Electric drives: torque control, losses, motor control skills

Torque control leaves efficiency on the table

What happens. The drive produces the requested torque, but with a standard control structure and fixed parameter tables. Machine parameters change with temperature and saturation, and the tables are a compromise across the whole operating range.

What it costs. Losses in the motor and inverter, extra heat that must be removed and a margin on the rating that makes the drive bigger than needed. For battery-powered applications the loss is range.

Why the usual fix fails. Larger tables and finer calibration take weeks per variant and still do not follow the machine as it warms up and ages. More measurement hardware adds cost to a product that is sold on price.

How a fixed-price remote diagnosis checks it. From test-bench logs, current and voltage records, efficiency maps and the parameter tables in use, I check where the losses come from and how far a model-based torque control would move the operating points. The fixed-price remote diagnosis tells you whether predictive drive control is worth the development, and what the control hardware would have to deliver.

Background: Predictive control of drives and motors

Motor control for converters and servo amplifiers needs scarce specialists

What happens. A maker of drive electronics needs engineers who can design motor control for converters and servo amplifiers: current loops, observers, tuning on the test bench. They are few, and the role stays open while new product generations wait.

What it costs. Delayed product releases, existing staff stretched over too many variants and development knowledge that sits with one or two people.

Why the usual fix fails. Recruiting takes long in a thin market. A control library from a supplier gets a product running but not the last bit of performance, and it cannot be adapted to the hardware without someone who understands it.

How a fixed-price remote diagnosis checks it. From your current control structure, test results and the product specification, I check where the limit of performance lies, in the control, the sensors or the hardware. The fixed-price remote diagnosis tells you what could be improved without a new hire, and what a longer engagement would have to cover.

Background: Predictive control of drives and motors

Welded hairpin stator joints cannot be tested cheaply inline

What happens. In motors with hairpin windings, the laser welds between the copper pins decide the performance of the motor. There is no fast, cheap test for every joint inside the line. Tiny insulation residue on the weld spot weakens it, and positioning errors of the pins carry over from one process step to the next.

What it costs. Scrap of finished stators, the cost of end-of-line tests that catch faults late, and a worry about joints that pass the test but fail in service.

Why the usual fix fails. Testing every joint with slow methods cuts into cycle time. Sampling misses the rare bad joint. Tightening each single step does not remove errors that add up across steps.

How a fixed-price remote diagnosis checks it. From process data of the stripping, positioning and welding steps, with the results of the end-of-line test, I check which signals predict a bad joint and how the errors travel between steps. The fixed-price remote diagnosis tells you whether a computed quality value can flag a doubtful joint early, and what data you would have to start recording.

Background: Computed measurements