New CoreXYU build. Need advice on bed leveling..

I’m currently designing a CoreXYU printer (dual X gantry).

(Picture: early WIP)

Main goals/features:
-Print high-end materials like PEEK/Ultem with support material
-Core XYU, Dual X gantry
-DyzeXTruder Pro + DyzeEnd Pro, 400 °C hotend (water cooled)
-160 °C actively heated and insulated build chamber, blow hot air at the top on currently printed layer
-Duet Wifi
-Print size ~ 300 x 300 x 300 mm
-Super rigid frame



Feel free to leave comments on the overall project. But now to my main question:

BED LEVELING:
This is a big topic I have yet to decide. I really don’t want a PEEK print or first layer to fail, especially after heating the chamber all the way up etc.
I see these 3 options:

1. (Heat resistant) Proximity sensor + True bed leveling (with 3 steppers+leadscrews)
It is supported by DuetWifi, but does this actually bring that much of an advantage? Is stepper synchronization a big factor? Cost is not that big of a factor in my case.

2. (Heat resistant) Proximity sensor + Z compensation (1 stepper, 3 leadscrews)
Let’s say I print a cube: When compensating over the first x layers, the bottom and top layers aren’t parallel right? When compensating during the whole print, do I end up with a parallelogram or are the vertical walls printed perpendicular to the bed? (I know, exaggerated of course)

3. No bed leveling, table screws (1 stepper, 3 leadscrews)

4. ?

(I assume the bed surface in itself is flat)
I would love to hear your thoughts on this

with a 160c chamber you got about 140c temp difference inside and outside but didn't see how you deal with the thermal expansion?

Are drive belts OK to work in those conditions as well I wonder ?

Edited 2 time(s). Last edit at 03/13/2019 08:46AM by Pippy.

Quote
Ernest1122
with a 160c chamber you got about 140c temp difference inside and outside but didn't see how you deal with the thermal expansion?

Quote
Pippy
Are drive belts OK to work in those conditions as well I wonder ?

The frame and all electrical and mechanical parts are outside of the chamber. The chamber itself is on one side only loosely connected to the frame.
However the print bed is inside and thus exposed to thermal expansion... I guess I have to make some changes to the Z lift assembly

The bed plate thermal expansion it relatively easy to deal with if you put it on a kinematic mount. The thing you have to look out for is the expansion of the printer's frame which will move the linear guides apart in the Y axis and the guide rails apart in the Z axis. It may be less of a problem in Z because the rails will tend to flex.

You'll have to water cool the hot-end or bring in outside air with a hose.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The bed plate thermal expansion it relatively easy to deal with if you put it on a kinematic mount. The thing you have to look out for is the expansion of the printer's frame which will move the linear guides apart in the Y axis and the guide rails apart in the Z axis. It may be less of a problem in Z because the rails will tend to flex.

You'll have to water cool the hot-end or bring in outside air with a hose.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

According to Gates, their GT2 drive belts are rated for a working temperature range of -54°C to +85°C.

Gates Polychain GT2

I guess you'll have to swap to leadscrews if the belts fail when you use them in your 160°C chamber ?

Quote
floesi
1. (Heat resistant) Proximity sensor + True bed leveling (with 3 steppers+leadscrews)
It is supported by DuetWifi, but does this actually bring that much of an advantage? Is stepper synchronization a big factor? Cost is not that big of a factor in my case.

2. (Heat resistant) Proximity sensor + Z compensation (1 stepper, 3 leadscrews)
Let’s say I print a cube: When compensating over the first x layers, the bottom and top layers aren’t parallel right? When compensating during the whole print, do I end up with a parallelogram or are the vertical walls printed perpendicular to the bed? (I know, exaggerated of course)

3. No bed leveling, table screws (1 stepper, 3 leadscrews)

Go for either #1 or #3. #1 is electrically more complex (i.e. 2 extra stepper motors + wires). #3 is mechanically more complex (extra belt, belt tension adjustment, thrust bearings, pulleys, levelling screws). Your choice. As you are building an IDEX machine, you already need at least a DueX2 expansion board. The DueX5 (which you will need instead of the DueX2 for the extra Z motors) doesn't cost a lot more, so the additional cost of #3 isn't so much.

Edited 1 time(s). Last edit at 03/13/2019 01:55PM 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].

Quote
the_digital_dentist
The bed plate thermal expansion it relatively easy to deal with if you put it on a kinematic mount. The thing you have to look out for is the expansion of the printer's frame which will move the linear guides apart in the Y axis and the guide rails apart in the Z axis. It may be less of a problem in Z because the rails will tend to flex.

Came up with this idea for the Y-gantry. I put in a bushing so the left end can now move in the X-direction. I realized that the left linear guide must not constrain in the X-direction (thats the right linear guide's job) but only has to "hold the weight" of the left end of the Y-gantry.

Quote
the_digital_dentist
You'll have to water cool the hot-end or bring in outside air with a hose.

yes

Quote
dc42

Quote
floesi
1. (Heat resistant) Proximity sensor + True bed leveling (with 3 steppers+leadscrews)
It is supported by DuetWifi, but does this actually bring that much of an advantage? Is stepper synchronization a big factor? Cost is not that big of a factor in my case.

2. (Heat resistant) Proximity sensor + Z compensation (1 stepper, 3 leadscrews)
Let’s say I print a cube: When compensating over the first x layers, the bottom and top layers aren’t parallel right? When compensating during the whole print, do I end up with a parallelogram or are the vertical walls printed perpendicular to the bed? (I know, exaggerated of course)

3. No bed leveling, table screws (1 stepper, 3 leadscrews)

Go for either #1 or #3. #1 is electrically more complex (i.e. 2 extra stepper motors + wires). #3 is mechanically more complex (extra belt, belt tension adjustment, thrust bearings, pulleys, levelling screws). Your choice. As you are building an IDEX machine, you already need at least a DueX2 expansion board. The DueX5 (which you will need instead of the DueX2 for the extra Z motors) doesn't cost a lot more, so the additional cost of #3 isn't so much.

After some pondering and tweaking I decided on #1 true bed leveling.
This brought me to the next question: which Z probe.
First I wanted to use an inductive proximity sensor, but they're huge and heat resistance makes them super expensive, plus they aren't the most accurate option. After a lot of additional reading on the topic I came across piezo electric sensors. Problem is heat again, so I have to keep them outside of the build chamber.
Where would I best place them? Does bed leveling work when theyre so far away from where the nozzle touches the glass (picture below)?


Or are there better options for high temperatures?

If you build the machine with a stable platform to support the bed plate and a stable leveling system for the plate itself, you won't need autoleveling and you won't have to manually relevel. 3 lead screws to lift the bed should be stable and will maintain constant position if the screws are all belted to one motor. 3 point kinematic mount on the bed plate will be stable, too. You'll set it level once, manually, and it will stay that way until you take the machine apart.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
floesi

Quote
dc42

Quote
floesi
1. (Heat resistant) Proximity sensor + True bed leveling (with 3 steppers+leadscrews)
It is supported by DuetWifi, but does this actually bring that much of an advantage? Is stepper synchronization a big factor? Cost is not that big of a factor in my case.

2. (Heat resistant) Proximity sensor + Z compensation (1 stepper, 3 leadscrews)
Let’s say I print a cube: When compensating over the first x layers, the bottom and top layers aren’t parallel right? When compensating during the whole print, do I end up with a parallelogram or are the vertical walls printed perpendicular to the bed? (I know, exaggerated of course)

3. No bed leveling, table screws (1 stepper, 3 leadscrews)

Go for either #1 or #3. #1 is electrically more complex (i.e. 2 extra stepper motors + wires). #3 is mechanically more complex (extra belt, belt tension adjustment, thrust bearings, pulleys, levelling screws). Your choice. As you are building an IDEX machine, you already need at least a DueX2 expansion board. The DueX5 (which you will need instead of the DueX2 for the extra Z motors) doesn't cost a lot more, so the additional cost of #3 isn't so much.

After some pondering and tweaking I decided on #1 true bed leveling.
This brought me to the next question: which Z probe.
First I wanted to use an inductive proximity sensor, but they're huge and heat resistance makes them super expensive, plus they aren't the most accurate option. After a lot of additional reading on the topic I came across piezo electric sensors. Problem is heat again, so I have to keep them outside of the build chamber.
Where would I best place them? Does bed leveling work when theyre so far away from where the nozzle touches the glass (picture below)?
[attachment 110633 8.PNG]

Or are there better options for high temperatures?

I hope there is enough play in the nuts and linear guides or at least frame flexibility to allow for this contraption to "level" without binding.

I think the same as DD, that I summarize as: ABL for a Cartesian 3D printer is a sign of a poor design/make.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
the_digital_dentist
If you build the machine with a stable platform to support the bed plate and a stable leveling system for the plate itself, you won't need autoleveling and you won't have to manually relevel. 3 lead screws to lift the bed should be stable and will maintain constant position if the screws are all belted to one motor. 3 point kinematic mount on the bed plate will be stable, too. You'll set it level once, manually, and it will stay that way until you take the machine apart.

Quote
MKSA
I think the same as DD, that I summarize as: ABL for a Cartesian 3D printer is a sign of a poor design/make.

Yes, I think you're right, the bed will stay level.
I still was a little afraid because I want to print with cold and hot chamber. But thermal expansion could only have an effect on the general Z distance between nozzle and bed, not on the levelness.
This however could be handled by using the nozzle tip as as homing reference at the beginning of each print, right?
This might even be easy to implement, as my extruders have ~70mm "parking space" left and right of the print bed.....
I will search for an accurate microswitch, heat resistant

Could these do the trick? https://docs-emea.rs-online.com/webdocs/0514/0900766b80514b6b.pdf

Edited 1 time(s). Last edit at 03/21/2019 07:22AM by floesi.

I wouldn't count on that type of switch to be repeatable.

You can use an opto interruptor at one of the arms that connects to the bed support outside the oven to set Z=0. If the bed plate is 10 mm thick and you raise its temperature 150C the vertical expansion is only about 37 um. It will be consistent, so set the sensor to trigger at Z=0 at room temperature, then in the filament profiles that set the temperature of the bed and hot-end, set a Z=0 offset to compensate for that filament's bed temperature. If you're printing ABS with a bed temperature of 100C, set one Z=0 offset for 100C. If you're printing PEEK with the bed set to 180C, use another Z=0 offset for that.

You may need to look at different adhesives to attach the heater to the bed plate. The 468MP that is commonly used on the Keenovo heaters may not like the maximum bed temperature you intend to use. It might be better to use a mechanical clamp to hold the heater on the bed plate so you don't have to worry about adhesive burning up and letting go.

You can eliminate bed leveling and heating if you print on a sacrificial foam bed...

Edited 1 time(s). Last edit at 03/21/2019 07:43AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
I wouldn't count on that type of switch to be repeatable.

You can use an opto interruptor at one of the arms that connects to the bed support outside the oven to set Z=0. If the bed plate is 10 mm thick and you raise its temperature 150C the vertical expansion is only about 37 um. It will be consistent, so set the sensor to trigger at Z=0 at room temperature, then in the filament profiles that set the temperature of the bed and hot-end, set a Z=0 offset to compensate for that filament's bed temperature. If you're printing ABS with a bed temperature of 100C, set one Z=0 offset for 100C. If you're printing PEEK with the bed set to 180C, use another Z=0 offset for that.

You may need to look at different adhesives to attach the heater to the bed plate. The 468MP that is commonly used on the Keenovo heaters may not like the maximum bed temperature you intend to use. It might be better to use a mechanical clamp to hold the heater on the bed plate so you don't have to worry about adhesive burning up and letting go.

You can eliminate bed leveling and heating if you print on a sacrificial foam bed...

I will give those opto interruptor / light barriers a try.

Yes, setting up different profiles with offsets sounds reasonable. (Although probably not just by calculation the expansion of the bed. The heat might to some extend creep onto the rest of the printer and expand some other parts as well.)

I plan on using one of these 200°C E3D heated beds https://e3d-online.com/high-temperature-heated-beds

I read your google forum on foam beds, will definitely test it as well