Excellent work! I started working on a similar concept several years ago, but while it taught me a lot about PCB design I didn't have the full skillset to complete it. The projects I was hoping to use it on all went unfinished, but now maybe I can get them done!
It would be interesting to see how easy it is to add the CLN17 to an existing product. Without any change, I assume the motor will start to make sure it doesn't lose a step by itself. Of course, the full benefit is not just to mod the stepper but having a central board to discuss with the stepper.
I think from the point of view of integration into a printer for example, it should be quite simple. For example you can make the driver working on step-dir interface as an clone of the commands from TMC2209. Then you can use it as a normal driver. This way it can replace the limit switch in case of ACTUALLY accurate sensorless parking. Also if any error occurs in the system, also it can create a proper interrupt and save the printing. This is a general concept for now, but I think it will be realised after some time) A lot of work is required 😅
Please obsolete my stallGuard and make encoderless stepper motors ancient. When closed-loop stepper motors with Trinamic-grade silence and accuracy become readily available at less than $20 each, it will be an “innovation” and beat industrial AC servos. I’ve subscribed hoping this is the very beginning of such wonderful new era.
I really hope to do so) Accessibility requires mass production. Mass production requires a large community. A large community requires versatility. So here we are)
Looking forward to the larger iterations (nema 34, etc). Make sure that what you come up with WILL work with Mach3 & 4. Opens up the whole market of people like myself with CNC mills, lathes, etc.
Currently in development is an advanced version for Nema 23 34 42 including hybrid 3 phase stepper motors. Im not really familiar with Mach3 but could potentially do support in the future) Join the community to give valuable advice when it comes to this)
As I love this subject, the bad thing is that I don't know English, even if it is to consume a written text, with wonderful documentation like this CLN17 driver, it would be great if we had a class on this, it could even be paid content, which so much beautiful information.... It would also be for larger engines nema 34 for example... Maybe one day I will learn
With a lot of struggle I'm manufacturing my Router, a little without money, but I'm almost finished, I have a Nema 34 85kgf engine, I can implement this on these engines, I'm going to use Linux CNC
Great work that I can noe even fully appreciate. Got here looking for open firmware, as most of available nema17 stepper closed loop controllers have bugged and not supported codebase. Would be glad to purchase when it comes out.
I thought there was a closed loop stepper driver board/platform project already, and I've been wondering why we haven't seen it widely implemented in desktop 3d printers. How does CLN17 differ from the S42B that came out a few years ago, for example?
I guess the main difference is the project's opensource and much more features. You can visit the project's website, there is a section including a summary table between all existing closed loop drivers of this type. To the question about prevalence, I think the answer is simple. Almost all such drivers are not good enough and/or expensive enough for mass use. So one of the goals of this project is to solve these problems by making the solution universal and mass-produced.
Sorry for my ignorance but This has become quite the time sink and before I delve into this project any further can you tell me how it compares to the MKS SERVO42, SERVO57 and AS5600 projects for high accelerations in 3d print applications?
Very cool! Contact a company like mellow, LDO, fysetc etc and ask if they can make them, I am interested in closed loop steppers but I do not want to make them myself.
The prototype source files are already published on github. I don't recommend to build it manually as the driver has 0402 parts. Also the current version does not have good software yet 😞
I'm aiming for 48v 10a operating mode. There may be limitations that I don't know about yet, but from the TMC5160 driver documentation it seems possible
Of course, I used Autodesk Fusion360 for rendering and then overlay of several images in Adobe Premiere Pro. The PCBA 3D model I took from KiCad. It's pretty simple, but not so convenient)
Can you explain what advantages your nice project over the SERVO42C and SERVO42D has? I am always interested in nice project like yours. I can inly speak for the 42c... and let me say, it works great with klipper in my hypercube evolution. Magnetic orientation field sensor works good. And its unbeleavable silent. Like tmc2209 drivers. Maybe more silent. I can hear mechanicial sounds on my rails, ive never noticed before.
That's a good question. Well, first of all, there are a lot more interfaces that are really useful. For example can-bus can be obtained without fuss with an external uart converter. Also there is USB, which is convenient for flashing or debugging during programming. And at the same time the microcontroller is much faster) Maybe after some time I will make a version that will also support FOC and will work better than tmc2209, but it takes time
@@creapunk the can bus sounds promising. But i think that klipper can not coexist in the cpu? Usually we have to compile a klipper firmware for mcu like ebb36 and so on. I think your firmware uses special algos with magnetic angle sensor check and so on. To reach people interested in nema17 servos i think it would be nice when it is compatible with klipper firmware over can bus for 3d printer.
Okay. I checked all that well documented informarions on your homepage. Nice job, appreciate it. Its interesting how deep and powerful the functions are.i like that. I found one difference. You are using a tmc2209 and the servo42c's cpu takes controll over a strong h-bridge with own algorithms, what in your case the tmc does. Theoreticly, in far future, you could change the tmc to a solid state h bridge or transistors/mosfets like the 42c does. The software on this thing is interestung. And it can handle the h-bridges silent as tmc does. Its open source as well. Maybe you can grab the h bridge control or let it inspirate you. 😅😊
Yes, by design these are aluminum parts that act as a heat sink. The part at the bottom is the main heat sink from the driver to the large mass of the motor. As for the top half, it's more for mechanical protection. You can learn more on the website cln17.creapunk.com/sources/enclosure/
Good question This is a big enough topic for discussion, but to summarize, these are the maximum current and voltage limits that the driver chip can apply to the motor. Limits are caused by the fact that it is very difficult to place all the necessary elements in such a small outline and provide sufficient cooling to the system. The maximum current of 2.5A is limited by the connector. So the nema23 will not be able to be used at full power. For larger motors I am designing a version that will fit much better and will be able to drive almost all motors)
Great stuff!!!! I am in the same boat, I was about to begin my journey. Now there is one thing that I may suggest. Integrating esp32 either removable with headerpins, or directly to it. This will be able to mesh them wirelessly (minus power). Maybe that will be a future version, CLN17-W / CLN17-W Pro.
Thanks, glad you like my project! The esp32 support is planned. There is a 1.27p header with uart support for expansion boards, including but not limited to esp32 as a wifi module and coprocessor. I don't plan to integrate esp32 into the board itself as that would increase the cost of the board. And the wifi functionality itself is quite specific and is not a generic feature like can-bus or step-dir. Btw besides wifi esp32 can support BT, BLE, ZigBee and Thread, so it will be a more comprehensive solution!
Infographics made in Inkscape Schematics and PCBs in KiCAD Board and enclosure renderings in Fusion360 All based on open source software and a bit of perfectionism magic)
If the community is interested, this could become a commercial open source product. I am trying to do my best to make the boards cheaper. Single board will cost around $60. However, if you order a set of several, like 3 pieces, the price will be much less. It also depends on how many people need it. The more there are, the lower the price😁 It's even possible that the PCBA will cost less than $25, and the simplified version (step-dir only) less than $20. But this is just for reference 🙃
It's just a typo... To be fair, I'm not a native English speaker and this is my first video ever, so I hope this little thing doesn't spoil the overall impression 🥲🥲🥲
@@creapunk and typos are fine, no big deal. I would just strive away from using the word “perfect” to describe anything other than in-the-moment experiences.
There are a number of reasons) 1) It is one of the cheapest microcontrollers. 2) It allows closed-loop algorithms to work. 3) It leaves time and compute powers for other computations that may be useful in some cases. 4) It gives the possibility of stand alone operation. 5) It gives a large set of interfaces, which expands the number of application domains. So it's more about maximum features for minimum money rather than overkill)
Unfortunately the project is not available yet. Please wait a bit until the development is over) I will definitely release a video when the project is finalized!
You keep saying "closed-loop", but you never mention what is the sensor for the motor position. Seems like you are emphasizing a lot of minor details, but skipping over significant issues like cost, and precision.
You can read all the information on the website) Including principle of operation, encoder type, etc. For example - cln17.creapunk.com/hardware/motion-and-position-control/ The cost is not mentioned in the video, as it is only a concept, without having a fully finished solution for sale. The purpose of the video is to understand the needs of the community for the features of this project
@creapunk I think a neat aspect of this is the ability to remove the driver and driver control off of the central mainboard. CLN17 is compact enough that i can imagine each motor communicating over canbus to a hub which connects to klipper host allowing the mainboard to be smaller/cheaper. Even if implementing step recovery ends up not being possible, having the ability to detect lost steps and cancel the print would be good enough.
That's definitely what I'm aiming for. You are right, and even I am not 100% sure that it will be possible to make real-time active skip compensation so that it would not affect the print quality at all. However, in your scenario cancelling printing is too radical. So the better solution as it seems to me will be a pause command to the main controller and to all drivers in the network, make a retract and move Z up. This would give the ability to quickly regain the lost position with almost no impact on quality and geometry. By the way, in theory any driver in the network can act as a main controller, but this needs to be tested
Not *yet*) This would make the final solution almost 2 times more expensive, which goes against the basic concept of affordability. It is common for nema17 motors that run on 48v to get very hot. It is also commonly used in thermal chambers. Placing the driver on the back of the motor will lead to driver overheating and possible problems. So in this case, it would be possible to use a more advanced version of the driver with an external encoder that would support 48v. But it is still under development.
@@creapunk I completely understand. This is an awesome project and I the work you have put into it! I just already converted to 48V and it's amazing. On the note of a 48V version. You could leave room for a water block(with a central hole) to be mounted to the nema17 and mount the driver on the back to cool both with the same block. I will eventually be converting my setup to water for this very reason. Motors also don't get nearly as hot as you might be thinking at 1.8A and 48V my motors 60mm length only get to about 50C(no enclosure) which is hot but nowhere near hot enough to damage anything. Another concern I have is the lack of large output capacitors. On my current drivers there are two large 470uF caps(well 100 but between 100-550uF is what people are using these days) 470uF is the value on the TMC5160 devboard. This may present some challenges in development as big caps aren't "space efficient". Regardless this is an amazing project and I will be keeping my eye out for future development!
its not error free. In order to correct the position there needs to be an error. Not even machines with hydrostatic ways and magnetic direct drives are error free.
You're right. In real mechanics there are always inaccuracies, backlashes and so on, so theoretical error-free is not achievable. However as a concept or some abstraction used to describe in a video for a general audience this term can be correct i think. For example in 3D printing an error of 0.1° and instant (150us) detection of the error and informing the main processor about it could be called error-free. I hope this nuance will not divide your whole impression of the project by 0 😅
Tropical Labs (Mechaduino) had a concept 7 years ago on such add-on close loop controller for Nema 17 and 23 Stepper Motors. All open source. This looks very promising and I hope you succeed.
Mechaduino is a really interesting project, but at the moment it looks outdated, as well as ustepper. Both of these drivers have very large dimensions, but even so they don't have full-fledged and convenient connectors for connecting to external devices. They also don't have can-bus, which is now becoming the new trendsetter in 3D printers and CNCs. I want to erase the boundary between industrial and cheap drivers at the expense of versatility and optimization of design 🙃
Thank you for open sourcing your work. In an ideal world, people would donate to this project to help keep the support up and running.
A really cool open-source project with awesome websites and documents!
Excellent work! I started working on a similar concept several years ago, but while it taught me a lot about PCB design I didn't have the full skillset to complete it. The projects I was hoping to use it on all went unfinished, but now maybe I can get them done!
It would be interesting to see how easy it is to add the CLN17 to an existing product.
Without any change, I assume the motor will start to make sure it doesn't lose a step by itself.
Of course, the full benefit is not just to mod the stepper but having a central board to discuss with the stepper.
I think from the point of view of integration into a printer for example, it should be quite simple.
For example you can make the driver working on step-dir interface as an clone of the commands from TMC2209. Then you can use it as a normal driver.
This way it can replace the limit switch in case of ACTUALLY accurate sensorless parking. Also if any error occurs in the system, also it can create a proper interrupt and save the printing.
This is a general concept for now, but I think it will be realised after some time) A lot of work is required 😅
Please obsolete my stallGuard and make encoderless stepper motors ancient. When closed-loop stepper motors with Trinamic-grade silence and accuracy become readily available at less than $20 each, it will be an “innovation” and beat industrial AC servos. I’ve subscribed hoping this is the very beginning of such wonderful new era.
I really hope to do so)
Accessibility requires mass production. Mass production requires a large community. A large community requires versatility.
So here we are)
You found typo in “preciese” in the video and opened comments to see if someone did too 😂. Great product BTW
🥹😭😭😭😭😭
I see trinamic and i upvote.
Awesome, thanks for making this available
This is so GOOD! Do more uploads pls 🤩🤩
Excellent work. If you put these up for sale I'll buy a few.
That's awesome, can't wait to see where this goes!.
Looking forward to the larger iterations (nema 34, etc). Make sure that what you come up with WILL work with Mach3 & 4. Opens up the whole market of people like myself with CNC mills, lathes, etc.
Currently in development is an advanced version for Nema 23 34 42 including hybrid 3 phase stepper motors.
Im not really familiar with Mach3 but could potentially do support in the future)
Join the community to give valuable advice when it comes to this)
Fantastic work! Love the professional design, and the video is just great! Can't wait to test it!
As I love this subject, the bad thing is that I don't know English, even if it is to consume a written text, with wonderful documentation like this CLN17 driver, it would be great if we had a class on this, it could even be paid content, which so much beautiful information.... It would also be for larger engines nema 34 for example... Maybe one day I will learn
This is exactly what I’ve been looking for! I have so many projects that could benefit from this
Awesome work, you should open a shop online !
What a stunning project and presentation! I’ll follow the project
Awesome! Amazing work!
its amazing to see what an individual can create
this is amazing contribution grate work!!
Great project! I have no use for closed-loop, but it's very useful for some projects, and it's super if there is open solution.
With a lot of struggle I'm manufacturing my Router, a little without money, but I'm almost finished, I have a Nema 34 85kgf engine, I can implement this on these engines, I'm going to use Linux CNC
Genius🤯
Great work that I can noe even fully appreciate. Got here looking for open firmware, as most of available nema17 stepper closed loop controllers have bugged and not supported codebase.
Would be glad to purchase when it comes out.
Can't wait to put this on my ender 3! Maximum overkill!
Man amazing work!!!
THANK YOU!
I thought there was a closed loop stepper driver board/platform project already, and I've been wondering why we haven't seen it widely implemented in desktop 3d printers. How does CLN17 differ from the S42B that came out a few years ago, for example?
I guess the main difference is the project's opensource and much more features. You can visit the project's website, there is a section including a summary table between all existing closed loop drivers of this type.
To the question about prevalence, I think the answer is simple. Almost all such drivers are not good enough and/or expensive enough for mass use. So one of the goals of this project is to solve these problems by making the solution universal and mass-produced.
Awesome, I hope the project gets a lot of support because it would be great to see this technology become mainstream.@@creapunk
Really cool!
Sorry for my ignorance but This has become quite the time sink and before I delve into this project any further can you tell me how it compares to the MKS SERVO42, SERVO57 and AS5600 projects for high accelerations in 3d print applications?
where can i buy a few of these bad boys!!! nice work 👏👏👏
Very cool!
Contact a company like mellow, LDO, fysetc etc and ask if they can make them, I am interested in closed loop steppers but I do not want to make them myself.
Now this is exciting.
Really nice！can't wait to make one myself！
The prototype source files are already published on github.
I don't recommend to build it manually as the driver has 0402 parts. Also the current version does not have good software yet 😞
@@creapunk 😄, 0402 is okay, I have made many boards using 0402 components. But I’m really not good at software. looking forward to your updates.🥳
Looks good! ;)
Very cool!
Kudos, Cool idea!
I approve of the project, but couldn't help chuckling at 'Preceise'
Waiting for pmsm/bldc servo with step/dir
Waiting for NEMA23 version for my CNC.
nice ,looking forward for tmc5160 version
Already in development 😎
@@creapunk what voltage you planing stock or something with mosfets?
I'm aiming for 48v 10a operating mode. There may be limitations that I don't know about yet, but from the TMC5160 driver documentation it seems possible
Love this
You should share this wirh the 3d printer community and voron community
this could be huge for speed printing
What is the maximum speed? This is a problem with many closed loop drivers.
Interesting stuff
How does it compare to the standard MKS $10 closed loop driver, or the newer $15 SERVO42C one?
You had me at canbus! Are you selling these yet?
Cool!!!
Great project ! I would love to get a course to learn that but I don't know how to start
FANTASIC WORK! If you dont mind me asking, which software did you use to 3d render the models and create that section of video from 01:35 to 01:40
Of course, I used Autodesk Fusion360 for rendering and then overlay of several images in Adobe Premiere Pro. The PCBA 3D model I took from KiCad. It's pretty simple, but not so convenient)
There is a typo on Precise. Like the project
Oh, thanks, my bad :)
Please make an affordable nema 23 and nema 34 version. We will donate
Already in progress 😏
Can you explain what advantages your nice project over the SERVO42C and SERVO42D has? I am always interested in nice project like yours. I can inly speak for the 42c... and let me say, it works great with klipper in my hypercube evolution. Magnetic orientation field sensor works good. And its unbeleavable silent. Like tmc2209 drivers. Maybe more silent. I can hear mechanicial sounds on my rails, ive never noticed before.
That's a good question.
Well, first of all, there are a lot more interfaces that are really useful. For example can-bus can be obtained without fuss with an external uart converter. Also there is USB, which is convenient for flashing or debugging during programming. And at the same time the microcontroller is much faster)
Maybe after some time I will make a version that will also support FOC and will work better than tmc2209, but it takes time
@@creapunk the can bus sounds promising. But i think that klipper can not coexist in the cpu? Usually we have to compile a klipper firmware for mcu like ebb36 and so on. I think your firmware uses special algos with magnetic angle sensor check and so on. To reach people interested in nema17 servos i think it would be nice when it is compatible with klipper firmware over can bus for 3d printer.
@@OrgathmTech Yes, I too hope to port klipper to this driver in the future, but there's a lot more to do at a low code level)
@@creapunk i wish you great success.
Okay. I checked all that well documented informarions on your homepage. Nice job, appreciate it. Its interesting how deep and powerful the functions are.i like that. I found one difference. You are using a tmc2209 and the servo42c's cpu takes controll over a strong h-bridge with own algorithms, what in your case the tmc does. Theoreticly, in far future, you could change the tmc to a solid state h bridge or transistors/mosfets like the 42c does. The software on this thing is interestung. And it can handle the h-bridges silent as tmc does. Its open source as well. Maybe you can grab the h bridge control or let it inspirate you. 😅😊
What are those covers you have around the pcb, the to one looks like a heat sync, right? But what is the one underneath?
Yes, by design these are aluminum parts that act as a heat sink. The part at the bottom is the main heat sink from the driver to the large mass of the motor. As for the top half, it's more for mechanical protection. You can learn more on the website cln17.creapunk.com/sources/enclosure/
fantastic job 👏🏼
any chance in the future to add wifi ? (esp32 version ? )
This is possible using an expansion board for which there is a 1.27mm connector 🙃
Can this work with klipper?
Noob question: what are the challenges in making this work with bigger motors?
Good question
This is a big enough topic for discussion, but to summarize, these are the maximum current and voltage limits that the driver chip can apply to the motor.
Limits are caused by the fact that it is very difficult to place all the necessary elements in such a small outline and provide sufficient cooling to the system.
The maximum current of 2.5A is limited by the connector. So the nema23 will not be able to be used at full power.
For larger motors I am designing a version that will fit much better and will be able to drive almost all motors)
48V-Version would be nice to
It's in development!
Awesome but lets see some examples of how this is better, use comparisons. Larry
Will it work with klipper os for 3d printers?
There are no obstacles to do this, but I need to make the firmware for that)
What's wrong with makerbase, bigtreetech, closed loop boards...?
Interfaces, features and computing speed for custom tasks
cln17.creapunk.com/hardware/alternatives/
@@creapunknice man! I read through the website awesome work. Very interested, have these been put on the market?
Great stuff!!!! I am in the same boat, I was about to begin my journey. Now there is one thing that I may suggest. Integrating esp32 either removable with headerpins, or directly to it. This will be able to mesh them wirelessly (minus power). Maybe that will be a future version, CLN17-W / CLN17-W Pro.
Thanks, glad you like my project!
The esp32 support is planned. There is a 1.27p header with uart support for expansion boards, including but not limited to esp32 as a wifi module and coprocessor. I don't plan to integrate esp32 into the board itself as that would increase the cost of the board. And the wifi functionality itself is quite specific and is not a generic feature like can-bus or step-dir. Btw besides wifi esp32 can support BT, BLE, ZigBee and Thread, so it will be a more comprehensive solution!
What software did you use for board overview and design pictures?
Infographics made in Inkscape
Schematics and PCBs in KiCAD
Board and enclosure renderings in Fusion360
All based on open source software and a bit of perfectionism magic)
ok, most important of it all: How much will it cost?*if you are going to sell boards*
If the community is interested, this could become a commercial open source product. I am trying to do my best to make the boards cheaper. Single board will cost around $60. However, if you order a set of several, like 3 pieces, the price will be much less.
It also depends on how many people need it. The more there are, the lower the price😁
It's even possible that the PCBA will cost less than $25, and the simplified version (step-dir only) less than $20.
But this is just for reference 🙃
Seems promising, gotta check that hardware to see what else we can get other than position control
Not sure if some meta-joke that went over my head, but the misspelling of “precise” bugs me when the creator is lauding their “perfectionism “.
It's just a typo...
To be fair, I'm not a native English speaker and this is my first video ever, so I hope this little thing doesn't spoil the overall impression 🥲🥲🥲
@@creapunk and typos are fine, no big deal. I would just strive away from using the word “perfect” to describe anything other than in-the-moment experiences.
How is this controller closed loop? Is there an encoder?
Yep, on the back side of PCBA)
I see the github pcb design files but doesn’t have the stm32 firmware source code. Is that available at a different location?
The code files are still very raw. I will publish them in the same repository when they are ready
That seems overkill. Why would each stepper require so much processing power?
There are a number of reasons)
1) It is one of the cheapest microcontrollers.
2) It allows closed-loop algorithms to work.
3) It leaves time and compute powers for other computations that may be useful in some cases.
4) It gives the possibility of stand alone operation.
5) It gives a large set of interfaces, which expands the number of application domains.
So it's more about maximum features for minimum money rather than overkill)
What's the best way to order it?
Unfortunately the project is not available yet.
Please wait a bit until the development is over)
I will definitely release a video when the project is finalized!
@@creapunk thanks.. before I use custom firmware for s42b but btt stop producing them... And there is no firmware for other pcb's
You keep saying "closed-loop", but you never mention what is the sensor for the motor position. Seems like you are emphasizing a lot of minor details, but skipping over significant issues like cost, and precision.
You can read all the information on the website)
Including principle of operation, encoder type, etc. For example - cln17.creapunk.com/hardware/motion-and-position-control/
The cost is not mentioned in the video, as it is only a concept, without having a fully finished solution for sale. The purpose of the video is to understand the needs of the community for the features of this project
0:52 I have no idea what this thing is called, but I want one... Where can I buy it?
If you're talking about a hot plate, it's a Miniware MHP30
@@creapunkI thought it had a special name for soldering. Thank you!
What software is this 0:40?
KiCad with dark schematic theme 🙃
the problem is that it's not economic
Ok so who's gonna put Klipper on it
Maybe myself or the community will make Klipper support in the future. The driver seems to have enough processing power and speed for that
@creapunk I think a neat aspect of this is the ability to remove the driver and driver control off of the central mainboard. CLN17 is compact enough that i can imagine each motor communicating over canbus to a hub which connects to klipper host allowing the mainboard to be smaller/cheaper. Even if implementing step recovery ends up not being possible, having the ability to detect lost steps and cancel the print would be good enough.
That's definitely what I'm aiming for.
You are right, and even I am not 100% sure that it will be possible to make real-time active skip compensation so that it would not affect the print quality at all.
However, in your scenario cancelling printing is too radical. So the better solution as it seems to me will be a pause command to the main controller and to all drivers in the network, make a retract and move Z up. This would give the ability to quickly regain the lost position with almost no impact on quality and geometry.
By the way, in theory any driver in the network can act as a main controller, but this needs to be tested
No support for 48V 😢
Not *yet*)
This would make the final solution almost 2 times more expensive, which goes against the basic concept of affordability.
It is common for nema17 motors that run on 48v to get very hot. It is also commonly used in thermal chambers. Placing the driver on the back of the motor will lead to driver overheating and possible problems.
So in this case, it would be possible to use a more advanced version of the driver with an external encoder that would support 48v. But it is still under development.
@@creapunk I completely understand. This is an awesome project and I the work you have put into it! I just already converted to 48V and it's amazing. On the note of a 48V version. You could leave room for a water block(with a central hole) to be mounted to the nema17 and mount the driver on the back to cool both with the same block. I will eventually be converting my setup to water for this very reason. Motors also don't get nearly as hot as you might be thinking at 1.8A and 48V my motors 60mm length only get to about 50C(no enclosure) which is hot but nowhere near hot enough to damage anything. Another concern I have is the lack of large output capacitors. On my current drivers there are two large 470uF caps(well 100 but between 100-550uF is what people are using these days) 470uF is the value on the TMC5160 devboard. This may present some challenges in development as big caps aren't "space efficient". Regardless this is an amazing project and I will be keeping my eye out for future development!
its not error free. In order to correct the position there needs to be an error. Not even machines with hydrostatic ways and magnetic direct drives are error free.
You're right. In real mechanics there are always inaccuracies, backlashes and so on, so theoretical error-free is not achievable.
However as a concept or some abstraction used to describe in a video for a general audience this term can be correct i think. For example in 3D printing an error of 0.1° and instant (150us) detection of the error and informing the main processor about it could be called error-free.
I hope this nuance will not divide your whole impression of the project by 0 😅
@@creapunk no dont get me wrong, its a nice project, i just wanted to clarify this :)
I just respect your approach and strive to avoid inaccurate interpretation too 🙃
Here is the real question. But can it run crysis/klipper? 🤣
One day it will be! But first I need to finish the beta version)
is it ironic that you spelt precise wrong?
Nah, it's just a typo)
Tropical Labs (Mechaduino) had a concept 7 years ago on such add-on close loop controller for Nema 17 and 23 Stepper Motors. All open source.
This looks very promising and I hope you succeed.
Mechaduino is a really interesting project, but at the moment it looks outdated, as well as ustepper. Both of these drivers have very large dimensions, but even so they don't have full-fledged and convenient connectors for connecting to external devices. They also don't have can-bus, which is now becoming the new trendsetter in 3D printers and CNCs.
I want to erase the boundary between industrial and cheap drivers at the expense of versatility and optimization of design 🙃
..how bout you sort my You Tube out, YT...
Where were you when I created StepperServoCAN project ;)
Awesome project btw