$300 120Nm robotic actuator from aliexpress - ROBSTRIDE04

[@ZedDead-n3s]

May know this already: * Rated current on actuator is peak phase current (which is not related to supply current unless at peak rated speed of motor at 48V, which is kinda unachievable) * Therefore your supply should handle peak torque (even at 30V) at low velocity, so just dial down those gains. * Your supply may have cut out due to an over voltage event due to BEMF generation during backdriving of the actuator (not an issue so mych with battery powered) * To handle these BEMF event your need a "regen shunt" (check o-drive for that) * The BEMF exhiibits itself on the voltage bus by exceeding the voltage setpoint of the supply. * You'll note when the weight is being lowered, due to the potential energy of the lifted mass, the current in the supply should drop ~0amps at times. * the above is predominantly just a warning so not to blow the control board or other devices connected to the voltage bus if they're not able to handle the transients and over voltage events. * testing at 24v, and getting a baseline of behaviour is safer, you have a much higher chance of damaging the controller when running at max operating voltage and then having a large Regen event (backdrive)

[@kindaovermyhead]

this is seriously very helpful, thank you!

[@ZedDead-n3s]

@@kindaovermyhead All good, keep at the testing and I'll be folliowing the progress, you're doing good. Shout out if you hit any hurdles, happy to give direction. Breaking things is all part of the process.

[@austinbetters8730]

​@ZedDead-n3s zeds dead baby, pulp fiction and the music big fan

[@gerokatseros]

@@ZedDead-n3s Sir, i face the same problem ... when my motor accelerates with some weight on it, and i want it to stop in a specifiv position , it shuts down when it is supposed to stop. I was thinking that there is reverse voltage that with the one added from the one that comes into the motor results in low voltage so the motor shuts down. In the odrive they give you a breaking resistor. But in my motor i don't see how i can implement it. Is there another way to avoid this problem? what is the regen shunt that you mentioned above?

[@gerokatseros]

Sir, i face the same problem ... when my motor accelerates with some weight on it, and i want it to stop in a specifiv position , it shuts down when it is supposed to stop. I was thinking that there is reverse voltage that with the one added from the one that comes into the motor results in low voltage so the motor shuts down. In the odrive they give you a breaking resistor. But in my motor i don't see how i can implement it. Is there another way to avoid this problem? what is the regen shunt that you mentioned above?

[@jamieclarke321]

This video legit got scary towards the end. With the risk of smashing the tile floor, kitchen cabinets, counter tops and then standing next to the weights with socks on at the end really got worrying for me

[@kindaovermyhead]

cookin

[@GLAJMAN]

He was wearing safety squints. It's fine :D.

[@TheFlatronify]

Yup, that thing can basically send those disks flying in an instance. Awesome! :D

[@josephsmith601]

​@kindaovermyhead I guess you live up to your channel's names 😂😂😂

[@jksjrgfpsjgr]

@@kindaovermyhead some constructive criticism, you designed the printed part as if it was to be milled from metal, the hub area stepping down like that is a weak point. if you keep the part that is clamped onto the motor as thick as the whole arm, you will compress the printed part with long AF screws and add to its z-strength instead. that actuator looks like a beast, and the software intuitive enough the figure it out, but you also get full control over everything, at least thats what it looks like. great video!

[@vincentnguyen8083]

A very good and incredibly cheap actuator, even after the export tax (it's something like 165$ in china). If you open it, it's also quite interesting to see how they optimized the design for mass production

[@hoodio]

do you have a link to some pictures, that sounds really interesting

[@conorstewart2214]

A stronger mount and limiting the actuator speed would probably be a good idea, it was moving a bit fast for that kind of load.

[@mosth8ed]

That turned out both exactly, and better than I expected at the same time.

[@Alex_Soare]

Love how it ends🫠

[@ulforcemegamon3094]

Reminds me of the ones Unitree robotics use in their heavier quadrupeds , i recall that the heavier quadrupeds do have motors that achieve 120nm (or more) , thought these motors cost 1500 dollars without taxes or shipping fees

[@gerokatseros]

so did you make any more checks? does the actuator actually handle 120Nm or near this value? What commands are supported throught the can bus?

[@kindaovermyhead]

the commands are available in the docs on their site, and i’ll find out soon if it’s capable but gotta improve my setup first cause..safety :)

[@gerokatseros]

I see they have a ROBSTRIDE03 as well .. this one looks very interesting too!

[@kindaovermyhead]

it does, i hope to try it!

[@3ffrige]

Actually, can you do me a favor and program that thing to wave around 20lbs at full speed?

[@kindaovermyhead]

will be doing that dw

[@Arek_R.]

$300 product and the give you only 10cm of cable like come on, spend 50c more and give a meter...

[@JarrodM29]

I just got a robstride04 and am having the same issue you did, it's not connecting. Can you explain what you did to get your to connect?

[@squa_81]

Question : Do you think this could output 30Nm for ballscrews?

[@makex_se]

Ham and cheese outro is interesting :D

[@FilmFactry]

So this is driven by CAN signals. The "controller" is built 8into the drive? You are just supplying Voltage and CAN?

[@kindaovermyhead]

yes just supplying power and a CAN signal via their own CAN to USB adapter

[@marceloenriquechirinosmedi6002]

120 Nm, in kgf/cm please

[@kaydenk]

I have the O3 and just keep getting this error: 1 [2024/11/05 21:04:18] Connect overtime, try baud rate: 9600 2 [2024/11/05 21:04:19] USB-CAN module configuration failed, please check physical device! I have tried swapping can high and low wires, unplugging usbc etc. I was able to get it connected once but now it wont connect at all, any ideas?

[@daveoconnor1413]

I have exactly the same problem! Does anyone have a solution? I tried contacting Robstride for support, no reply yet.

[@kaydenk]

Yeah you have to change the baud rate I fixed it

[@daveoconnor1413]

@@kaydenk where did you change the baud rate? In windows settings or somewhere else? What did you set it to?

[@Ray88G]

120Nm because of gear reduction

[@kindaovermyhead]

yes, it advertises a planetary gear reduction

[@maalonszuman491]

realy nice!! There is also a much cheaper one the xiaomi actuator

[@tfoutfou21000]

i am the landlord , it will be 4000$ to change those breaked tiles 😅😅

[@Śiśna3633]

"breaked", Uneducated moron like you can't be the owner of any house.

[@docdailey]

does that continuously roatate?

[@kindaovermyhead]

yes

[@docdailey]

@@kindaovermyhead perfect. Would be great to drive a small capstan.

[@AnnaVannieuwenhuyse]

@@docdailey sure, a small capstan lol

[@docdailey]

@@AnnaVannieuwenhuyse exactly what I need. was going to repurpose hoverboard motors... but i need tight control of position. Multiple capstains working in concert.

[@regun2434]

Dozen of these for robodog 🤩

[@RchGrav]

Sweet actuator... what broke? which mount?

[@kindaovermyhead]

nothing broke, I believe the actuator drew more power than my power supply could handle and disabled.

[@Tosti_bakker]

So did it manage 120 Nm or not? what was that ending lmao

[@kindaovermyhead]

i will be finding out soon with a stronger testing setup

[@DgtlCnsltng]

No statement anywhere about backlash...

[@kindaovermyhead]

knew I forgot something! Its backlash isn’t advertised, it feels near zero, but I didn’t have my dial indicator on me to test this time around

[@DgtlCnsltng]

@@kindaovermyheadfor lots of applications this is more important than pure torque. It is a planetary gearbox so probably not that good I'm afraid...

[@vincentnguyen8083]

This actuator is meant for dynamic applications (e.g. robot dogs) where backlash is not as critical as transparency (i.e. reflected inertia/efficiency) and torque

[@gtsianakas]

​@@vincentnguyen8083This means that using it for a robotics arm is a bad idea? Any thoughts on what kind of repeatability should anyone expect from an arm out of these?

[@vincentnguyen8083]

@@gtsianakas an arm out of these would certainly be possible, and this is what some companies do to accomodate impact-rich environments (see the DuaArm from Duatic or anymal's robot arm, which both use QDD actuators with a planerary geartrain). However to get good repeatability you will need either a well-matched gearbox, or an external control-loop (e.g. visual or precise output encoder) to mitigate backlash

[@kadly6657]

Good but not great. Snapping jump cut needs work.

[@johnmccardle]

Market Basket #22 in Tewksbury?! How was that cheese?

[@MADMAX7330]

Seems like a great motor, but $300 is awfully expensive, given you'd need like 4 of them for a humanoid, if not more. What do you plan to use this for?

[@kindaovermyhead]

i mean you’d need at least 5

[@engineeredaf1920]

probably should have got it in position and then slowly added the weight

[@conorstewart2214]

There is a difference between static and dynamic load. Just because a motor can hold a load that doesn’t mean it can move it.

[@engineeredaf1920]

@@conorstewart2214 theyre probably rating a static load tho

[@pixiepaws99]

This is just a CAN bus geared servo. Nothing special here... The gain looks a bit low and its position holding looks really bad (by CNC machine servo standards anyway) so you need to adjust the PID parameters.

[@kindaovermyhead]

yeah the pid tuning is my own and not great yet, i’ll find out soon if it’s special or too good to be true

[@TheBozn]

idk at what voltage you ran it but of course you cant hold 120Nm or even 12Nm with that 2 cm leverage... you are trying to counter equivalent to over 1000 pounds with your fingers there.. Itd probably be impossible for you to do even with a normal sized wrench(30cm). Cool motor tho. I would love to see the gearing and more of the inside. I wonder if its just one set of planetaries or multiple and what kind of machining and material quality its for that price. Sometimes chinazon really delivers

[@5doolar]

u print that at 0.24 layer? what filament u use?

[@kindaovermyhead]

yeah 0.24 and generic pla for a temporary mount

[@JeffreyMartin]

why is this video only available in 360P / potato resolution?

[@Nobody-Nowhere]

are you new in youtubes?

[@neeksor]

pootube transcodes original to many formats.. lower quality takes less time to process

[@Picks_Productions]

Because that’s all your device allows 😂

[@FLORIDIANMILLIONAIRE]

Simple fact If it weighs a lot it cannot be good for robotics im making an actuator that will be super light weight and strong

[@Model3GenerativeANdroid]

keep it up 🥰

[@nicecrash]

The software looks like the Xiaomi cyberdog Motor Software...be careful 🤞

[@gerokatseros]

why did you say to be carefull? what is wrong with xiaomi software?

[@Nobody-Nowhere]

You could have just spent 3 months 3d printing one that would have performed at least 5 times worse. ... and now i checked your earlier videos and you already tried that :) There has been this trend in youtube of people trying to 3d print these.. guess its "content".

[@satibel]

also 300 bucks vs like 50 bucks (maybe less if you aren't prototyping)

[@bartoszswieciak8341]

What's wrong with people who do this on their own?

[@kindaovermyhead]

this content is in search of the best actuator possible for the price, if it’s gotta be my own, so be it

[@qo7ems]

W ENDING

[@wizrom3046]

120 Nm... I know! Why not make a hub out of flimsy 3D printed plastic goop 😑

[@kindaovermyhead]

just gotta use stronger goop

[@rbyt2010]

probably interesting but couldn't watch. video made me sick