Відео

Sorry, I haven't posted the source files

Thanks for the comment! My GUI interface is made with rQT and implemented to communicate with the ROS layer

can i get the code ?@@Jorgefer88

at minute 1:36, can you share the code?@@Jorgefer88

at minute 1:36, can you share the code?@@Jorgefer88

Odrive son muy buenos controladores, pero es cierto que son un poco caros. Yo estoy usando un controlador echo por mi basado en el proyecto de código abierto SimpleFOC y el precio de cada placa estaría en torno a los 45€, pero no vendo PCBs. SimpleFOC funciona con mucho hardware compatible y puede llegar a ser muy asequible en precio, te recomiendo que le eches un vistazo ;-)

@@Jorgefer88 conocí ese proyecto hace poco, pero no encuentro como comprar sus tarjetas por aliexpress u otro mercado donde las pueda obtener y que realmente cuesten esos 45 euros

@user-vc5ny6rn7g Los controladores que yo uso están fabricados y soldados por mi, no los vas a poder encontrar en ningún sitio. En aliexpress por ejemplo hay algunas PCBs que son compatibles con SimpleFOC. Te aconsejo que investigues por su página para ver cuáles modelos podrías usar. Otra opción (que es la que yo hice), fue fabricar PCBs desde cero y hacerlas funcionar con ese Firmware.

Muchas gracias 😄

Or you keep everything the same but you increase the gear ratio and the speed of the motor

@@ap2372 Thanks for the comment! Yes, increasing the stator diameter to 10010 would get much more torque, but winding all the actuators again would be very tedious. I'm currently working on making a new all-metal actuator with a small homemade CNC and I'm going to increase the ratio from 4:1 to 6:1 I will have less speed, but from my tests I think it is still enough.

@@Jorgefer88 I am building a motor (with little knowledge) how to calculate windings, gauge and voltage for 500RPM (and max torque)

Thank you so much! I am limited to a maximum of 5A in the driver. That's why I chose the AWG of the cable so that it could handle those amps without getting overheated. From there, try to introduce the greatest number of turns to achieve the greatest possible torque.

@@Jorgefer88 I see! That's a great idea

Thank you so much! I'll keep working!

First of all, thank you very much! Are you making a quadruped robot? What electronics/motors do you use? Making this actuator with 3D printing (PLA) has taken me many iterations to get it to work more or less reliably. To this day I still have problems with it, due to the heat and the efforts it has to endure, but it half works :-D I answer your questions: 1. The only piece that I have not built is the stator, it was purchased on Aliexpress. It could be made DIY with cut steel sheets, but I think it would not compensate for the work it takes to do it unless you want something very specific. 2. Small magnets create a "continuous" magnetic flow. This is called the "Halbach Array". Because I don't have a steel external yoke (it's PLA plastic), a lot of magnetic flux escapes outside. In this way it is concentrated inside, towards the stator, causing the torque to increase. Using Halbach Array or external yoke is similar, you can see the detailed explanation here: things-in-motion.blogspot.com/2018/12/the-advantages-and-disadvantages-of.html 3. Switching in FOC is done through 6 steps, yes. You can see the details of how the library I am using works: docs.simplefoc.com/foc_theory 4. Indeed, using more voltage could make it run cooler. It's something I was testing, but starting at 30V I burned a couple of PCB boards, I didn't take that into account when I designed them. That's why I'm currently limited by electronics to 24V. 5. The continuous torque that I can handle is around 2.5N·m, the maximum at 8N·m. I have not calculated the performance and the nominal power will be around 50W. I have considered on several occasions making it open source, but I am a little wary of what you say, that someone will commercialize it. I know that there are different licenses and I'm sure that some may suit me, but for now, it is still in continuous development. Currently I have bought a small CNC to be able to make the actuators completely out of metal, because I have reached a point with plastic and 3D printing, where I cannot make them 100% reliable and durable.

Bro obviously it's his hard work and he can't just publish it for free .... And even if you would ask him respectfully then maybe he would think of sharing his hardwork with you ..... But for that attitude..... No...

Bro obviously it's his hard work and he can't just publish it for free .... And even if you would ask him respectfully then maybe he would think of sharing his hardwork with you ..... But for that attitude..... No...

@@pukhrajdhiman8041 lol dumb take. You should feel bad.

​@@pukhrajdhiman8041 The open source community is full of hard working people who freely contribute to society.

Look at a channel called aaed musa, they did an internal cycloidal robotic actuator similar to this but open source

Thanks!! The stator is the only commercial part, purchased on Aliexpress

@@Jorgefer88 I see...are the STLs open-source by any chance? I'm also curious to know about the total price for all the parts used in the design.

Muchas gracias! Si, es una de las cosas que haría si empezara desde cero, porque me ha llevado (y me sigue llevando) mucho más tiempo del imaginado. Ahora estoy diseñado y fabricando unos nuevos actuadores de aluminio fresados con una pequeña CNC 😄

No has pensado en montar una empresa o solo es hobby ?@@Jorgefer88

Muchas gracias! Si, están diseñadas y soldadas por mi para este proyecto concreto.

Thanks!

Yes, it's a long job to do all that by myself. But I've learned a lot along the way :-D Now I'm trying to make the actuators out of metal to make it more reliable in the long term.

Thank you very much for your words! My idea is to learn everything I can from it. I'm working on making it in metal, but it's pretty slow work. In the future you will see the results!

It is model 3020 from the Vevor brand

thank's !@@Jorgefer88

The maximum torque I can currently generate is 8N.m at 24V I have tested the actuators with 32V and they generate 12N.m, but with that voltage my PCBs are not reliable and I have broken them.

@@Jorgefer88 Oh nice thank you!

Thank you so much! At first I tried to do what you mentioned, but in my case I can't. I have the magnet for the encoder and the fan on the shaft, so it is impossible for me. If you do not use a fan and use an off-axis encoder it can be done without problems.

@@Jorgefer88 You can use two 49E linear hall sensors for an off-axis encoder. Spaced 90 electrical degrees (when one is centered on a rotor magnet, the other is half way between magnets), subtract midpoint from the readings and use atan2 to get the angle. They need to be close to the magnets to get a good signal, but not too close or it saturates. Usually 1-2mm is good.

It is quite a time job. The important thing is to be able to make a robotic actuator that works correctly.

Thanks so much!

Thank you so much! Yes, I glue the magnets directly to the plastic with cyanoacrylate

And what dimensions is the rotor core?

The graphical interface is specifically designed by me too. This is made with QT