Frequently Asked Questions

If you have a question about our products or technology that is not answered in the below FAQs, please contact and our team will be happy to help.

Datasheets typically contain basic motor parameters like torque constant, voltage constant, resistance, and inductance which are all measurable parameters. It is not recommended to use any of the calculated parameters on the datasheet (which is anything that is not measurable).

Km is a well known and important figure of merit for a motor, unfortunately suppliers calculate this term differently and their calculation is sometimes not in the correct units to compare to other motors.  Motor rated torque is probably the most important figure of merit, and it is the least accurate and least understood.

There are several other things that are not present like cogging torque, hysteresis torque, torque linearity, eddy current torque, thermal time constant, thermal resistance, and torque versus angle curves (an indication of torque ripple). If you are designing a system with a motor in it, either select a motor that is well above what you need, or talk directly to the supplier to get all of these things that are not in the datasheet.

Motors are easy to get quickly online from shell companies, and in that case you get even less data and your second purchase may not result in the same performance.

Using a CAD model you can get the system inertia, if the materials are correct, in turn acceleration rate required will allow a calculation of torque needed to acceleration (Torque = Inertia * acceleration rate). Friction is much more complicated. If there is an existing or similar system you can put a torque wrench on it, if not you need to use information from bearing suppliers (which tend to be conservative). Total torque is typically covered by these two. There is occasionally viscous friction (damping) torque, which is also similar to windage torque that is present that is proportional to speed.

Typical ranges for motors are -50 to +150C. This means that the motor supplier will use materials that are rated above these to achieve a safety factor. If there is an encoder in the system the  limits may change from operation at 0C to 80C due to limits on the encoder and it depends on where it is in the system. Bearings area is also limited by colder temperatures. Electrical insulation systems and magnets can be chosen to operate over 200C if the rest of the mechanical systems can handle this temperature.

No.  A motor with cogging torque will always have torque ripple when running. A motor without cogging torque (slotless or air-core motor) may still exhibit torque ripple if the phases are not balanced, (in amplitude and phase relationship), and the torque versus angle curves are not sinusoidal with low harmonic content. It can also happen if current waveforms from the driver are not well controlled sinusoids. If a six state older style motor controller/driver is used, any motor will exhibit torque ripple.

In vacuum there is no convection. Only conduction. This means the mount and material used is critical to heat flow. Smaller motors utilize conduction to a larger extent than large motors with lots of surface area that use more convection. So, if it is a small motor <50mm diameter than its derating in vacuum would be less. A large motor >50mm would require a more significant derating. General rule of thumb would be to assume 50% of the heat transfer. This mean that the motor would be derated to 75% of the torque output because power is mainly a function of copper losses that proportional to the square of the current.

Eddy currents in a motor are related to the rate of change of the magnetic field in the iron. They look like a damping term if you are doing any modelling, NM/rad/sec is an example. Motor designers typically laminate the magnetic structure to minimize the flow of eddy currents. Thinner laminations provide a higher resistance to eddy current flow. Any eddy currents present will produce heating in the motor and result in a drag torque that is proportional to speed.

Yes. Hysteresis in the motor laminations is caused by the changing magnetic fields as the motor rotates. Motors with high levels of magnetic flux will have higher hysteresis. Motor or transformer grade steel with high Ni content is typically chosen to minimize magnetic hysteresis. Hysteresis is a friction term if you are doing any modeling. Most motors have hysteretic friction. The higher performance motors with higher strength magnets have the higher hysteresis. There are some trends moving away from laminations to SMC (sintered metal compound) materials.

Most motors have a torque constant you can find in a datasheet. This is only first order approximation for torque output as a function of current. In reality many motors have some saturation and at higher currents actually produce less torque/amp. This is called torque linearity. It is not shown on any datasheets, but can impact system performance and make your control system very non-linear. Only Air-Core motors and slotless motors have good torque linearity up to current levels well above the continuous rated current.

Click to out more about torque motors.

Most motors have a continuous torque rating. This rating was arrived at under some thermal test conditions at the supplier. Rated torque is least understood and the most corrupt parameter for any motor. Heat is the enemy of torque. Know the test conditions that a particular motor was rated at and make an engineering judgement whether you can use those ratings. We recommend against using any ratings unless the test conditions were exactly like your project (never the case).  Torque rating is a reference point only. Thermal modeling is difficult and needs empirical input to be useful. Many solid modeling software tools have a thermal FEA module this can help determine the likely thermal resistance of the system you are developing. Peak torque rating should not be used until you know how close the motor torque output is to your needs.

Most projects start with a size and payload for the robot. This drives axis torque requirements, speed requirements, and size requirements. The biggest question is whether a gear will be needed.

Direct drive works for certain robots that have relatively light loading. Direct drive is also very attractive for collaborative robots, because any gearing has implications in reflected inertia limiting stopping time. After the decision on gearing, a decision whether to integrate a frameless motor kit or buy a housed servo motor is next. Housed motors are easier to use but tend to be long and thin versus short and wide.

Almost all robots big and small have moved to highly integrated frameless motors that share the main bearing systems, eliminate couplings and offer shorter axial packaging. Lastly, based on the payload, bearings will need to be selected to handle the loading, speeds and any through hole that a robot joint might require. Lastly, the robot joint output needs a high resolution accurate encoder, AND the input side needs an encoder with medium resolution to run the motor.  High resolution is 20+ bits and medium is about 16 bits, absolute encoders are always preferred for less wires, and the best information. They are not much more expensive than incremental encoders these days.  You basically pay for accuracy, high resolution and an absolute interface is standard.