Help Choosing Rotary Torque Sensor for Low Torque Accuracy

Hi, I am currently working on a project in the Mobile Robotics Lab that requires a rotary torque sensor. We’re considering the ATO-TQS-D03 rotary torque sensor (100 Nm version) with a 4-20 mA analog output.

From my understanding, 4–20 mA gives 16 divisions, and with interpolation, we can get 10 subdivisions per mA, resulting in 160 divisions total. So with a 100 Nm range, the minimum readable value would be 0.625 Nm.

This is acceptable for testing actuators above 60 Nm. However, we also work with actuators of 8 Nm, where 0.625 Nm resolution is not satisfactory.

So before we proceed with procurement, I’d like to know:
1. Can the lower and upper sensing range be altered before use (e.g., 4 mA = 0 Nm and 20 mA = 20 Nm) to increase resolution for lower torques?
2. If not, is there any feasible solution for observing smaller changes (e.g., 0.1 Nm) while testing smaller actuators?

Any help or clarification would be great! Thanks!

Cassius wrote:Hi Cassius,
Thanks for your interest in our product.
For the 100 Nm model, the minimum readable value is 0.625 Nm with a 4–20 mA output. You do not need to alter the sensing range of the sensor. However, the ATO-TQS-D03 cannot be reprogrammed to output a scaled signal like 0–20 Nm. If you need to measure torques around 8 Nm more accurately, we recommend purchasing a 10 Nm range model.

ATO wrote:Thank you for the details answer! I have a few follow-up questions:
1. What is the minimum readable value of the sensor for the full range with default settings?
2. We want to get a minimum readable value of 0.1 Nm with the same 100 Nm sensor. Is that possible?
3. If not for the full range, then can we alter the working range to improve the minimum readable value?

Waiting for your response.

Cassius wrote:Cassius, The absolute accuracy of the ATO-TQS-D03 is 0.1% F.S., and the 4–20 mA output is precise enough to reflect the sensed torque. So, in theory, the minimum readable value for the 100 Nm model is around 0.1 Nm. However, please note that this accuracy may be affected by interference and noise in practice.
Again, this model cannot be reprogrammed to output a 0–20 Nm signal. For higher precision in the 8 Nm range, we suggest using the 10 Nm range sensor.

ATO wrote:Understood. One last question: What would be the minimum readable value for the 200 Nm and 300 Nm models? We will decide accordingly.

Cassius wrote:For all range models of the ATO-TQS-D03, the absolute accuracy is 0.1% F.S.