Very nice reducer, I was always under the impression that a planetary gear system would always mean serious backlash. Also that big airgap motor is very nice with the ODrive. Can't wait to see this thing pushed to its limits or used in some example application. Seems like this arm can be both fast AND strong, would be fun to see it hammer, throw or move something. Keep up the good work Skyentific!
Hi, I like the channel. However, I had problems finding how the various video's are related, for instance I watched the first video about the OpenTorque Actuator, but could not easily find this one. I tried many playlists, but no luck. One thing you can do is add links to other videos of the same topic/project. The links at the end of your video seem to be more or less random.
4 года назад
Very interesting video. Please use the proper unit for torqe, Nm. I know a lot of motor vendors still use kg*m in their spec sheets, but it has been deprecated with the introduction of the SI system many decades ago (1978 in Germany).
commenting to boost YT ratings because i love your content. Also for those who are new to the content watch at 1.5 speed makes the pacing of the content slightly more engaging.
Nice, thanks for the test. I have a 24V version of the ODrive. What do you think should I pull up 10kg mass with this actuator with 24V Odrive if I put a 10cm diameter pulley on it? You measured 3.5kg @20A with 24V, so it is 1,143kg*m. If I count well, then if use a 10cm diameter pulley on the actuator, then 1,143+0,047/0,05 = 23,8kg. Am I right or missed something?
Thank you for your comment! You are almost right. If you will have a circular pulley, then: 1.143/0.05 = 22.7kg. (you should not use 0.047, as it is for unbalanced arm. You pulley would be balanced (because it circular))
Hi, I have a question, at min 9:10 you push the lever and it moves, because the current is low the holding torque is low, is there a way to force the current so the motor stays in that place? (increase holding torque). And why the reaction time of returning to the holding position takes so long? By the way awesome videos, thanks a lot!
Love watching these D.I.Y. servo and actuator videos and learning how things work. Wouldn't a stalled motor try to draw all the current it can get? Since you have limited the current to 10/20/30/40 Amps the power draw at stall should be 450/900/1350/1800 Watts (@ 45 Volts). I haven't played with BLDC motors or Odrive before, so if i am wrong please let me know.
Current limiting is done by reducing the voltage until "all the current it can get" is the target value. If you fed it the full 48V while stalled, either the power supply would blow a fuse, ODrive's transistors would burn, or solder joints would melt :) But I'm not sure exactly what's going on, because Skyentific's power measurements are higher than I'd expect, and it should only take a few volts at the motor to produce 40 amps so I don't know why the 24V ODrive wasn't getting the full torque.
Wow, 25Nm is impressive for that size plastic gearbox! And with such low reduction ratio, gear speed should be low enough for real world usability without worrying about melting them. How much does it weigh, excluding the weight of the motor?
Great video! I love the project and the approach. I am soon going to embark upon a very similar build. One thing that makes me wince almost as much as the sound of the aluminum arm hitting the scales is when people say decimals as if they are full integers. Saying 3.65 as "three point sixty five" can be misleading. This is "three point six five"! Why? How about this 3.2 "three point two", 3.20 "three point twenty*", 3.200 "three point two hundred". Ouch.
так допустим мотор 100kv 45v 8к1 редукция то по идее имеем примерно 560 оборотов максимум на выходе (при всех ваших подсчетах крутящего момента) нужно ли робо руке 560 оборотов в минуту скорость???? как по мне нужна большая редукция и на тех же слабых тока будем иметь больше момента (напрашивается эксцентриковый редуктор(циклоидальный))
Это разные подходы. Есть классический подход: большое передаточное число, и слабый мотор. Есть более современный подход: низкое передаточное число, мощный мотор. У них есть свои плюсы и минусы. С высоким передаточным числом, редуктор меньше потребляет при одинаковой нагрузке. Но если эта рука с такими редуктором врежется во что-нибудь, то скорее всего она просто поломает редуктор. А с низким передаточным числом, если она врежется, то просто мотор прокрутится, но актуатор не сломается. Поэтому на ногах роботов часто используют второй подход (с низким передаточным числом) что бы робот мог прыгать, не ломая себе колени при каждом приземлении. Сейчас стали появляться целые робот руки с низким передаточным числом, это дешевле, динамичнее, и надежнее.
Could someone explain to me how the max calculated wattage was only 153 watts at 40 amps but the max power is written 2800 watts? Would the power really be anywhere close to 2800 watts at 67 max amperage?
I am interested in the cost of this drive. My guess is that these may be better costed as a pair because the O drive runs two motors? So is it safe to say that this setup (Power supply (or lipos), fans, O Drive, two motors, bearings) cost about $150 per motor? Are you moving to Lipo power next?
What is the maximum torque ouput of the BLDC? Looking at the gearbox it should have an 8:1 ratio but at 40A if the actuator is able to exert 25.91Nm then thee BLDC must be outputting 3.16Nm. For 40A this would require a Kv of 12.64?
You're off by a factor of 10 there. The quick torque estimation formula is current / (KV/10), so that would be 126.4KV. But a more accurate formula is current * 8.3 / KV (from this article things-in-motion.blogspot.com/2018/12/how-to-estimate-torque-of-bldc-pmsm.html ) which estimates 3.32Nm for 100KV, which I think is the version of that motor in use here. The HobbyKing page lists the max current at 57A, so max torque should be about 4.73Nm.
You should measure the force with the arm in the vertical orientation so the weight of the arm doesn't contribute to the force. Get a real load cell that can handle 10+ kilos of force.
Please do not hesitate to leave the comments. This will help to boost this video! Thank you!
you can make someone to print gears in tungsten ;)
Great results. Imagine how cool the arm can be with quiet and fast movements this actuator provides
Thank you! I agree! It should be great! :)
Very nice reducer, I was always under the impression that a planetary gear system would always mean serious backlash. Also that big airgap motor is very nice with the ODrive. Can't wait to see this thing pushed to its limits or used in some example application. Seems like this arm can be both fast AND strong, would be fun to see it hammer, throw or move something.
Keep up the good work Skyentific!
Excellent work sir. Super progress.
Its inspiring to see what is possible today with robotics by a dedicated individual; keep it up!
nice scientific approach. good to someone building on the work of others instead of scratch design from ground up cheers
Hi, I like the channel. However, I had problems finding how the various video's are related, for instance I watched the first video about the OpenTorque Actuator, but could not easily find this one. I tried many playlists, but no luck. One thing you can do is add links to other videos of the same topic/project. The links at the end of your video seem to be more or less random.
Very interesting video. Please use the proper unit for torqe, Nm. I know a lot of motor vendors still use kg*m in their spec sheets, but it has been deprecated with the introduction of the SI system many decades ago (1978 in Germany).
commenting to boost YT ratings because i love your content.
Also for those who are new to the content watch at 1.5 speed makes the pacing of the content slightly more engaging.
The set point was clearly over 9000 :P
The numbers look amazing. I wonder if you can scale this down to use for the smaller (further from the base) joints.
Nice, thanks for the test. I have a 24V version of the ODrive. What do you think should I pull up 10kg mass with this actuator with 24V Odrive if I put a 10cm diameter pulley on it? You measured 3.5kg @20A with 24V, so it is 1,143kg*m.
If I count well, then if use a 10cm diameter pulley on the actuator, then 1,143+0,047/0,05 = 23,8kg. Am I right or missed something?
Thank you for your comment!
You are almost right. If you will have a circular pulley, then: 1.143/0.05 = 22.7kg. (you should not use 0.047, as it is for unbalanced arm. You pulley would be balanced (because it circular))
I always laugh when I see your selection of foods in the videos. What's your favorite?
Of course potatoes! :)))
Katerina is alive?
Perfect!
Thank you!
Could you please link the power supplies that you use? I'm not sure what ones to use myself.
Hi, I have a question, at min 9:10 you push the lever and it moves, because the current is low the holding torque is low, is there a way to force the current so the motor stays in that place? (increase holding torque). And why the reaction time of returning to the holding position takes so long?
By the way awesome videos, thanks a lot!
@Skyentific
Love watching these D.I.Y. servo and actuator videos and learning how things work.
Wouldn't a stalled motor try to draw all the current it can get? Since you have limited the current to 10/20/30/40 Amps the power draw at stall should be 450/900/1350/1800 Watts (@ 45 Volts).
I haven't played with BLDC motors or Odrive before, so if i am wrong please let me know.
Current limiting is done by reducing the voltage until "all the current it can get" is the target value. If you fed it the full 48V while stalled, either the power supply would blow a fuse, ODrive's transistors would burn, or solder joints would melt :) But I'm not sure exactly what's going on, because Skyentific's power measurements are higher than I'd expect, and it should only take a few volts at the motor to produce 40 amps so I don't know why the 24V ODrive wasn't getting the full torque.
@@dekutree64 thanks.
Wow, 25Nm is impressive for that size plastic gearbox! And with such low reduction ratio, gear speed should be low enough for real world usability without worrying about melting them. How much does it weigh, excluding the weight of the motor?
What is the reference of that actuator sir?
Can I increase holding Torque of stepper motor by any way. I am using it for my industrial robotic arm.
Great video! I love the project and the approach. I am soon going to embark upon a very similar build. One thing that makes me wince almost as much as the sound of the aluminum arm hitting the scales is when people say decimals as if they are full integers. Saying 3.65 as "three point sixty five" can be misleading. This is "three point six five"! Why? How about this 3.2 "three point two", 3.20 "three point twenty*", 3.200 "three point two hundred". Ouch.
так
допустим мотор 100kv 45v 8к1 редукция
то по идее имеем примерно 560 оборотов максимум на выходе (при всех ваших подсчетах крутящего момента)
нужно ли робо руке 560 оборотов в минуту скорость????
как по мне нужна большая редукция и на тех же слабых тока будем иметь больше момента (напрашивается эксцентриковый редуктор(циклоидальный))
он поймет тебя?
Понял :)
Это разные подходы. Есть классический подход: большое передаточное число, и слабый мотор. Есть более современный подход: низкое передаточное число, мощный мотор. У них есть свои плюсы и минусы.
С высоким передаточным числом, редуктор меньше потребляет при одинаковой нагрузке. Но если эта рука с такими редуктором врежется во что-нибудь, то скорее всего она просто поломает редуктор. А с низким передаточным числом, если она врежется, то просто мотор прокрутится, но актуатор не сломается.
Поэтому на ногах роботов часто используют второй подход (с низким передаточным числом) что бы робот мог прыгать, не ломая себе колени при каждом приземлении.
Сейчас стали появляться целые робот руки с низким передаточным числом, это дешевле, динамичнее, и надежнее.
How did you calculate the theoretical torque?
Could someone explain to me how the max calculated wattage was only 153 watts at 40 amps but the max power is written 2800 watts? Would the power really be anywhere close to 2800 watts at 67 max amperage?
I am interested in the cost of this drive. My guess is that these may be better costed as a pair because the O drive runs two motors? So is it safe to say that this setup (Power supply (or lipos), fans, O Drive, two motors, bearings) cost about $150 per motor?
Are you moving to Lipo power next?
Are you plaining to add cooling to actuator?
nice video
keep going!
What model of printer do you use?
I use Prusa i3 MK3S
would love to see the arm lifting some weight instead of pressing the scale
What is the maximum torque ouput of the BLDC? Looking at the gearbox it should have an 8:1 ratio but at 40A if the actuator is able to exert 25.91Nm then thee BLDC must be outputting 3.16Nm. For 40A this would require a Kv of 12.64?
You're off by a factor of 10 there. The quick torque estimation formula is current / (KV/10), so that would be 126.4KV. But a more accurate formula is current * 8.3 / KV (from this article things-in-motion.blogspot.com/2018/12/how-to-estimate-torque-of-bldc-pmsm.html ) which estimates 3.32Nm for 100KV, which I think is the version of that motor in use here. The HobbyKing page lists the max current at 57A, so max torque should be about 4.73Nm.
You should measure the force with the arm in the vertical orientation so the weight of the arm doesn't contribute to the force. Get a real load cell that can handle 10+ kilos of force.
Resolution ???
When you convert metric pointing with fingers to BES pointing with fingers, the conversion is 2:1. 😂
Resolution
Tiny scientific remark: torque is *not* expressed *per* meter. Leave away the "per".
You have an expensive oscilloscope and temperature meter...
It is not mine, this is equipment from my university.
@@Skyentific oh okay. Do I need a braking resistor on the odrive for robotic applications?
Акцент, как у русского
Точно! :)
@@Skyentific Раз такое дело, осмелюсь попросить русские субтитры от автора хотя бы для новых видео =)
Или английские, но от автора. Потому что гугл плохо распознает речь