Overcoming Torque Deadzone on EV3 Large Motor

[reillys9]

Hello,

I hope you're all keeping well!

I am an engineer and I'm creating a demonstrator using industrial parts, but I'm actually testing the machine with LEGO Technic.

I'm trying to create a simulation of a system (I don't think the specifics are important but I'll gladly explain it more if anyone needs to know) using LEGO Technic. I have flashed an EV3 controller, and I'm writing code for it in VS Code via the inbuilt extension. I am using a Large EV3 motor to simulate the system. However, one issue that I am facing is that the duty cycle of the motor when it goes from 0 percent to 1 percent creates a massive jump in torque, i.e the difference between the coasting and the actual torque control being engaged does not increase linearly, rather it spikes, and then it goes linearly going from 1 to 2 percent to 2 to 3 percent to 3 to 4 percent, etc.

Does anyone have any fix for this? It's no bother buying a different LEGO motor or a third party LEGO motor (as long as it still has connectivity with the EV3).

Any help would be greatly appreciated

[Project516]

This might be an overkill for this, but a PID controller may help regulate this. Also depending on your situation a medium motor (45503-1) may be better.

[dlech]

Can you share the code you are using (or even better a minimal example)? At 1% duty cycle, there should not even be enough torque to make the motor turn. If you are using the "run-direct" command, the duty cycle should be controlling the PWM output directly and a duty cycle of 0 should be the same as ev3dev's "brake" rather than "coast".

[reillys9]

Sorry for taking so long to get back

@Project516, we're one step ahead of you! We already implemented one because the resistance was very jerky.

@dlech, just a bit of context. We are trying to simulate a spring mass damper. So we have an industrial motor, and when that industrial motor goes to reach it's set position, we want the LEGO EV3 motor to resist that motion in proportion to the industrial motor's position (how far away it has gone from the starting point), its velocity and its acceleration.

So when the motor goes from coast to engaged, there is a big jump in the "stall" torque (this is the torque that appears to keep something locked in place). This torque is about 0.1Nm, and the max rated torque for the EV3 motor is 0.17Nm, which gives us a very narrow range.

If there are any ideas or thoughts on how to fix this issue (probably no way to overcome it using that EV3 motor (?), it's all very much appreciated!

Thanks very much

[rhempel]

There's too much "stiction" in the LEGO Motors for that application - I think you want a purely magnetic brake for what you are trying to accomplish ...

…

-- Ralph Hempel ***@***.***

On Thu, Jul 31, 2025, at 11:10 AM, Seán Reilly wrote: Sorry for taking so long to get back @Project516 <https://github.com/Project516>, we're one step ahead of you! We already implemented one because the resistance was very jerky. @dlech <https://github.com/dlech>, just a bit of context. We are trying to simulate a spring mass damper. So we have an industrial motor, and when that industrial motor goes to reach it's set position, we want the LEGO EV3 motor to resist that motion in proportion to the industrial motor's position (how far away it has gone from the starting point), its velocity and its acceleration. So when the motor goes from coast to engaged, there is a big jump in the "stall" torque (this is the torque that appears to keep something locked in place). This torque is about 0.1Nm, and the max rated torque for the EV3 motor is 0.17Nm, which gives us a very narrow range. If there are any ideas or thoughts on how to fix this issue (probably no way to overcome it using that EV3 motor (?), it's all very much appreciated! Thanks very much — Reply to this email directly, view it on GitHub <#1675 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABKQIJF7T6UQQ47RKS4YL433LIWVZAVCNFSM6AAAAACCQ5JPBSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGOJUHA4DMOI>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[Growflavor]

We are trying to simulate a spring mass damper. So we have an industrial motor, and when that industrial motor goes to reach it's set position, we want the LEGO EV3 motor to resist that motion

@reillys9
side note: I am curious about the layout of this setup, would you be willing to share a photo or two that my son & I could examine together for conceptual/brainstorming fun?

Thanks for this discussion...I see I'll need to keep starting "stiction" in mind.
I did not think of that when viewing the interesting graphs for various LEGO motors at this page: https://philohome.com/motors/motorcomp.htm

PDF save of the above page in case it ever disappears: LEGO9VTechnicMotorsComparedCharacteristics.pdf

[reillys9]

Hello,

Untitled.video.-.Made.with.Clipchamp.1.mp4

I have attached a picture and a video of where we're currently at.

The PLC (Programmable Logic Controller) we are using is the X20CP1584, the drive is an ACOPOS P3 (I can't remember the model number of the top of my head) and then motor is an 8lsa25 motor. All those components come from a company called B&R, who specialise in creating industrial automation solutions.

The goal of the demonstrator is to show off some of the incredible function blocks that B&R have created, that have powerful methods of modelling and controlling systems.

We need to create an adaptive and dynamic system to test these function blocks on, so we choose LEGO as it's very easy to rapidly prototype with it, it's innate excellent visual demonstrator ability and the its overall ease of use.

I have to say, seeing LEGO recreating actual engineering principles is astounding. I bought a book, "Unofficial LEGO Builder's Guide, 2nd Edition" by Pawel Sariel Kmiec, when checking the feasibility of building the system out of LEGO, and I was astounded to see the incredible things in which LEGO can demonstrate. It's definitely well worth getting if you have a Technic set.