Using NEMA 34 Stepper Motors in a Robotic Platform

Thread Starter

klaus25

Joined Jun 10, 2025
1
Hi all,

I’m working on a new robotics project and I’m considering using NEMA 34 stepper motors for the drivetrain. The robot is a ground-based mobile platform, designed to carry a payload of around 40–60 kg. I'm aiming for a balance between torque and precision, which is why I’m leaning toward steppers over DC motors for now.

System Overview:
Motors: NEMA 34 stepper motors (around 8–12 Nm holding torque)

Motor drivers: Digital stepper drivers (like the Leadshine DM860 or similar)

Power supply: 48–60V, up to 10A

Controller: Arduino Mega + external step pulse generator (possibly upgrading to STM32 or Raspberry Pi with real-time kernel)

Drive system: 2WD with large rubber wheels, direct drive or via belt reduction

What I’m looking for:
Advice from experience: Has anyone here used NEMA 34 motors in mobile robots before? I’d love to hear about your setups.

Overkill or smart choice?: Are NEMA 34s too much for a 50kg robot? I want high torque, but I also want efficiency and manageable power consumption.

Closed-loop vs. open-loop: Would you recommend going with a closed-loop stepper system? I’m concerned about missed steps during uneven terrain movement.

Heat/noise concerns: Any tips on managing heat build-up? Do they really get that loud under load?

Alternative suggestions: Would I be better off with high-torque DC motors + encoders and PID control for this application?

I’d really appreciate any insights, photos of similar builds, or even just "don’t do this" warnings. Thanks for reading — looking forward to hearing your thoughts!
 

Irving

Joined Jan 30, 2016
5,070
Nothing fundamentally wrong with your thinking, though you've not mentioned one vital component - batteries. Ok, you said 48 - 60v @ 10A, but that's too generic. Whats your required run time/duty cycle?

You say a 'balance between torque and precision' and 'I want high torque' but what's your actual requirement? Speed, acceleration, braking, etc need to be spec'd before motors can be spec'd. What do you mean by precision, and why do you need it? Also don't rule out DC motors, they are, arguably, a lot simpler to control and manage than steppers and give a better bang per buck than a stepper for drive train applications. You may find they don't need to be as high a torque as you think...
 

MrChips

Joined Oct 2, 2009
34,714
NEMA 35 specification gives the physical dimensions of the faceplate and mounting holes.
It does not specify the size, weight, electrical data and torque of the motor.
 

DNA Robotics

Joined Jun 13, 2014
670
That should be plenty.
I use some like that on a milling machine at 24 volts.
There is no stopping them and no missed steps.
Providing you ramp up the speed.
They will accelerate fast, but not Zero to 60 instantly.
Go for it.
 
Last edited:

MaxHeadRoom

Joined Jul 18, 2013
30,605
@klaus25 usually the reason for going the stepper route is no closed loop required.
If you want precision from DC servo's then you require PID loop and all it entails.
Having said that, industrial robots are generally closed loop servo's.
 

MrChips

Joined Oct 2, 2009
34,714
You also need to consider holding torque.
The holding torque on a stepper motor is fixed and depends on the current applied and full-step or micro-step condition.
The holding torque on a DC servo is dynamic and depends on the PIC controller as well as the power of the motor.
 

Irving

Joined Jan 30, 2016
5,070
@klaus25 usually the reason for going the stepper route is no closed loop required.
If you want precision from DC servo's then you require PID loop and all it entails.
Having said that, industrial robots are generally closed loop servo's.
You also need to consider holding torque.
The holding torque on a stepper motor is fixed and depends on the current applied and full-step or micro-step condition.
The holding torque on a DC servo is dynamic and depends on the PIC controller as well as the power of the motor.
All of the above is true..

However, this is for the drive train of a robotic cart carrying a 50-60kg load across a varying terrain. Issues such as wheelspin will need to be considered. IMHO a DC (brushed or brushless) motor with fine torque control is better suited to that requirement.
 

Irving

Joined Jan 30, 2016
5,070
A worm drive reduces speed, increases torque, self locks and prevents slippage.

View attachment 350823
Agreed, ignoring backlash and end-lash, the worm drive prevents slippage in the drive train, but does nothing to address wheel-spin at boundary of wheels and terrain. Stepper motors in the drive train as a means of measuring position is a pointless & fruitless approach as any attempt to accurately measure position by counting steps, or using an output-side rotational encoder, is doomed to failure other than at very low speed/acceleration on a high stiction surface. You need an external 2-or 3-D frame of reference to measure position to any required degree of accuracy. For example, in our Motion and Mobility lab we have a bank of 16 cameras and IR scanners to record the 3-D position of multiple 'location targets' attached to the object (person, wheelchair, robot, etc.) to an accuracy of approx ±1mm in a 10m x 10m x 3m volume. Industrial robot carriers such as in warehousing or production lines either run on rails or use a combination of wheel rotation encoders plus machine vision, or other sensors (eg edge detectors for a white stripe on a dark floor), to provide absolute positional accuracy for loading/unloading or charging purposes.
 
Top