Robotics and motion electronics
Motor drives generate the two problems that define this field — energy coming back up the bus when the load decelerates, and PWM edges radiating from every meter of motor cable.
Evidence
What we can substantiate
Everything below is work we have actually done, described at a level that respects the client’s confidentiality. If you need detail, ask under NDA.
- Regenerative energy handled deliberately A decelerating load pushes energy back into the DC bus. With a battery it charges; with a rectified supply it has nowhere to go and the bus climbs until something clamps it or fails. Brake resistor sizing, bus capacitance, and an overvoltage response that is designed rather than emergent.
- Motor cable treated as the radiator it is Fast PWM edges on a meter of motor cable are an efficient common-mode antenna, and the same dv/dt drives bearing currents that pit races over months. Shielded cable with 360° termination at both ends, common-mode chokes, and edge rate slowed to the slowest the switching loss budget allows.
- Encoder signals that survive a moving cable Incremental encoder channels degrade with flex fatigue, crosstalk from the adjacent motor pair and ground bounce. Differential signaling, correct termination, separation from the power pair inside the drag chain, and a cable rated for the actual flex cycle count.
- Current sensing that is accurate under PWM Shunt placement, amplifier bandwidth and sample timing relative to the switching edge decide whether current control is stable. Sampling in the wrong part of the cycle gives a beautiful-looking signal that is systematically wrong.
- Safe torque off as a hardware path Where a machine can hurt someone, the safe state has to be reachable without trusting firmware — a gate-drive disable that hardware alone can assert, with diagnostics that prove it works.
- Impact survivability as a design discipline For combat and field robotics: component mass and orientation planned against the shock axis, connectors that latch, potting where inspection allows it, and boards mounted so the chassis takes the load rather than the solder joints.
- Hot-plug transients root-caused rather than derated around Connecting a live battery to bulk capacitance rings the wiring inductance against it, and the overshoot can approach twice the supply — briefly, invisibly, and fatally for the input capacitors. We size the damping rather than replacing capacitors on a schedule.
Common questions
Straight answers
My drive trips on overvoltage when the axis decelerates. Why?
Because the motor becomes a generator and the energy has to go somewhere. On a battery system it charges the pack, which is fine until the pack is full. On a rectified mains supply the diodes block it entirely, so the bus capacitance absorbs everything and the voltage climbs until the drive protects itself. The fix is a brake resistor sized for the actual kinetic energy and duty cycle, or a regenerative front end if the duty is high enough to justify it. Adding bus capacitance only moves the trip point.
Why does my robot fail EMC when the bench prototype passed?
The bench prototype had short motor leads. In the machine those become a meter or more of cable carrying PWM edges, and common-mode current on that cable is what the chamber measures. Shielded motor cable with proper 360° termination at both ends, a common-mode choke, and — usually the most effective single change — slowing the gate drive to the slowest edge your switching losses can tolerate.
Do you do the firmware as well as the board?
We do the bring-up firmware and enough motor control to prove the hardware works — commutation, current loop, communications. A full motion-control application with trajectory planning and application logic is usually better placed with whoever owns the machine behavior, and we hand over a board with a documented, tested low-level layer for them to build on.
Got a board to design — or one that won’t boot?
You talk to the engineer who would do the work. Reply within one business day, and we’ll sign your NDA before you go into detail.