How do you even quantify 'feel' when comparing two actuators?

Harmonic drives, quasi-direct drive, series elastic actuators, motor selection, gearing, and everything that makes joints move.
thomas65
Posts: 59
Joined: Sat Feb 14, 2026 11:59 am

Re: How do you even quantify 'feel' when comparing two actuators?

Post by thomas65 »

I dealt with almost this exact situation. Quasi-direct drive (QDD) actuators use lower gear ratios (roughly 6:1 to 10:1) paired with high-torque-density motors, trading some peak force density for backdrivability and low reflected inertia, which matters a lot for impact tolerance and fall recovery. For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon.
Watching this space closely since 2019.
jwang
Posts: 189
Joined: Wed Jan 22, 2025 7:28 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by jwang »

Slight correction, though the overall point stands: Reflected inertia is what a limb 'feels like' to the outside world through the gearbox - a high-ratio harmonic drive reflects a lot of the rotor's inertia back to the joint, making the limb feel stiffer and less forgiving on unexpected impacts.
olga_lind
Posts: 170
Joined: Tue Dec 24, 2024 12:11 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by olga_lind »

@jwang Here's the relevant bit as far as I understand it: Torque sensing is frequently integrated directly into the actuator housing via strain gauges near the output flange or spring element, rather than bolted on as an external sensor, to save weight and reduce noise from mechanical slop. Reflected inertia is what a limb 'feels like' to the outside world through the gearbox - a high-ratio harmonic drive reflects a lot of the rotor's inertia back to the joint, making the limb feel stiffer and less forgiving on unexpected impacts. Makes me wonder how this looks in another five years.
benjaminsanchez
Posts: 180
Joined: Fri Apr 18, 2025 1:58 am

Re: How do you even quantify 'feel' when comparing two actuators?

Post by benjaminsanchez »

This raises a question for me - For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon. Encoder resolution requirements scale with how tightly you need to control low-speed motion - coarse encoders are fine for open-loop swing phases but cause visible stutter during precise placement or fine force control.
rivera14
Posts: 54
Joined: Thu Mar 19, 2026 1:48 am

Re: How do you even quantify 'feel' when comparing two actuators?

Post by rivera14 »

Follow-up question though - Anti-backlash techniques in production actuators range from mechanical preloading (spring-loaded gear meshes) to purely software compensation that models and corrects for known backlash in the control loop. For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon.
Ex-automotive, now full-time robots.
ethan.lewis5
Posts: 65
Joined: Sat Apr 04, 2026 7:35 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by ethan.lewis5 »

@rivera14 Minor factual note: Motor cogging torque (the 'notchy' feel from magnet-to-slot interactions) shows up as small periodic torque ripple that can translate into visible joint jitter at low speeds - it's usually mitigated with skewed magnets, better current control, or software compensation tables. For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon.
Opinions my own, not my employer's.
garcia51
Posts: 94
Joined: Fri Oct 24, 2025 3:36 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by garcia51 »

@ethan.lewis5 Small correction on one detail: For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon.
they/them
ethan17
Posts: 138
Joined: Mon May 19, 2025 8:21 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by ethan17 »

@garcia51 This lines up with my experience. Anti-backlash techniques in production actuators range from mechanical preloading (spring-loaded gear meshes) to purely software compensation that models and corrects for known backlash in the control loop. Totally unrelated but has anyone else noticed how fast component costs are dropping this year.
"Torque is a lifestyle."
william.moreau6
Posts: 23
Joined: Thu Jul 09, 2026 12:47 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by william.moreau6 »

Slightly off-topic, but related: It's common for a single humanoid to mix actuator types by joint - harmonic drives or hybrid geared actuators at the hips and shoulders where sustained torque matters most, and lower-ratio QDD-style actuators at ankles and knees where backdrivability and impact tolerance matter more. Boston Dynamics' commercial electric Atlas switched fully from hydraulic to QDD-style electric actuation, ending up quieter and lighter with comparable or better dynamic performance than the old hydraulic platform.
jwang
Posts: 189
Joined: Wed Jan 22, 2025 7:28 pm

Re: How do you even quantify 'feel' when comparing two actuators?

Post by jwang »

@william.moreau6 One nitpick - Torque sensing is frequently integrated directly into the actuator housing via strain gauges near the output flange or spring element, rather than bolted on as an external sensor, to save weight and reduce noise from mechanical slop. Harmonic drives use high reduction ratios (often around 100:1), which gives excellent force density, stiffness, and thermal endurance for continuous-duty lifting - the tradeoff is low backdrivability and some backlash.
Post Reply