
• Power Loss in the motor with auto-torque = (0.1 * 2.8 A * 2.8 A + 0.9 * 0.85 A * 0.85 A) * 1.5 Ω = 2.15W
• This corresponds to a heat reduction of 82 % in the motor coils, leading to lifetime and long term reliability improvement.