Options for auto leveling?

I am currently designing a large scale Core XY 3D Printer, about 18 inches x 18 inches for the print bed. So obviously leveling a bed that large would take a long time and be frustrating, so I plan to use a type of auto leveling system. I was thinking about using an all aluminum bed fixed to the x-axis chassis, but I checked the price for a pre-cut size of 18 x 18, it would cost around $150. Another idea I had was was to get a pre-cut acrylic bed that would be also fixed to the chassis. But the problem is, what would I use to auto level the acrylic bed? Is there a probe like the ones induction probe, that would sense the acrylic bed? Or would I have to resort of the auto leveling arms that are connected to a leveling switch.

Also, would adding aluminum tape around the corners of the acrylic bed, would the induction sensor sense the acrylic bed?
Example, this SmartrapCore. It looks like they use aluminum tape on the side of the glass bed.
[shopping.smartfriendz.com]

Edited 1 time(s). Last edit at 03/05/2015 11:53PM by TheBoilingDumpling.

Tom S. on youtube said he tried metal tape with his inductive sensor and did not have luck. Of course, your mileage will vary depending on the tape and the sensitivity of the sensor.

You could always glue some magnets to the other side of the bed, but I think that would perhaps leave you with a very accurate reading biased somewhere to the left or right of where it should be (by way of error/differences between magnet attachment)

I'm an inexperienced reprapper, but I think forcing your way into an inductive sensor solution is more trouble than it's worth.

I'm using a microswitch without the metal horn on it. I don't have much in the way of repeatability issues. It helps that I'm sending G29 E which keeps my probe extended throughout the probe sequence. This prevents a cheap servo's failure to exactly replicate the extension position from influencing your tilt compensation.

Yeah, I think I'll probably just go with the limit switch style. But which one do you have? Is it like this?[www.thingiverse.com] Or on like this [www.thingiverse.com]
Another question, how reliable has it been so far? Has it been accurate?

Edited 1 time(s). Last edit at 03/06/2015 01:41AM by TheBoilingDumpling.

I have tried lots of different types of height sensor. My favourite so far is a differential infrared sensor. The details are at [miscsolutions.wordpress.com]. It works by reflecting two infrared beams off the bed surface. So it works on most bed surfaces except very dark ones. I manufactured and sold more than 100 of these to Ormerod printer owners, and due to continuing demand I intend to make some more soon.

The other probe type I am looking at is a mechanical probe on the end of a miniature solenoid. Contact with the bed is detected by an optical sensor or by measuring the change in inductance.

I have also experimented with capacitive, inductive and ultrasonic sensors, but these all have disadvantages.

Edited 1 time(s). Last edit at 03/06/2015 03:05AM 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].

Hmm, so many choices, which one do I chose! Which one would be the simplest and easiest that would require not much maintenance to keep it running throughout the years?

One other consideration is that a mechanical probe that moves against a spring to operate a switch typically needs quite a lot of force to activate it. This isn't much of a problem when using a glass bed, however on an acrylic bed, repeated probing at the same places may cause wear. So I suggest either a non-contact probe such as the differential IR one, or a contact probe that needs only a small force to activate it.

What colour would the acrylic be? The only colour acrylic sheet I have here to test an IR sensor on is black. What do you intend to coat the acrylic with to help the print to stick?

When you had that quote of $150, what thickness was the sheet? I can get marine grade 500mm x 500mm x 5mm sheet on eBay UK for about half that price.

---

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].

I mean the $150 including shipping, was for the aluminum sheet (Metalsdepot), it is 3/8 thick. It may be overboard, but I was to resist bending. For the acrylic (Tapplastics), its about $35 for a 18 x 18, 3/8 thick but I did not check the shipping.
If I chose the acrylic sheet, it would be clear. I would be either using Elmers gluestick or Blue Painterstape.
Honestly, I'm not too sure if I should go with the aluminum fixed bed, and use the induction sensor. Or use a acrylic bed with leveling springs, or use a type of contact probe, or as you mentioned a IR sensor.

Whatever you go with, I would avoid using bed springs and a mechanical probe. Perhaps I had weak springs, but I found that the force required to press the switch caused the springs to compress on the edge which was being probed. Obviously not good.

Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

3/8" seems thick for a bed. I'm using 1/8" thick.

Quote
ggherbaz
Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

Not so. If you print on a glass bed as I do, it senses the distance to the material under the glass (aluminium in my case), although putting tape, glue etc. on the surface also affects the reading. Not only does this make it less repeatable because of the greater distance, but I use several glass plates, and I can't guarantee that they are all exactly the same thickness. The same considerations apply to inductive sensors. So I only consider sensors that measure the distance to the top of the printing surface, such as contact and IR.

If you print directly on to an aluminium bed (perhaps with a thin coating on top), then both capacitive and inductive should work well.

---

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].

This is also a pretty cool idea

Precise Z probe

I have two printers that work fine with the inductive probe. The inductive probe requires a aluminum bed to work properly, one printer has a 1/8" aluminum bed and the other has a MK3 aluminum heated bed. They both detect the bed at about 2mm from the surface of the bed. The cost of the probe and resistors necessary for the correct signal voltage was about $5 each. Then all that is needed is a mount for your specific printer.

[www.youtube.com]

Edited 1 time(s). Last edit at 03/06/2015 12:32PM by normw.

dc42,

I don't share your opinion about capacitive sensors. They sense any material including water. My 2 printers used this method and one have a pcb board and the other an aluminum bed, both with glass on top.

Repeatability: I test every 5 to 10 prints over 100 prints so far and not once my spacing between hotend and bed have been greater or smaller than 0.01 from the 0.1 I use. The difference between a 70 an a 110 degrees bed it's so small that doesn't require adjustment.

The only issue they have, and I'm willing to live with it is weight, they are heavy and they are not suitable for deltas or if your intentions are to print at really high speeds.

Capacitive Proximity Sensors
Capacitive Proximity Sensors can detect nonmetal solids and liquids in addition to standard metal targets. They can even sense the presence of some targets through certain other materials, making them an ideal choice in some applications where other sensing technologies simply will not work.

You can also read this:

[m.machinedesign.com]

Edited 1 time(s). Last edit at 03/06/2015 12:55PM by ggherbaz.

I am both a physicist and an electronics engineer. I can assure you that a capacitive proximity sensor will not sense the surface of a glass bed, or any other good insulator. It can detect a conductive or semi-conductive plate under a glass bed. It can also detect the difference between glass and no glass between the sensor and the underlying conductive surface - but that is of no use for a height sensor. This is all basic physics. And yes it can detect water, because water is a relatively good conductor. I have used capacitive proximity sensors to measure water level. I have tried using one as a height sensing device for a 3D printer. It worked, but not consistently enough for my situation (3mm or 4mm glass on top of aluminium, with interchangeable glass plates and various coatings on the plates to hep the print adhere, temperature range 20C to 100C).

Bear mind that we are not just trying to detect whether the bed is close to the sensor. We are trying to get a trigger height that is reproducible to within 0.05mm over a range of temperature, humidity and bed coating.

It's also entirely possible to make a capacitive sensor that it light enough to use on a delta printer - the one I made for height sensing weighed only a few grams.

Edited 1 time(s). Last edit at 03/06/2015 01:38PM 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].

Capacitive sensors detect insulators at a different distance to conductors. If you have an insulator over a conductor it will trigger at a bigger distance than either on their own. So if you have glass over metal it responds partly to the proximity of the metal and partly to the thickness and proximity of the glass, not to the top surface of the glass as you would want.

Capacitive, inductive, hall effect, and optical are all analogue measurements compared with a threshold. They can be affected by supply voltage, temperature and stray fields.

My preferred solution is a pair of gold contacts, you don't get much more digital than that. Pulled together by a magnet, pushed apart by a rod and pushed further to latch in an up position held by the magnet. The only electronics requires is a digital input pin.

---

[www.hydraraptor.blogspot.com]

Quote
DRobs86
Whatever you go with, I would avoid using bed springs and a mechanical probe. Perhaps I had weak springs, but I found that the force required to press the switch caused the springs to compress on the edge which was being probed. Obviously not good.

I mean the traditional of the 4 screws with tension spring on each corner, and use the paper leveling technique. But I think that would be really frustrating considering the large bed area.

Quote
ggherbaz
Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

Hmm, I'm not quite sure about that!

Quote
TheBoilingDumpling
I mean the traditional of the 4 screws with tension spring on each corner, and use the paper leveling technique. But I think that would be really frustrating considering the large bed area.

You should do that anyway to start with, so that the automatic bed compensation is just used as a fine adjustment.

---

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
madmike8
3/8" seems thick for a bed. I'm using 1/8" thick.

It is to make up for the large bed area, so I can avoid warping or sagging.

Quote
dc42

Quote
ggherbaz
Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

Not so. If you print on a glass bed as I do, it senses the distance to the material under the glass (aluminium in my case), although putting tape, glue etc. on the surface also affects the reading. Not only does this make it less repeatable because of the greater distance, but I use several glass plates, and I can't guarantee that they are all exactly the same thickness. The same considerations apply to inductive sensors. So I only consider sensors that measure the distance to the top of the printing surface, such as contact and IR.

If you print directly on to an aluminium bed (perhaps with a thin coating on top), then both capacitive and inductive should work well.

So does that mean, if I were to use Blue Painters Tape on top of the bed, a leveling system that would make contact to the actual bed would be more preferred?

Quote
madmike8
This is also a pretty cool idea

Precise Z probe

That's pretty cool, I will think about it!

Quote
TheBoilingDumpling

Quote
ggherbaz
Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

Hmm, I'm not quite sure about that!

Capacitive will detect any conductive surface. It will also detect any insulator with a high relative permittivity compared to air but it responds to its bulk, not its surface. Both increase the capacitance of a single plate capacitor.

---

[www.hydraraptor.blogspot.com]

Quote
normw
I have two printers that work fine with the inductive probe. The inductive probe requires a aluminum bed to work properly, one printer has a 1/8" aluminum bed and the other has a MK3 aluminum heated bed. They both detect the bed at about 2mm from the surface of the bed. The cost of the probe and resistors necessary for the correct signal voltage was about $5 each. Then all that is needed is a mount for your specific printer.

[www.youtube.com]

I haven't seen a EZ3D in a while, nice to see it working well!

Quote
ggherbaz
dc42,

I don't share your opinion about capacitive sensors. They sense any material including water. My 2 printers used this method and one have a pcb board and the other an aluminum bed, both with glass on top.

Repeatability: I test every 5 to 10 prints over 100 prints so far and not once my spacing between hotend and bed have been greater or smaller than 0.01 from the 0.1 I use. The difference between a 70 an a 110 degrees bed it's so small that doesn't require adjustment.

The only issue they have, and I'm willing to live with it is weight, they are heavy and they are not suitable for deltas or if your intentions are to print at really high speeds.

I think I will be able to print at high speed, because I am using a Core XY system, so the bed won't be moving as much as the hotend will be.

Quote
dc42
I am both a physicist and an electronics engineer. I can assure you that a capacitive proximity sensor will not sense the surface of a glass bed, or any other good insulator. It can detect a conductive or semi-conductive plate under a glass bed. It can also detect the difference between glass and no glass between the sensor and the underlying conductive surface - but that is of no use for a height sensor. This is all basic physics. And yes it can detect water, because water is a relatively good conductor. I have used capacitive proximity sensors to measure water level. I have tried using one as a height sensing device for a 3D printer. It worked, but not consistently enough for my situation (3mm or 4mm glass on top of aluminium, with interchangeable glass plates and various coatings on the plates to hep the print adhere, temperature range 20C to 100C).

Bear mind that we are not just trying to detect whether the bed is close to the sensor. We are trying to get a trigger height that is reproducible to within 0.05mm over a range of temperature, humidity and bed coating.

It's also entirely possible to make a capacitive sensor that it light enough to use on a delta printer - the one I made for height sensing weighed only a few grams.

Apparently from the the graph on [m.machinedesign.com] a Photoelectric can sense Silicon, Plastic, Paper, Metal, etc... Wouldn't it be able to sense a acrylic bed if it was a color other then clear? Because it use a light transmitter to sense a object, and a receiver to catch the beam I think.

Quote
nophead
Capacitive sensors detect insulators at a different distance to conductors. If you have an insulator over a conductor it will trigger at a bigger distance than either on their own. So if you have glass over metal it responds partly to the proximity of the metal and partly to the thickness and proximity of the glass, not to the top surface of the glass as you would want.

Capacitive, inductive, hall effect, and optical are all analogue measurements compared with a threshold. They can be affected by supply voltage, temperature and stray fields.

My preferred solution is a pair of gold contacts, you don't get much more digital than that. Pulled together by a magnet, pushed apart by a rod and pushed further to latch in an up position held by the magnet. The only electronics requires is a digital input pin.

You're talking about this right? [www.openbeamusa.com] I have seen a bunch of variation of auto leveling on Delta bots.

Quote
nophead

Quote
TheBoilingDumpling

Quote
ggherbaz
Capacitive proximity sensor instead of inductive, simple to set and will detect any material.

Hmm, I'm not quite sure about that!

Capacitive will detect any conductive surface. It will also detect any insulator with a high relative permittivity compared to air but it responds to its bulk, not its surface. Both increase the capacitance of a single plate capacitor.

I wont be using a glass bed over the aluminum bed, if I do decided to go with the aluminum bed.

Quote
TheBoilingDumpling
Apparently from the the graph on [m.machinedesign.com] a Photoelectric can sense Silicon, Plastic, Paper, Metal, etc... Wouldn't it be able to sense a acrylic bed if it was a color other then clear? Because it use a light transmitter to sense a object, and a receiver to catch the beam I think.

Yes. A single beam is OK as a proximity sensor, but poor as a height monitoring device (which is what we need) because its sensitivity depends on how reflective the surface is. The RepRapPro Ormerod uses such a sensor, and you have to put white targets at the corners of the bed (which are the probing points). My differential IR sensor uses 2 beams to avoid that. So it works on a surface of plain glass, Kapton tape, glue etc. However, if the bed is transparent and the surface underneath is very reflective, the trigger height is affected a little by the reflection from behind the bed.

---

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].