Closed loop Nema17

[PD0RUZ]

Anybody seen this!!  Closed loop system on nema 17 steppers..
https://www.kickstarter.com/projects/tropicallabs/mechaduino-powerful-open-source-industrial-servo-m
Looks great, nice solution on 3d printers missing steps while printing, or on a laser carver/engraver.
It's called Mechadino closed loop system

[Zax]

Very neat. I'm not sure you need encoders on lasers or 3D printers but for CNC mills and other applications servos are fantastic.

[PD0RUZ]

His designs for the board(OSH park)  and schematic incl firmware are available on his site..
They run very well...
By the way, i printed once a big blok, where my vertex lost some steps, (head touched the material) whole project in the trashcan after 16 hours printing.
Layer shifted a little 
This would be the solution!!
Ps @ Zax , waiting for reply of the laser supplyer in the UK. if these 40watts have a Z-axis... Let you know

[ggallant571]

These are steppers with encoders, not traditional servos.

[Zax]

That is true, but you get some advantages of each but not the speed and no-steps of true servos.

[Lob0426]

His designs for the board(OSH park)  and schematic incl firmware are available on his site..
They run very well...
By the way, i printed once a big blok, where my vertex lost some steps, (head touched the material) whole project in the trashcan after 16 hours printing.
Layer shifted a little 
This would be the solution!!
Ps @ Zax , waiting for reply of the laser supplyer in the UK. if these 40watts have a Z-axis... Let you know

Yes getting "real" position feedback would be nice, BUT your controller has to know how to use that feedback. Most 3D printer and Laser controller firmware's do not use, and do not know how to use feedback. Hotend crashes are always going to be a problem. Even if you return to proper position each time you hit the obstruction it is going eventually ruin the "print". If you get it back on the line it will just pull the work from the build plate, or crack the work, break the hotend, or any of the dozens of failures that we love, that ruin our prints.

Now the fact that they appear to return to position when acted upon is nice. But in the statement above you can see what would happen if it keeps returning to the proper position when there is an obstruction in it's path. If you were lucky and the obstruction was not to tall you might clear it enough on the next pass to save the print, or maybe not!

Now its ability to sense "tension" may be useable, again, the Printer firmware would have to be capable, It could pause the print and alert you there is a problem. Then you could "cure" the issue and restart the print.

Another problem is that you need an extra board to interpret for the steppers and still need a controller. So you would be best to build a new controller that is made for these new "drivers".

Also you could end up taking a step back if you use the wrong "mode". Most firmware's already have "motion control". If you enable it in the stepper driver then the controller will have not control over it. On our Lasers that would mean long work times. It would be like turning "Laser Mode" of in grbl.

These boards are great, But they would need to be part of a system to really make them work right for the 3D printers or the Laser kits. They are Intelligent drivers being connected to an ignorant controller!

[PD0RUZ]

Thanks Lob0426, explained  very well..
Maybe this would be nice in the near future for the developers of this board..
Would like to try some, but with feedback to NC like you explained above.

What would be nice too, is a display, connected with gbrl, like marlin, to see axis positioning, refrencing, and maybe run G-code from SD...
Does somebody have that allready, ??

[Administrator]

There are two major classes of laser machines:

1. The low powered laser machines that this forum is all about.

2. The more powerful CO2 (gas) lasers. I'm ignoring the really expensive and really powerful Fiber, Crystal and Plasma industrial lasers.

The more powerful CO2 lasers are expensive, ranging from, perhaps a cheap kit for roughly $750.00 to well over $10,000.00+. The more expensive machines have all of these bells and whistles and they also have very to extremely expensive CAM software.

If a company were to add this kind of stepper motor to a 'low cost', low powered laser machine, I don't think it would be 'low cost' anymore. As @Lob0426 indicated, some software programming and testing would be needed.

I'm not sure from reading this post that you realize that...maybe you do. But my point is, and I could be off base here, is that you're just thinking out loud of how nice it would be if our machines has this kind of stepper -- but -- I think to engineer that into our kind of DIY kits would drive up the cost, perhaps, too much.

Just my 2 cents.

[Zax]

I see this as a cheap replacement for the servos with step/direction drivers, it's much easier to interface this solution as you can still use the existing software and firmware but obviously it has limitations since the control is done at the motor with no feedback.

It is a closed loop system so an improvement over open loop steppers, but perhaps not suitable for all applications (especially cheap ones as Ralph mentioned).

[Lob0426]

Good points;
At $130 x3 the boards cost more than one of the Laser kits I purchased. And $175 for three "servo's" is more than the cost of most peoples cost for a kit. Even if they were "out" for volume purchases they would raise the cost significantly of the Laser kits and the 3D printers/kits.

I did not even cover the software changes that would have to be made.

PD0RUZ:
I do not have "positioning" as the software/firmware does not provide real time feedback, but using a tablet as a dedicated device on a 3D printer you can watch the gcode visually as it runs through software. That is a function of virtual position by the software, and lags well behind the "real" position most of the time. I have not looked in BCL or T2Laser to see if they have a similar function.

[Administrator]

>I have not looked in BCL or T2Laser to see if they have a similar function.

BCL does have that, but I recently deactivated XYZ real time feed back positioning display for the following reasons:

1. As you said, there is noticeable lag in fetching the XYZ positioning data.
2. To obtain accurate, real time feedback requires:
--priority CPU time processing the returned data (this priority has to be higher than gcode processing priority which is undesirable).
--low powered PCs such as tablets and notebooks (I have several different kinds to test with) simply do not have the horse power to do this. This is the main problem.
3. Additional Grbl overhead complexity to parse the returned data string format based on which version of Grbl you are using. As more and more firmware is added, more coding has to be added to parse the firmware data string correctly.

All in all, I feel on these low powered laser machines with unknown PC CPU power, I have come to the conclusion that displaying the real time position of the XYZ axis, while nice to have, is not 'mandatory' to have.

The very highest CPU priority needs to be on:
--processing the gcode.
--sending the gcode to the controller board.
--checking the gcode return status from the controller board.

My testing has shown that on low powered PCs, to handle the gcode sending/status receiving AND to process the real time XYZ axis data at the same time, simply bogs down the PC CPU. Thus, I have made the decision to deactivate this feature in BCL.

[PD0RUZ]

ok point made.
No feedback system possible for our setup, and no realtime axis position read.
But my second question was, if it would be possible to create a hardware setup like the ramps 1.4 with gbrl, like the marlin with menu and sd card acces, so that it can run standalone. Reading the generated g-code from sd and not direct from a pc.
Or does this need more power from de atmega too?
And whats the difference between the Marlin firmware and the GBRL..
I know that the Marlin can dnc from sd card.

Sorrie for  don't understanding it right, but you guys work a long time with gbrl and know all ins and outs of it.
Just looking for some more options, what can and can't be done

[Administrator]

I have a couple of users running BCL on a RAMPS 1.4/Marlin firmware setup. All I needed to do was to add some gcode words in BCL Setup to allow changing power from S gcode word to P and change the M3 and M5 gcode words to something else. They got the updated BCL and made the changes in Setup and was then able to use BCL to send gcode to their laser machines!

I don't know anything about the SD card issue with Marlin, but if the firmware supports it, then yes, you should be able to do that and easily.

BTW my 3D Printer uses a RAMPS 1.4 board and I love it...and I'd much prefer using a RAMPS 1.4/Mega2560 on my laser machine...maybe someday I will just add the board, connect the wires, flash the firmware with Grbl 1.1e and then just use it!!!

[PD0RUZ]

I have a kind of Ramps in my Velleman Vertex 3d printer, based on 2560, but they designed the board different. And like the Marlin firmware with the menu and the different kind of languages. Have a spare laying around with SD card reader an large display that i want to use on the A3 laser.
But looking for suitable Marlin software, because i have different stepper drivers Not the A4988 but the DRV8825, these have some pins swapped around i think.. I'll let you know when it runs on the marlin base firmware.
Pm me for the standalone license of your modified software, so i know what i can do... Still planning on buying the 40watt co2 like Zax has...

Thanks for the info..

[wild.bill]

I have some things to say about using the RAMPS board with a laser system but it will get lost in this thread. I will start a new topic in a few minutes with my comments.