Tarot 4108 380KV

[Swann_Schilling]

Thanks a lot for the info…I just found the AS5600 encoder? Also AliExpress and dirt cheap!!

I guess I will go for those GB4008, if you can confirm they are well suited…a common ground is always nice, that way we might also be able to exchange parts, so we safe some time 3D modeling.

Is there anything that you can show, because I am also working on a smaller humanoid robot…

The bigger robot arm is still early WIP…but already loads of fun!

I got this post hidden, the first time I posted can you advice me on where this system is sensitive, when it comes to links? Probably just vending sites like AliExpress? And links to other forums or YouTube are safe to post?

[Swann_Schilling]

OMG this forum is really bad…I cannot even link my own prints, so I guess this is just meant to be strictly SimpleFOC!

Maybe we should connect on Discord?
Feel free to add me…

Mickey666Maus#3529

Oh but it looks like they are really looking into the content and are restoring post, if the links are not harmful…maybe I will just have to adjust?
But feel free to add me on Discord anyways, there are some nice channels, I was actually a bit confused when I realized that SimpleFOC is not on Discord…

[Antun_Skuric]

@Swann_Schilling

OMG this forum is really bad…I cannot even link my own prints, so I guess this is just meant to be strictly SimpleFOC!

Hey @Swann_Schilling,
The forum is automatically removing the posts which it finds questionable
I am not sure whay it has chosen your posts, maybe because the links, I am not sure.

Anyway it should be fine now.
Feel free to contact me directly if you have any other problems posting!

[Swann_Schilling]

@Antun_Skuric Thanks a lot, I was a bit shocked first…but then I realized that the post are getting checked and being put back, once it is clear that there is no harmful content! Nothing wrong with that, and also very nice to let me know directly!

I had a post when I started up that allowed only two links…so I guess when post with links are being removed, I should just stay patient and the will come back after the were being checked, right?

[runger]

@Swann_Schilling

And a low KV rated motor will be more sensitive to a voltage change and spin slower at the same voltage, but will also produce more torque at the same volts and draw more amps?

A great question, and if David has time to give some thoughts (or anyone) I would also be very interested…

I think KV is a kind of artificial measure that is determined by the motor’s impedance. I don’t think low KV motors necessarily have higher torque, just different torque-speed curves. You can get better torque out of them at lower speeds maybe?

Basically the way I see it, the current through the windings determines the strength of the magnetic field, and thus the (somehow theoretical) maximum torque.

But given sufficient voltage (again in theory) you can put any current through any winding (up to physical limits) - so the winding resistance determines how much voltage is needed to obtain a particular current, but also, and perhaps more importantly, the resistance is a component of the impedance, which determines how quickly the current can rise through the windings.

How all this relates to torque is even more complicated, since in a “static” situation, i.e. if the motor is not turning but the windings are energised, the magnetic pull will be “outwards from the centre”, and the torque in terms of rotation will be quite low.

To turn the motor, you energise (commutate) the windings in such a way that the magnetic pull acts “sideways” on the rotor, ideally at 90deg (electrical angle). By doing this in exactly the right pattern and speed to always match the rotors current position, and keeping the electrical angle at always 90deg, you “pull” the rotor around with maximum torque - that’s what’s so cool about SimpleFOC.

So I think the winding resistance has more to do with the speed you can attain - low winding resistance means lower impedance, means faster rise-time of the current. And if the resistance is too low, it means the current can easily rise too high, and everything (motor, cables, PCB traces, MOSFETs, etc…) can burn up, unless you build in some protection. But it also means you can spin the motor quickly with good torque, since each (short) commutation step will deliver enough current.

On the other hand high-ohm windings mean slower speeds, and if the resistance is high enough, it simplifies the control electronics and software because you don’t need the current limiting. It also means you can “hold” positions better, and presumably it is easier to commutate the motor slowly and smoothly.

This is why these motors are used more with SimpleFOC, because the current limiting is work in development, and you can run the gimbal motors at their native voltage without fear of burning things down. Low ohm windings have to be run at very low voltages, which comes with its own set of problems…

[runger]

@Swann_Schilling

I guess I will go for those GB4008, if you can confirm they are well suited…

Hey, it depends your task… for a humanoid, I dare not say, I will leave that for you to find out and then learn from you… at the moment I am still on wheeled robots.

Anyway, I set it up today:

The motor:

Mounted to a “holder” and with Magnet attached:

With encoder mounted:

On its desk-stand:

as promised a short Video of the GB4008 in action:

SimpleFOC on Emax GB4008

(My first ever YouTube video. Please forgive my incompetence.)
Note the 1m/s speed setting demoed in the Video is assuming 6cm diameter wheels - I should have mentioned that.

[David_Gonzalez]

@Swann_Schilling

am I right assuming that a gimbal motor will produce less torque at the same volts?

Yes. Generally speaking the best way to think about BLDCs is in terms of current, so yes, to achieve a higher torque/ current in a gimbal motor higher voltages are usually required because of their typical higher phase resistance.

Furthermore, gimbal motors are optimized for low cogging torque (for fine and smooth movements) and not optimized for peak power, while drone motors are optimized for high power, not low cogging torque.

@runger

I think KV is a kind of artificial measure that is determined by the motor’s impedance. I don’t think low KV motors necessarily have higher torque, just different torque-speed curves. You can get better torque out of them at lower speeds maybe?

You can approximate a brushless torque with the following equation:

This is a great way to visualize that having the same current on two motors with different KV will produce different torques.

[Antun_Skuric]

Hey guys,
I wanted to add a bit of info about the torque and KV. I think it is an interesting topic!

@runger

I think KV is a kind of artificial measure that is determined by the motor’s impedance. I don’t think low KV motors necessarily have higher torque, just different torque-speed curves. You can get better torque out of them at lower speeds maybe?

Yes it is an artificial value that is easy to identify empirically. KV has a unit which is rpm/Volt. To find a kV of the motor you spin it with 1 Volt and measure the velocity it reached and than just divide the two and you will get the KV:

KV = velocity_in_rpm / voltage;

@David_Gonzalez

You can approximate a brushless torque with the following equation:

This is a great way to visualize that having the same current on two motors with different KV will produce different torques.

This equation is true for static conditions. It is relatively easy to derive it also. The power is defined as effort times velocity, for linear movement it is P = F.v and for rotation it is P=T.w where T is torque and w is angular velocity. So if you want to calculate the torque you will need to calculate T = P / w. And in the case of electric motor, the electrical power can be calculated as P = I.U, so current times voltage. So at the end you can find the equation

T = I.U / w = I / (w_rpm/U * 2pi/60) = I / KV * 60/2pi

You can see here that the torque of an electric motor is directly proportional to the current:

T = k.I

and in many cases that is what you will control, not really the torque but the current. Or if you really wish to control the torque, you can measure it with some simple method for 2 or 3 points and find the value k.

[Swann_Schilling]

So I am not at all someone that knows how all this works, but I had some rules of thumb to kinda get a feeling of what is going on…maybe you can take a look and let me know if I am horribly wrong?

• use kt = 8.27 / kv as a good approximation for torque constant (Nm/A) of a motor
• use kv * v_in to determine no-load rpm on motor for given supply voltage
• use T% = 1 - (rpm / no_load_rpm), limited from 0 to 1 to estimate usable torque percentage during the fastest move of the motor
• use C = (T / T%) / kt to determine required current for the given output torque considering possible torque output of the bldc motor at given speed

The first one came from Oskar Weigel who started ODrive, the second was confirmed above by @Antun_Skuric …the other two are just piked up from a discussion over at the ODrive forum!

Since my application is not set in a specific scenario, I am currently still trying to see what different motors and mcus are capable of and how I can deploy them best!

I will have to read the last part that @Antun_Skuric posted a few more times I guess…
I am really still learning and am trying to sponge up everything I can comprehend!

[Swann_Schilling]

Btw great video and really awesome setup @runger I already subscribed and am hoping you will put out more content, it is really so much better to have some visuals when talking about a certain setup!!

[runger]

Hey thanks so much! I have to be honest I’m not sure I’m made for YouTube, but I will do my best. I certainly have to get a better lighting setup if I’m to make more videos…

@Antun_Skuric and @David_Gonzalez thanks for those explanations - my background is computer science, I did robotics from that side, so the mechanics and things like motors are things that I’m exploring now as a hobby, but never covered at university… it’s a vast world of knowledge, and I often have the feeling that I’m missing the “101” courses that provide the foundation for all this.