Extending the Y-Axis  

Re: Extending the Y-Axis

The heatbed also has 9 discs embeded in it for the calibration of levelling and the Z axis. For the design files see GitHub

Regards,
Martin

Re: Extending the Y-Axis

Hi Joe,

1, your heat bed travel needs to be ballanced around the frame so that you can move the heat bed far enough back for the extruder to reach the front extremity and far enough forward so that your extruder can reach the back of the bed.
as the extruder is in front of the main frame, this results in the front being longer than the back, as per the standard prusa

2, your bed will need longer rails, as you have discovered 8mm is likely to be too flimsy, if you use rods like the mk3, you have to support them from the ends, so the point of maximum deflection, is pretty much directly under the extruder... you could consider:
https://www.ebay.co.uk/itm/SBR12-600mm-12MM-Fully-Supported-Linear-Bearing-Rails-Shaft-Rod-2-SBR12UU-Block/152629693935?_trkparms=aid%3D555018%26algo%3DPL.SIM%26ao%3D1%26asc%3D49131%26meid%3D43b62d032eea4cd6b3fd7b56b8a8237a%26pid%3D100010%26rk%3D6%26rkt%3D7%26mehot%3Dpp%26sd%3D352070502580%26itm%3D152629693935&_trksid=p2047675.c100010.m2109
or
https://www.ebay.co.uk/itm/MGN12-12mm-Miniature-Linear-Rail-Slide-MGN12H-Carriage-for-Kossel-3D-Printer/292569736117?hash=item441e83e3b5:g:3hoAAOSwdDBZh9-1
in which case you could support them along their length and have a more robust solution with potentially lower overall weight

3, magnets are affected my temperature, you might find that magnetic sheet has unsatisfactory, magnetic and adhesive properties. some info on the link below
https://www.first4magnets.com/high-temperature-resistance-magnets-t162

thin magnets are more susceptible than thicker ones.... magnetic sheet is very thin... however Angus at Makers Muse, has successfully used this method.
you could consider using "Swiss Clips"

4, Mosfets are generally low voltage devices, a large heatbed may require a lot of current at low voltage, you may have to consider using solid state relays and mains heaters.

5, Heaters, you may have to consider splitting your heaters so that you can get mountings between the bed and the Y platform.

6, reciprocating mass, the bigger you make the bed, the more difficult it will be to accelerate and decelerate, so you will probably need heavier belts and perhaps bigger motors.

7, if you change the drive train, you may need to change the stepper driver chips, and revise the acceleration, jerk, and probably other settings too

8, Pinda, senses pads on the original heatbed, to measure assembly accuracy. it is unlikely that you will be able to emulate this, with an aluminium heatbed.
the pinda also does 9 point mesh bed levelling, the test point locations are coded into firmware, you would need to adjust these. I am assuming you propose to use spring steel sheet and pei, so the pinda should be able to sense that for mesh bed levelling, ok

9, cable management, the longer bed will require a longer work surface, if you keep the cable management to the rear of the frame, you will need to allow additional space for that too... maybe consider using a cable chain in front of the main frame. like this...

I have seen this done on a prusa mk3, but can't easily locate the picture!

10, Bed support, the current bed is supported on three bearings to reduce the chances of carriage jamming due to inaccuracies in assembly, there is nothing to stop you trying 4 point support, but you are advised to include holes for 3 bearing support, during machining, in case you need to go back.
it is likely that the two bearings are too close together on the original prusa, to support the longer bed, but increasing this dimension will lead to a longer Y rail requirement.

11, extruder. you plan to increase the potential model size significantly, you may need a bigger extruder to keep build times, reasonable...

apart from that, and anything I may have missed, the upgrade will be dead easy,

Best wishes, Joan

Re: Extending the Y-Axis

Oh, you may have to revert to traditional Y axis end stops, as the current detection may be affected by the inertia of the new bed.

regards Joan

Re: Extending the Y-Axis

Thanks Joan! Great reply. I have some follow-up questions below if you have time.

Hi Joe,

1, your heat bed travel needs to be ballanced around the frame so that you can move the heat bed far enough back for the extruder to reach the front extremity and far enough forward so that your extruder can reach the back of the bed.
as the extruder is in front of the main frame, this results in the front being longer than the back, as per the standard prusa

Right. Since I am adding a total of 90mm in travel I was planning to add 45mm of travel to the front and 45 to the back, so the front would maintain the same length differential as it has now.

2, your bed will need longer rails, as you have discovered 8mm is likely to be too flimsy, if you use rods like the mk3, you have to support them from the ends, so the point of maximum deflection, is pretty much directly under the extruder... you could consider:
https://www.ebay.co.uk/itm/SBR12-600mm-12MM-Fully-Supported-Linear-Bearing-Rails-Shaft-Rod-2-SBR12UU-Block/152629693935?_trkparms=aid%3D555018%26algo%3DPL.SIM%26ao%3D1%26asc%3D49131%26meid%3D43b62d032eea4cd6b3fd7b56b8a8237a%26pid%3D100010%26rk%3D6%26rkt%3D7%26mehot%3Dpp%26sd%3D352070502580%26itm%3D152629693935&_trksid=p2047675.c100010.m2109
or
https://www.ebay.co.uk/itm/MGN12-12mm-Miniature-Linear-Rail-Slide-MGN12H-Carriage-for-Kossel-3D-Printer/292569736117?hash=item441e83e3b5:g:3hoAAOSwdDBZh9-1
in which case you could support them along their length and have a more robust solution with potentially lower overall weight

Do you know if the quality of the MGN12 linear rails at this price point are acceptable? The Hiwin rails, which seem to come fairly highly recommended, are much more costly.

3, magnets are affected my temperature, you might find that magnetic sheet has unsatisfactory, magnetic and adhesive properties. some info on the link below
https://www.first4magnets.com/high-temperature-resistance-magnets-t162

thin magnets are more susceptible than thicker ones.... magnetic sheet is very thin... however Angus at Makers Muse, has successfully used this method.
you could consider using "Swiss Clips"

I'm using a Buildtak build system on another printer which uses a thin magnetic sheet. I am typically printing at 60C and it hasn't been an issue, but I will take this into consideration.

4, Mosfets are generally low voltage devices, a large heatbed may require a lot of current at low voltage, you may have to consider using solid state relays and mains heaters.

Understood. I have a number of high current SSRs on the shelf from another project. They could easily serve this function.

5, Heaters, you may have to consider splitting your heaters so that you can get mountings between the bed and the Y platform.

Yes, this will be a challenge.

6, reciprocating mass, the bigger you make the bed, the more difficult it will be to accelerate and decelerate, so you will probably need heavier belts and perhaps bigger motors.

7, if you change the drive train, you may need to change the stepper driver chips, and revise the acceleration, jerk, and probably other settings too

I'll keep an eye on the additional mass. I'd like to stick with the original motor and drivers, but I'll keep that in mind. I am concerned about getting the settings dialed in correctly. I imagine I'll be be looking for advice on good calibration approaches some day in the future.

8, Pinda, senses pads on the original heatbed, to measure assembly accuracy. it is unlikely that you will be able to emulate this, with an aluminium heatbed.
the pinda also does 9 point mesh bed levelling, the test point locations are coded into firmware, you would need to adjust these. I am assuming you propose to use spring steel sheet and pei, so the pinda should be able to sense that for mesh bed levelling, ok

OK, let me know if I have this right. The probe detects the calibration points on the heatbed because there are 4 metal pads embedded into the MK52 at specific points. That's what you're saying will be difficult to emulate on an aluminum plate. Is the right? I will be using a spring steel sheet, so normal bed leveling won't be an issue.

9, cable management, the longer bed will require a longer work surface, if you keep the cable management to the rear of the frame, you will need to allow additional space for that too... maybe consider using a cable chain in front of the main frame. like this...
cable chain.JPG

I have seen this done on a prusa mk3, but can't easily locate the picture!

I hadn't started thinking about cable management yet. I'll look into this, though it's a bit bulky.

10, Bed support, the current bed is supported on three bearings to reduce the chances of carriage jamming due to inaccuracies in assembly, there is nothing to stop you trying 4 point support, but you are advised to include holes for 3 bearing support, during machining, in case you need to go back.
it is likely that the two bearings are too close together on the original prusa, to support the longer bed, but increasing this dimension will lead to a longer Y rail requirement.

I saw a CR-10 mod where the builder replaced the existing wheel assembly with linear rails for better operation. He used two bearings on each rail successfully, so I am encouraged by that.

11, extruder. you plan to increase the potential model size significantly, you may need a bigger extruder to keep build times, reasonable...

Understood.

apart from that, and anything I may have missed, the upgrade will be dead easy,

Best wishes, Joan

I'm confident it won't be easy, but I'm also confident I will have fun trying! The biggest challenge I face is finding the time to attempt all this.

Thanks again for your very thorough response.

Re: Extending the Y-Axis

Hi Joe,

those rails were only chosen to demonstrate supported rails as opposed to the suspended system that Prusa use.

Hiwin will be better quality.

whether the ones I posted will be satisfactory to you is a question I can't answer,

four rails, as queried in your first post, would be much more difficult to set up.

the embedded pads on the prusa heatbed are only required for checking squareness of the chassis. if you build carefully you won't need them.

bed levelling would sense the buildtak carrier metal. you need to instruct the firmware to use test points, further apart...

regards Joan

Re: Extending the Y-Axis

Got it. Thank you.

Re: Extending the Y-Axis

8, Pinda, senses pads on the original heatbed, to measure assembly accuracy. it is unlikely that you will be able to emulate this, with an aluminium heatbed.
the pinda also does 9 point mesh bed levelling, the test point locations are coded into firmware, you would need to adjust these. I am assuming you propose to use spring steel sheet and pei, so the pinda should be able to sense that for mesh bed levelling, ok

A longer Y axis would help me out big time in something i would like to do. My thinking was along the lines of 2 mk52 beds together, still lots to figure out.
But could the 9 extra point be added to the mesh leveling or the points of the 9 be altered to find the new positions?
I'm guessing the firmware is set to the 210mm size. Can this be alterd easliy?

RE: Extending the Y-Axis

Joe,

Did you end up doing anything with this?  Asking because it does seem like a good mod to make.  I'm thinking of getting another mk3 kit and extending it to 400mm Y right out of the box. 

In my thinking, it would require a complete re-make of the Y-axis, but almost no change to the X and Z at all.  I'd have to modify the overhead frame to clear a thicker bed.  Without going through a huge design/testing effort, the simplest way to do a heated bed might be just a machined aluminum plate, heated at the bottom.  Recesses could be cut in for magnets to attach a steel printing sheet.  Use the same basic design for traverse as the Prusa, but with larger diameter rods, bigger belt, and bigger stepper.  Probably 2x rod size, at least one motor size up, and maybe triple the belt strength. 

Making it stiff enough to not have any additional flex in the Z direction would most likely be the toughest physical challenge, and sorting out how to tweak the firmware would probably be the biggest PITA.  

To me, this is a really great idea.  Similar to the way the 707 and other aircraft were stretched in length.  A relatively simple way to give it more capacity.  It's also a bit like how other machine tools work.  It's normal for a mill to have much more travel in one direction than the other (xy), because increasing just one dimension doesn't require increasing rigidity in the other 2 axis.

Tom