[AutoBed Levelling] New prespective

Hello there,
Here is what I think, 3D printers still have years and years till they become normal home appliance. The barrier to entry is still very high; your grandma still can't do CAD but she can your TV. I think the focus should be more in getting the technology itself closer to the eventual vision rather than sell it as is right now as a consumer friendly product.

Anyway,.... Let's get to the topic. I think one step in the printing process that can be eliminated/simplified is AutoBed levelling. I came up with some idea and I want to know what you think. The concept here is all you need is the coordinates of the Normal line on the Plane of the bed. OR you need a simple signal when the nozzle hits the printing surface. So here goes:

- [ 1] Use 5v output from Arduino to connect between Z endstop and Nozzle.
All you need is to send a 5V signal to the arduino board when the nozzle hits the "Metal" printing surface. If you wire up the 5V output from the arduino to the metal surface and the nozzle and back to the z-min endstop, et voila... maybe we'd need to isolate the stepper in the extruder not to get the 5V into it, yet otherwise this seems very simple and may actually work. That's why it's number 1 on the list. I'll build a delta in a couple of month and I'll try this for sure.

- [ 2] Use an accelerometer wired to the bed and thermo isolated to directly give the direction of the normal to bed.
Again all you need to know is the normal on the bed plane level. That's actually, if I'm not mistaken, the output of the G28 code. So one of the small accelerometers used in RC planes could do the trick here. We need to thermally insulate it and it might be expensive although we could do with a 3 axis one.

- [ 3] Set the bed on 3 screws with potentiometer attached to them. When power is on, Arduino can tell how much inclined each is telling the direction of the plane.
I need someone to correct me on this, but a potentiometer wired up to the screw will always hold it's value even if the machine is turned off, right? So when you turn it back on, Arduino reads the input from the Potentiometers plugs in an equation, gets the plane. done.

- [ 4] Set the bed on 3 screws with each connected to a black and white strip exposed to optical encoder. The optical encoder can tell how much each screw is shifted by scanning through the code.
Same idea but with optical encoders instead. Not sure if this will be shut-down proof.

What do you guys think?

[1] Oozing can and will break this idea. There will be some plastic which doesn't conduct between the nozzle and plate.

[2] What would be good place for sensor? How to thermally insulate sensor and still get even heat spread on bed? If bed is 20cm wide, sensor in centre of it and one end needs to be lifted 0.2 mm With simple math I get 0.2/100 = sin(a), where a is difference of the angle of about 0.1 degrees. It is still detectable, but smaller required changes aren't (cheaply).

Honestly, I can't tell what is so hard with bed levelling. If your machine is square, go to one corner near the adjustment bolt and but it to right distance. Go to second screw and put it to same distance. And third one should set the other axis straight too. Rest is imperfections in build plate, non square machine and flex which might need to some small compensation from ideal level. People don't need more automagic leveling, they need more square machines

In normal Prusa i3 I would think one advancement. Having under the Y-axis rest place for hot-end. So that build plate would be run out of the way and Z would be lowered on it. This way Z axis wouldn't have any forces and screws wouldn't go out of sync when power is switched off. But this isn't mechanically possible.

Most printers are equipped with a non conductive bed surface to achieve better adhesion. There would be also a high risk for a short circuit if something goes wrong with the thermistor.
If there's a damage inside the PCB of the heated bed you risk a short circuit between 12Volts and 5 Volts.

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

How about this: build a quality printer that doesn't need to be leveled before every print?

The problem with auto leveling is that it relies on assumptions such as orthogonal axes and a flat bed surface, neither of which are likely to be found in the cheap printers that are typically equipped with autoleveling.

I know it's heresy to suggest quality construction, or that people should consider paying more than $300 for a printer, or that 3D printed part content should be minimized in favor of metal parts, but quality prints start with quality machines.

---

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

Quote
the_digital_dentist
How about this: build a quality printer that doesn't need to be leveled before every print?

QFA
I have yet to find a reason to add ABL to any of my printers. For my most used printer i did the bed leveling the last time over three years ago.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Every printer could be equipped with an ABL and most of them could be calibrated by hand. However if you use Kapton tape for example the height could be differ with every change of the tape. However I use a permanent bed surface which should be the best options.

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

If you change the covering on the bed, whether it is glass or tape, you may have to rezero. Most auto leveling systems use inductive sensors that can't cope with such changes and require manual tweaking to zero the bed anyway.

---

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

What I would like to see is just an image of my bed current situation. During the heat up the printer could do something useful and not just sit and wait.

It could go around the plate and get the points of bed and produce wireframe or heatmap of the print surface. Just to show user everything is OK.

Twisted bed would be symptom of uneven heating or poor build quality.

Naturally the heat expansion changes dimensions of the build plate but current assumed square machine ABL isn't the final solution.

I would never use an inductive or a capacitive sensor for ABL. In fact I meanwhile use a special IR-sensor for ABL as well as for homing my Z-axis and x-axis. The precision of this device is amazing.

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

To the OP I agree with you that to get 3d printers to your grandma then they are going to have to do certain things automatically that we currently like to do ourselves, because we're 3d printer enthusiasts and like technical stuff, your granny or my dad is not. It's not really just about bed levelling.
1. Auto nozzle prep, like an ink jet, the printer will have to heat and auto clean the nozzle and wipe it just before beginning to print in the same way I'd manually remove ooze just before the print starts.
2. Auto adjust the bed to level. The best idea I've seen is motorised screws at 3 points on the bed that physically level the bed, which would be as flat as affordable. Really flat is expensive. Which sensing technology is unimportant if it works.
3. A print surface that is reusable, durable, offers good adhesion and good release. Printbite does all that, so maybe we have that covered.
4. Filament diameter and run out sensing, this is being done. Also filament flow rate feedback or something similar, calibrating the axes is easy, calibrating the extruder is very hard.
5. Heated (to prevent warping), filtered (for volatiles and ultra fine particles) and enclosed (to keep fingers out) build volume, can be done by enthusiasts but there are patent issues when it comes to making a fully enclosed commercial printer.
6. Easy to use software, this is possible but as anyone who has used slic3r on simple mode knows, oversimplify a complicated thing and you are going to get much more average and generic results.

I hope it does become widespread but maybe 3d printing is just too complicated to be something for the general public, maybe the appetite for it isn't there either. No one has made a welder aimed at the general public, or a crane or a cnc Mill. I can imagine a situation where cheap readily available 3d printing shops exist in urban areas for people to get things printed, because that's already here.

Edited 2 time(s). Last edit at 03/28/2016 05:47AM by DjDemonD.

Quote
DjDemonD
2. Auto adjust the bed to level. The best idea I've seen is motorised screws at 3 points on the bed that physically level the bed, which would be as flat as affordable. Really flat is expensive. Which sensing technology is unimportant if it works.

One idea I had for printers that use a bed moving in the Z direction is to support the bed using 3 leadscrews driven by separate motors. Then you could probe three points on the bed, for example using my differential IR Z probe, and adjust the 3 motors to get the bed level. You would need to use ball joints or similar to avoid stressing the leadscrews and the bed.

For small consumer 3D printers, I think the delta configuration is more practical. It doesn't need bed levelling, just auto delta calibration. The RepRapPro Fisher had this, and it was possible to define macros for such tasks as filament loading and nozzle priming. I think it came close to being a machine suitable for consumers, although it needed improvement in a few areas.

---

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].

There is a printer which does not sense the bed, but pads aside the bed, just by electric contact. This is cheap and reliable, and if done with heated hotend, you have no problem with the filament protruding from hotend. As it sense the pads, the bed shall be calibrated against these pads, but the variation between pads and bed can be very low. This simple sensing eliminate a lot of others errors and notably thermal expansion.
On some design like deltas, thermal expansion is important as the switches are at the delta top and you cannot avoid calibration, whatever the construction quality.

As for mechanical force sensing, like the Fisher or FSR sensors, it does have a few drawbacks related to the force needed to operate the sensors which create offsets.
Also, it is quite sensitive to the filament at the hotend. Fisher use Buildtak and surface manufacturer tells to sense with cold hotend, which basically don't work because of the filament remains on the hotend and also thermal expansion while heating the hotend. Indeed, sensing with heated hotend tend to destroy the buildtak and if anything go wrong, you end up with a hole in the surface. I finally settled to a sensing temperature of 155°C, less destructive for the surface and eliminating the filament remains hardness.

I disagree that the answer is good mechanical construction. Sensors became really cheap, or you can do mapping at manufacturing workshop. Any default which can be known either during manufacturing or by sensing and can be eliminated by software is without importance. All consumer products like printers can exist because manufacturers found possibilities to deal with default linked with cheap manufacturing methods. Yes, it needs software development. While programmers are supposed to be paid, this is a one time cost.
The low cost of computing power will enforce this trend. I think that is the real deal. How to get precision from imprecise machines at the lowest possible cost.

Designing low cost stuff have always been more difficult than designing with unlimited budget and I am much more admirative on some low cost car design than on high end cars. And what people needs are low cost cars.

I am not okay with fixing mechanical issues with an electrical solution. My Ormerod 1 has a strange fault. The front of the bed is more than 2mm higher than the back. I had trouble to fix this. Meanwhile I could fix it, but I lost build height.
Before I could use ABL as the Ormerod is generally equipped with a more or less good solution. In theory it worked, but the height difference was too much so the nozzle kicked larger objects from the bed.
However I think you should level the bed nearly accurate or as much accurate as possible (within +/. 0.4mm). Then ABL should support you to eliminate the rest error.
It can be very difficult to achieve a tolernace less than 0.2mm by hand, because if you change one screw the other points will change, too.
My Ormerod 1 has a bed leveling error of about 0.5mm. I am using dc42's mini-IR-sensor-board and it works perfectly, so I will equip my own design with this sensor, too.
The second problem of any solution touching the bed surface is that it will not work if you have two nozzles for dual extrusions.
BTW I ran pre-tests with an FSR and my type was very sensitive so I think it would need a bearly touch only to be triggered if configured well, but my new design uses two nozzles so it will not work with it and my Ormerod 2 will probably equipped with two nozzles, too, after fixing it.

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

Hi again,
Sorry for late response things got busy on campus. I'm so glad with where we are in the discussion now. I'd like to firstly stress that what I agree with DjDemonD; the point is to build a printer for the masses.
For me my 3D printer is the ultimate toy! When it works it's awesome to print things, when it doesn't it's even more fun to fix. I have no problem whatsoever with ABL but for a normal consumer it's too much hassle. The point here is to push the technology to maturation.

@DjDemonD:

1] I think this won't be very hard, I've seen one printer with a metal brush of some sorts where the nozzle is cleaned after heating before each print. Sounds like a good idea.
2] This sounds good, but the cost of the system will skyrocket! you just added 3 motors! I believe 5 motors are too much, one the reasons I like Deltas more. You just added 3. My point is i'm not sure how the added accuracy would make sense against the added cost. How about using the potentiometer? wouldn't work in similar manner but more passive?
3] Good point, never thought about this one. I've always been using Glue and happy with it. But true this need a better consumer friendly way.
4] Do you think this is very critical to the quality of the print? I've seen very good print qualities w/o it.
5] & 6] Agreed.

Your point at the end is well said, not sure I'd agree with it though. I feel people are not interested in the technology only in the results. like imagine a machine that literally duplicates plastic objects. You put the object it, press a button, remove the object, 30 mins later you got another copy. That will grasp people's imagination. Normal consumers that it.


@dc42: 3 lead screws also sounds too complex. As an engineer I'm looking for a simple reliable, cheap way to fix this technical issue. I'm reading more about your ir sensor now. Seems like a good way to go. But it's still is active. I think we can solve this passively too.

@PRZ: I agree on the cost part. That's where the challenge and the fun are. I didn't get the pads thing. Do you have a link for more info?

@Treito: My interest in the nozzle touchy thing is that it's simpler and cost reducing. For now I want to focus on the core of the technology so I'm kinda of letting go of two extruder setup.

I searched a bit and finally found it :
see here on the Lulzbot mini review from Thomas Sanladerer, from 4:40
[www.youtube.com]

Quote
ishe7ata
@PRZ: I agree on the cost part. That's where the challenge and the fun are. I didn't get the pads thing. Do you have a link for more info?

Edited 1 time(s). Last edit at 03/29/2016 06:33AM by PRZ.

The autolevelling bed using motors is complicated I agree and would add cost. A delta with a fixed bed would be easier.

Regarding filament run out and diameter sensing, this is very useful. Your non-techy consumer needs to know if his/her filament has run out and ideally be able to just load a new reel and recommence printing, or possibly start the print again - but the machine could issue a warning when the filament reel is getting low, like an inkjet printer does, the consumer can ignore the warning and carry on, but they will be less irritated if it stops mid print if they were fair-warned. The customer is going to want to be able to use most filaments available. If you want to avoid having this technology on board, they might be happy with cartridges that you manufacture with filament of known quality/diameter etc... but then again its easier to market a machine that "can use most filaments available" rather than "please buy our overpriced consumables". The business case for the consumable approach is better, sell the machines more cheaply and make it back on the consumables, but as a customer it is annoying when manufacturers take this obvious approach, and in so far as I can I'd like to influence people not to go down this route - probably a waste of energy, if this model didn't work then it would not be so commonplace.

As for filament flow feedback or something similar, to actually measure the amount of filament extruded, surely this is essential. Every time I change a reel (unless its for the exact same filament) I have to recalibrate my extruder - the level of grip for any filament is slightly different, the amount of force required to feed it through the nozzle is different, and temperature/nozzle size dependent. So we need a system that basically provides feedback to the controller to calibrate the extruder automatically. This is the technically difficult part of 3d printing (IMHO anyway). Triffid Hunter's guide is great but I'm not sure my Granny is going to want to go through that process. Sure if you're selling the customer cartridges of filament and you can control the properties of it, or encode into the cartridge the settings required for that particular filament reel (as I write the esteps and temperature on the side of my filament reels to speed up the recalibration process) then that's great, but if you want a machine that can print any/most filaments, then solving the auto-calibrating extruder problem (much harder than calibrating axes/bed levelling) is a worthy challenge.

There have been some really good ideas tried using optical encoders on idler bearings and the like.

Quote
PRZ
I searched a bit and finally found it :
see here on the Lulzbot mini review from Thomas Sanladerer, from 4:40
[www.youtube.com]

HA HA!! So my idea works... It's been made before. Music to my ears! ) I have to give it a try then...

@DjDemonD:
I get it now... True I agree that the Filament Runout system should be implemented. How about filament change mid print? How to make it more user friendly? Actually I think it's better to eliminate it. Maybe we can have the printer pause itself, retract unuseable filament, asks for old filament removed then add new filament to the extruder, then the printer goes somewhere far from the object extrude a bit to get the filament going, then goes back to printing. Same procedure could be applied for multi color printing. I think Joseph Prusa is using something very similar if not exactly similar.
The diameter thing, is that really critical? I don't know actually, I've been stuck with one manufacture of filament since I started. It works, so I never thought of changing it although it's almost a no name chinese brand.

that's actually true, since you mentioned it. I keep calibrating my extruder so many time even when I change colors from same brand filament!!!! Agreed, I'd like to work on this one too.....

Diameter is important if it changes between filament spools, or (worse) within one spool. So cheaply manufactured filament had a reputation for having inconsistent diameter throughout the spool. Lets say the diameter goes from 1.5mm to 1.75mm within a spool, that's a 16% change but the change in filament per mm is a squared relationship so its 36% more filament per mm. in this (fairly extreme) example (1.75² / 1.5² = 1.36) There is also the less significant issue that if the diameter changes then the effective gearing of the extruder changes as the drive gear is engaging a wider/narrower second gear (in this case the filament itself is the second gear). Try feeding 1.75mm through a 3mm extruder and see if the same number of esteps always gives the same number of millimeters of filament extruded.

This will have a huge effect on print quality. I haven't encountered many spools with highly variable diameter within a spool so perhaps this is less of a problem than it used to be. I don't generally buy very expensive filament so hopefully they are improving. But there is still variability between spools especially from different sources.

There's also the issue of oval filament - if it is oval then its difficult, if not impossible to calculate its diameter (or effectively cross sectional area), by any easy method that you could build into an extruder. So for enthusiasts/professionals we just measure our filament diameter with callipers and then put this into our slicer software, we check it is not oval by taking a number of measurements at a number of angles to the filament. Sure you could have some optical/laser scanner measuring the cross sectional area of a filament on the fly but this is really going to add cost to the machine.

So you're back to making a consumable cartridge for your customers so control the filament and ensure it works in your machine, this is looking much more appealing in many ways except it is not the approach we'd love to take, i.e. make a machine that can just automatically adjust itself to the filament you feed into it.

Edited 2 time(s). Last edit at 03/29/2016 07:51AM by DjDemonD.

Hey guys I just switched from the inductive sensor for bed leveling to this. The bed leveling is great and I don't have to worry about the bed being metal or not. This works great for me with the printbite bed that I have on top of glass and then the aluminum bed. Check it out. I now don't have to worry about the sensor being so far away from the hotend. This thing is ~12mm away from the nozzle and works flawlessly.

[miscsolutions.wordpress.com]

Yes David's sensor is almost certainly a very useful bit of kit, haven't tried it myself but I am thinking about ordering one. The issue we're discussing though isn't really bed levelling, although that's part of it, it's what do you have to build to make a printer that your Granny (presuming she's not a techy, engineering type) can use to just print things.

Edited 1 time(s). Last edit at 03/29/2016 08:59AM by DjDemonD.

This is pretty close to a works out of the box printer [www.youtube.com] but its £1899/$2713!

Edited 1 time(s). Last edit at 03/29/2016 09:02AM by DjDemonD.

Quote
DjDemonD
This is pretty close to a works out of the box printer [www.youtube.com] but its £1899/$2713!

That is one piece of machinery right there! it's insane! I still think using motors for the bed is an overkill. It's what the engineering team opts for when they have unlimited budget and aim for perfect results. I'm still very interested in the Passive ABL concept. This is what I have in mind so far.


So the thing is, the blue plate is dead flat using some bubble balancing thing to balance the whole machine like that expensive thing you posted. Then you get your reference plane. On top there is the black hollow rectangles which are the potentiometers. When they are totally closed their surfaces make a parallel plane to the dead flat base. But of course they won't be. So you connect each bolt through the glass/metal bed to the potentiometers and screw them enough in place. maybe squeeze a spring in between. Then Arduino can calculate how unwinded the potentiometers are and thus how far away each is from dead flat plate. Then you get your plane. If this works, you get your plane every time all the time without needing any startup sequence. It's a passive system.

Will this work?

Edited 2 time(s). Last edit at 03/31/2016 10:34PM by ishe7ata.

I like the idea and the passive nature of it is good also. I think you can calculate a plane using the method you have proposed. I do not know how easy or quick it will be to build as you have to calibrate this mechanism somehow.

However whilst the potentiometers will be cheap, the dead flat base will not be. Nothing that is nearing truly flat i.e. tooling plate is cheap. Also if you go to the trouble of buying the dead flat plate why not print onto it? Why not build a machine that is orthogonal to the dead flat tooling plate?

You also need to consider that by using sensors you can do more than just calculate a plane. Some Marlin versions (and other firmwares I am sure) will calculate a full bed height map. So you can compensate for an uneven bed as well as a tilted bed. This will be of even more use for the consumer printer, since it will give good first layers even if the customer doesn't clean the build plate etc...

I think you have potentially a good idea, but whether it solves a problem that actually exists I do not know.

Edited 1 time(s). Last edit at 04/01/2016 05:05AM by DjDemonD.

I can see a delta benefiting from an auto level system as all 3 axis move when printing a layer, but I'm with the_digital_dentist on this one. For a cartesian style, build a sturdy printer that doesn't flex and you wont need "auto bed leveling" or rather "auto compensation for a crappy build"

Why on earth would you want to have z movement while printing a layer?
Can anyone give me an honest answer to that?

My Prusa i3 doesn't have it and will never need it. It takes me seconds to check and adjust the z height every dozen or so prints.

Edited 1 time(s). Last edit at 04/01/2016 09:11PM by SteveRoy.

---

My updated Instructable on our Prusa i3 Build
[www.instructables.com]

This is one of those issues where people who have solid printers and know how to level them think its a waste of time. People who have less rigid machines and are not sure how to level their bed think it's great. I know this as I used to be the second one and now I'm the first one.

The trouble is we're talking about consumer, buy it, open it and print something-printers so we need something that's going to remove the manual bed levelling from the process. You can set it at the factory but it isn't going to stay level in transit to the customer.

Deltas assuming they're rigid and have flat beds, probably don't need auto levelling, just auto calibration. But you still need some sort of z probe to do this.

You cannot assume that everyone with a not so good working printer will get rid of it. I probably have already invested 1300€ or more into my both machines.
Also there cannot be any stable I3 out there. Proper adjustment of the heated bed with four screws is not possible.
My printers uses three screws for adjustment but they both suffers on twisted rods. That means that it is not assured that both rods for the Y- axis are parallel.
In that case it's impossible to calibrate or level the bed by hand.
There are some tries out there to fix this, but I haven't found a suitable solution yet.
The only option at the moment is ABL with dc42's great sensor board.

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

Quote
DjDemonD
This is one of those issues where people who have solid printers and know how to level them think its a waste of time. People who have less rigid machines and are not sure how to level their bed think it's great. I know this as I used to be the second one and now I'm the first one.

The trouble is we're talking about consumer, buy it, open it and print something-printers so we need something that's going to remove the manual bed levelling from the process. You can set it at the factory but it isn't going to stay level in transit to the customer.

My local Hackspace has a Tinkerine Ditto Pro, a consumer grade printer. It has 3 manual screws to tram the bed - which generally isn't needed between prints.
Doesn't need "auto leveling". We set it up at our Maker Faire last year, trammed the bed once and printed the whole weekend. Same for the Prusa i3 we built from scratch.

This forum is people building their printers from either kit or from scratch. It seems that a lot of these people would rather spend hours on figuring how to setup "auto leveling" than how to make their printer more rigid. IMHO it's a waste of time - fix the issue that is causing the printer to need tramming every print rather than applying a band-aid solution.
The trend seems to be a race to the bottom in price a quality of kits then kludges to make them work half decently.

Quote
SteveRoy
This forum is people building their printers from either kit or from scratch. It seems that a lot of these people would rather spend hours on figuring how to setup "auto leveling" than how to make their printer more rigid. IMHO it's a waste of time - fix the issue that is causing the printer to need tramming every print rather than applying a band-aid solution.
The trend seems to be a race to the bottom in price a quality of kits then kludges to make them work half decently.

Not everybody wants to change the printer. My opinion is that there should be a suitable solution for the beginning - for the first steps.
I do not have to ABL my printers with each print. Only once powering them on, but if I would transfer the values into the config.g I would have to ABL maybe once a week or even once a month.
At the moment it is impossible to level my printers suitable especially the older one. With the newer one I will soon test a new solution. The printer was no kit I built it by myself and at the moment it is completely disassembled. But again there is a designing issue and I do not know how to solve it as I am no mechanical even though I am designing my own printer at the moment. That means that I have some mechanical background besides my electrical background.

There are for example solutions to fix the twisted rods problem using a screw pushing on the top of one small y-feet (stand). That may work, but I did not test it, because for me this is not a solution that is desperation only. There is one guy who had the same idea as me but his design is much better than my idea would have been. He designed stronger feeds so I am hoping that this will fix the problem. If so My Ormerod 1 will be transferred into an Ormerod 2.

But for first commisioning and first print results if you are new into 3D printing ABL is a great help or do you think you (grand-) parents would be able to use an I3 by themselves?

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

Forget consumers, especially grandma and grandpa. They have no interest in or use for 3D printing. 3D printing for for geeks like us.

One of the fatal flaws in most of our thinking that keeps us trapped in a perpetual state of childhood is an inability to face reality. You want evidence? Look how many adults print Yoda heads and other starwars crap, and cartoonish tugboats. We want things to be as we want them, not as they are, to the extent that we ignore realities such as the value of our time, and tend to over-value money. That leads to the perverse idea, and follow-on behavior, that something that costs $5 and takes 100 hours of labor is better than something that costs $50 and takes 1 hour of labor.

That's how cheap, crappy printers continue to sell even though the community's collective experience should have long ago forced the cheapo kit makers out of business. This "cheapness at all costs" attitude seems to be particularly strong in the US compared to places like the EU, and Japan where people typically demand higher quality for their money.

People will struggle for many hours to get autotramming to work because autotramming is cheaper than building the printer "the right way". It doesn't matter how many hours they spend trying to level and zero the bed or how many hours they spend trying to get autotramming to work. The only thing that has any value is money, and its outlay must be minimized.

---

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

I use quick change hotends on a dual extruder setup so that I can use different combinations of 1.75 and 3mm filament. It's also quicker to change out an entire hotend than to change nozzles for different diameters. In this scenario it would be nice to have Z probing so that I don't need to worry about manually setting the Z0 datum, and once I have it for that I'd probably try ABL as well. If the fractional Z travel looks like a problem I'll turn it off again, but constant Z motion doesn't seem to be a problem when printing in spiral vase mode, so I don't particularly see why it would be a problem with ABL. Because I want to set Z0 automatically it will need to be a nozzle to bed contact system - I plan on trying LeadingLights piezo sensor approach.

In other words, I don't think there's anything fundamentally bad about having some form of ABL *and* well built printers, as well as using it as a way of maximising the capabilities of printers with bed "issues".