Welcome! Log In Create A New Profile

Advanced

Does this CoreXY make sense?

Posted by Schild0r 
Does this CoreXY make sense?
April 02, 2019 06:57AM
Hello everyone,

about two years ago I bought the original Flsun cube (this one)

But shortly after that I learned about the concept of CoreXY and fell in love with that. I did not want to buy a completely new printer so I wanted to covert the cube to CoreXY.
I chose the design of jand1122 (this one) and modified it a bit to fit dual extruders + part cooling, endstops and so on.
I really like the design because of
- printed parts (lightweight)
- linear rods (cheap and easy)
- horizontally arranged x axis-rods with the hotend inbetween (compact + force distribution)
- scalability and so on...

I printed all the parts and assembled them (see picture ). then I had some problems with circles not being round but flat on the spots parallel two the x axis which I (hopefully) have resolved by reprinting the parts for better tolerances in less flexible pla and using linear bearrings instead of igus bushings to reduce friction (moving the hotend manually in y was harder than moving it in x).

I always try to optimize my machines as far as possible and always try to make it "the great alrounder".
And I am currently waiting for a few new parts so I can not test the machine right now but I want your opinion/ knowledge on this:

I am currently still using the mks gen l board which honestly fits all my needs right now.
The board is 8 bit and running Marlin 1.1.9
The motors are turning some 20 teeth gt2 pulleys which makes 80steps/mm
combined with 1.8° Motors and 1/16 microstepping (currently A4988 but I want to go for TMC2208 in the future)

I am not that kind of person which wants top speeds of 200mm/s and above.
I know that a 2020 cubic frame is not the sturdiest but it will do its job.
I also like quality over speed and since I do not run the machine 24/7 most of the time it is idle anyway.
So preferred *print* speeds are 60-80mm/s (applies for circles too) which also is a good compromise for temperature/part cooling/overhang capability IMO.
If my calculations are correct this results in a maximum step rate of 16kHz

I am also using the reprap lcd smart controller but maybe I will upgrade that to the LCD 12864 Full Graphic Smart Controller.

So the overall question is:
1.Will this board be capable of these print speeds? (Will upgrading the LCD have an impact on this)
2. Are there any things that I am overseeing or things which you would recommend me to do?

Thanks in advance for constructive contributions.

EDIT: heres a schematic of the kinematics without belts

Edited 1 time(s). Last edit at 04/02/2019 12:41PM by Schild0r.
Re: Does this CoreXY make sense?
April 02, 2019 05:07PM
I learned something from my previous kit:

Always choose the best and longest bearings you can afford. This helps you keep orthogonality in xy axis easier. Cheap and small bearings with clearance cost you even more time.

If you are not sure your printer can do 90 degree prints, it's better search for other machine's help. It's not impossible, but less confident to expect components self-replicated by a crooked machine will work well. (That's why I use metal service shop for building my machine.)


SecKit, all-metal linear rail CoreXY kit
[seckit3dp.design]
Re: Does this CoreXY make sense?
April 03, 2019 11:28AM
Quote
Ernest1122
If you are not sure your printer can do 90 degree prints, it's better search for other machine's help. It's not impossible, but less confident to expect components self-replicated by a crooked machine will work well. (That's why I use metal service shop for building my machine.)

I wanted to have the corner parts of the kinematic being milled... unfortunatley 1600€ is out of my range.
I am printing them on the ultimaker at work now
Re: Does this CoreXY make sense?
April 03, 2019 01:12PM
Actually, you start by making the axes square without the belts in place. When you tension the first belt, it will pull the axes out of square, mostly by flexing the X axis, then you tension the second belt just enough to pull them back into square. The length of the bearings shouldn't really matter.

As far as controllers go, the biggest problem with the 8 bit controllers is that if you send gcode via USB, the speed of the USB transfer can cause the extruder to pause occasionally during prints, especially if you're trying to print fast. The other problem is dealing with configuration and maintaining a version of the IDE that will keep compiling your firmware's source code.


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Does this CoreXY make sense?
April 04, 2019 09:43AM
There is at least one 8bit firmware that improves performance with regards to the USB transfer speed limits however it does require the use of a Raspberry Pi.

Edited 1 time(s). Last edit at 04/04/2019 12:06PM by obelisk79.
Re: Does this CoreXY make sense?
April 04, 2019 02:49PM
Quote
the_digital_dentist
Actually, you start by making the axes square without the belts in place. When you tension the first belt, it will pull the axes out of square, mostly by flexing the X axis, then you tension the second belt just enough to pull them back into square.
Yes that was my intention

Quote
the_digital_dentist
As far as controllers go, the biggest problem with the 8 bit controllers is that if you send gcode via USB, the speed of the USB transfer can cause the extruder to pause occasionally during prints, especially if you're trying to print fast. The other problem is dealing with configuration and maintaining a version of the IDE that will keep compiling your firmware's source code.
I read that there is only a maximum of steps a board can send to the drivers and that this is less steps for 8 bit...Does that only apply to USB printing? I always print from SD
Re: Does this CoreXY make sense?
April 04, 2019 03:03PM
SD cards should be OK.


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Does this CoreXY make sense?
April 05, 2019 06:07PM
Quote
Schild0r
I read that there is only a maximum of steps a board can send to the drivers and that this is less steps for 8 bit...Does that only apply to USB printing? I always print from SD

The better idea, IMHO, is using a TFT28 or TFT32 with a RAMPS 1.4 (low cost and works great) or MKS-Gen-L board. The TFT have 32-bit Arm CPUs and they just feed the main board the steps as needed. It is the same process as the Klipper without a Pi. But, you get the added advantage of the touch screen and, also you can use a USB thumb drive instead of SD cards.

The heavy lifting, processing steps from the GCode into Step/Dir for the motors is all done by the Arm CPU.
Re: Does this CoreXY make sense?
April 06, 2019 06:04AM
Quote
StlLooking

The better idea, IMHO, is using a TFT28 or TFT32 with a RAMPS 1.4 (low cost and works great) or MKS-Gen-L board. The TFT have 32-bit Arm CPUs and they just feed the main board the steps as needed. It is the same process as the Klipper without a Pi. But, you get the added advantage of the touch screen and, also you can use a USB thumb drive instead of SD cards.

The heavy lifting, processing steps from the GCode into Step/Dir for the motors is all done by the Arm CPU.

That actually sounds pretty good. I did not know that with the TFTs the gcode would be processed on their processor neither that they have a 32 bit one. What is the deal on 32 bit mainboards then anyway if this is a working and cheaper solution? grinning smiley

Edited 1 time(s). Last edit at 04/06/2019 06:04AM by Schild0r.
Re: Does this CoreXY make sense?
April 06, 2019 11:17AM
For some cheapness is the primary or only selection criteria. For people who value printer performance and their time, the cheapest solution isn't usually the best.

The 32 bit boards have motor driver chips soldered to the large circuit board that allows them to operate at higher current and still dissipate the heat they produce. Motor current is accurately set in the config file, not by turning a tiny, easily broken pot on a module that can't dissipate much heat, and guessing at the value of the resulting current. The Duet board uses advanced TMC drivers that interpolate 16:1 ustepping up to 256:1 which makes the motors/printer almost silent.

RAMPS looks cheap, and starts that way, but you'd better keep a bunch of spare driver modules handy or be prepared to wait for delivery while your printer sits idle with a dead driver.

Before you spring for a Chinese main board, check to see if there's an active support community for it that operates in your native tongue. Some of those boards look like a great deal, but you'll may never get any help if the thing doesn't behave the way you expect. SmoothieBoard and Duet both have large, active, and expert support communities that speak English and several European languages. You won't have to rely on Google translate.


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Does this CoreXY make sense?
April 07, 2019 09:21AM
Quote
StlLooking

The better idea, IMHO, is using a TFT28 or TFT32 with a RAMPS 1.4 (low cost and works great) or MKS-Gen-L board. The TFT have 32-bit Arm CPUs and they just feed the main board the steps as needed. It is the same process as the Klipper without a Pi. But, you get the added advantage of the touch screen and, also you can use a USB thumb drive instead of SD cards.

The heavy lifting, processing steps from the GCode into Step/Dir for the motors is all done by the Arm CPU.

WRONG


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Does this CoreXY make sense?
April 07, 2019 01:27PM
Quote
MKSA
WRONG

Well would you please correct the statement then or give any hints on what you think is better?
Just commenting "wrong" is neither very constructive nor very nice...
Re: Does this CoreXY make sense?
April 07, 2019 02:24PM
I think he means that the TFT units use 32-bit processors to manage the TFT displays, but they do not directly control the motors, heaters etc. of your 3D printer. Instead they send GCode commands to your controller electronics to get the status and to home the printer, jog the motors, set heater temperatures etc.

We are in the absurd situation that the display uses a much more powerful processor than the controller electronics. The reason that displays use 32-bit processors is that they cost less than 8-bit processors, except in applications requiring only a few I/O pins, where (last time I compared prices) a few of the ATTINY processors and perhaps some PICs are still competitive.

Edited 1 time(s). Last edit at 04/07/2019 02:25PM by dc42.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Does this CoreXY make sense?
April 09, 2019 11:49AM
Another question

I plan on upgrading to these TMC2208 which are rated for 1.2A rms and 2A peak current.
And Since the steppers I have are not labeled at all I want to get new ones.
I think this question has been asked many times but I am still uncertain about it: are Stepper Motors like these 17HS4401S rated for peak or for rms current?

Or simpler: Will these motors cause the drivers to run hotter than what is intended (if I really have to push them up to that stated 1.5A rms/peak to get enough torque)?

Edited 1 time(s). Last edit at 04/09/2019 11:52AM by Schild0r.
Re: Does this CoreXY make sense?
April 09, 2019 08:19PM
Quote
Schild0r
Quote
MKSA
WRONG

Well would you please correct the statement then or give any hints on what you think is better?
Just commenting "wrong" is neither very constructive nor very nice...

When you make a statement, you have to ensure it is correct in order not to mislead people !
You should make the correction, not me. Even APOLOGIZE when you see you are WRONG !

It isn't even in the TFT doc. No need to add to MKS ridiculous claims.

Constructive ? Yes, it is. This advice: check your facts don't ask others to prove you are wrong.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Does this CoreXY make sense?
April 10, 2019 08:53AM
Quote
MKSA

When you make a statement, you have to ensure it is correct in order not to mislead people !
You should make the correction, not me. Even APOLOGIZE when you see you are WRONG !

It isn't even in the TFT doc. No need to add to MKS ridiculous claims.

Constructive ? Yes, it is. This advice: check your facts don't ask others to prove you are wrong.

Are you aware, that it wasn't me making the false claim about the TFT? For me as the one asking the question it would be better to get a correct statement (I agree with you on that). But rather than having a false statement only identified as false I would get a correct statement no matter from whom

Edited 1 time(s). Last edit at 04/10/2019 08:55AM by Schild0r.
Re: Does this CoreXY make sense?
April 11, 2019 01:35AM
Best to ignore MkSA. He is very rude and brings a dark cloud to this forum. I come here for digital-dentists wisdom but get really turned off by MKSA’s outright abrasivemess.
Re: Does this CoreXY make sense?
April 11, 2019 09:12AM
Quote
Wescherry
Best to ignore MkSA. He is very rude and brings a dark cloud to this forum. I come here for digital-dentists wisdom but get really turned off by MKSA’s outright abrasivemess.

mmm, any other forum and they'd have been booted a long time ago.

The 'guy' (I presume ?) certainly has or has had a few life problems sad smiley

Edited 1 time(s). Last edit at 04/11/2019 09:15AM by Pippy.
Re: Does this CoreXY make sense?
April 11, 2019 11:27AM
Quote
Schild0r
Quote
MKSA

When you make a statement, you have to ensure it is correct in order not to mislead people !
You should make the correction, not me. Even APOLOGIZE when you see you are WRONG !

It isn't even in the TFT doc. No need to add to MKS ridiculous claims.

Constructive ? Yes, it is. This advice: check your facts don't ask others to prove you are wrong.

Are you aware, that it wasn't me making the false claim about the TFT? For me as the one asking the question it would be better to get a correct statement (I agree with you on that). But rather than having a false statement only identified as false I would get a correct statement no matter from whom

Indeed it was StlLooking who made the wrong statement about the TFT
"The heavy lifting, processing steps from the GCode into Step/Dir for the motors is all done by the Arm CPU".

Too bad, I replied to your last comment while in fact it was directed to him. Sorry.

Edited 2 time(s). Last edit at 04/11/2019 11:43AM by MKSA.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Does this CoreXY make sense?
April 13, 2019 08:59AM
Quote
Pippy
Quote
Wescherry
Best to ignore MkSA. He is very rude and brings a dark cloud to this forum. I come here for digital-dentists wisdom but get really turned off by MKSA’s outright abrasivemess.

mmm, any other forum and they'd have been booted a long time ago.

The 'guy' (I presume ?) certainly has or has had a few life problems sad smiley

Just trying to contain bad ideas, bad designs to propagate not to mention dangerous "designs" when high voltages, high powers are involved.
So far all the remarks I made about these proved true.

I too come here for experts, ex. VDX, nophead, DC42, sdavi ( mainly firmware). but also for the fun provided by most "designs". (just saw yours BTW)..

Don't worry about my life, so far so good, just a few facepalm injuries seeing, reading some of the stuff here.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Does this CoreXY make sense?
April 14, 2019 05:13AM
Quote
MKSA
Don't worry about my life, so far so good, just a few facepalm injuries seeing, reading some of the stuff here.

grinning smileygrinning smileygrinning smileygrinning smileygrinning smileygrinning smiley
Sorry, only registered users may post in this forum.

Click here to login