Best Auto level sensor & Extruder?

^{(Probably a repeat thread so apologies if it is!)}

Im working on a new printer build and wanted to get some ideas what you all thought on some parts.

What do you think is the best auto-level sensor is in terms of ease of use, repeatability, package size, bed material options, etc. Currently I'm looking at the BLTouch so I can use a glass bed but if its worth not having glass so I can use a PINDA or similar I'd be interested to hear.

Also what do you think is the best option for an extruder? Ive been looking at things like the nimble or the titan. Any thoughts? Anything open source?

Edited 1 time(s). Last edit at 11/10/2018 09:36AM by Ed3D.

The word best is a big red flag!!

those here that make a sensor will say theirs is best.
those that have a sensor will say the one they are using is best.


best is the one you personally like the best. For what ever reason that may be.

Right now I guess either the BLTouch or a piezo for a close second might fit most people's criteria as best, all-around.

I'm making the same decision right now and bought a genuine BLTouch.
I am currently using an inductive sensor with an aluminum bed, it's serviceable, I'm happy with it for the most part. It's limiting if I wanted to use different build surfaces like a thick sheet off glass or needing to be re calibrated when I added thick piece of PEI etc.

The BLTouch won over the Piezo in my case for two main reasons. The BLTouch has a somewhat simpler instillation, I already bought a direct drive Bondtech extruder with E3D V6 so adding a Piezo makes the carriage design that much more complex.
The second reason is the BLTouch can start the auto level feature before the nozzle is warmed up. No point in getting a fancy sensor if I still have to feel like I need to babysit it. A Piezo is a good concept though and in reality it's probably not a big deal for some to wait for the nozzel to heat up, but then again can a hot nozzel probe a PEI bed? Maybe someone will make a Piezo sensor with a flip down leg that works like the BLTouch.

Edited 1 time(s). Last edit at 11/10/2018 09:56AM by Number_5.

Quote
Dust
The word best is a big red flag!!

those here that make a sensor will say theirs is best.
those that have a sensor will say the one they are using is best.


best is the one you personally like the best. For what ever reason that may be.

Absolutely agree! Its a very loaded question. I just wanted to hear peoples thoughts / experiences with different sensors.

Quote

What do you think is the best auto-level sensor is in terms of ease of use, repeatability, package size, bed material options, etc.

My take: [duet3d.dozuki.com]

Tom's take: [www.youtube.com]

Edited 1 time(s). Last edit at 11/10/2018 03:52PM 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].

I have made my printers carefully, true and square. None have needed auto bed levelling. Not trying to toot my own horn but patience while building pays off. Try and make your build level to start off with and you don't have to fool around with this bed levelling nonsense. Your build plate can then have any surface you desire.

Quote
Holy1

I have made my printers carefully, true and square. None have needed auto bed levelling. Not trying to toot my own horn but patience while building pays off. Try and make your build level to start off with and you don't have to fool around with this bed levelling nonsense. Your build plate can then have any surface you desire.

It is true that automatic bed leveling is not strictly needed on a 3D printer, but it is useful.
The build plate may have gone out of flat because a little curlicule of filament had been trapped under it, or perhaps a bit of dandruff from the mustache or a crumb of biscuit shed by a noisome but curious child. Perhaps the nozzle was changed or running repairs done on the frame. Maybe poverty had forced the builder to use a framework unsuitable for 3D printers or maybe other unknown reasons had meant that thermal instability was at loose in the universe. Small earthquakes or persons unfamiliar with 3D printers "cleaning and tidying" could have adversely affected the likelihood of a good print.
More likely still though is the realization that too many prints had been spoiled by the above and other events. The bliss seen on the faces of others blessed with ABL as they started a print secure in the knowledge that it would all go smoothly may also have a lot to do with it.
Mike

If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

Now to set the Z home, a sensor makes thing easier. The best approach, and it is what I use is to detect the nozzle contacting the bed. "Seismic" sensor on the bed, sorry, more issues than worth it.

Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.
DC42 smart effector, deserves its "smart" label, but for a Delta. Piezo is fine too.
The others, the "proximity" are bulky, problematic, some even ridicule. The funniest, a microswitch mounted on a arm actuated by a servo.

---

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

It would be nice not to need ABL, fingers crossed I don't need it on my next build, however I find a good sensor useful for some of the reasons mentioned by leadinglights.

ABL is good for bootstrapping the last bit of imperfections out of my machine, specifically a slight dip in my 8" build plate that might change a depending on whether I decide to heat the bed or not. I still keep the machine as square as I can, the ABL just makes the first layer consistent on the slightly spoon shaped bed. The 300mm mk2a bed would have been unusable without ABL.

I change my first layer thickness often when switching from PETG to PLA so I like having my Z-offset calibrated to actual thickness so I can just change the first layer in software. I feel this requires a decent quality Z endstop, maybe MKSA gets by with a cam and microswitch? One of my machines came with a micro switch for the Z, it looked like a terrible idea, though I admit I did not use it for long. (I needed the ABL for that bed anyway)

Having a probe means one could swap out different build plates, maybe your PEI thickness varies, or even just have quicker rebuilds because with a probe you only have to level the bed, the Z-offset is already set with the probe and firmware. That alone seems worth it.

So I guess I am using the probe primarily as a Z endstop, perhaps the ABL will catch any thermal drift, or perhaps a slight sag in the gantries. I agree it would be nice not to use it, but I don't think I would build a machine without a nice probe and option for ABL.

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

Now to set the Z home, a sensor makes thing easier. The best approach, and it is what I use is to detect the nozzle contacting the bed. "Seismic" sensor on the bed, sorry, more issues than worth it.

Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.
DC42 smart effector, deserves its "smart" label, but for a Delta. Piezo is fine too.
The others, the "proximity" are bulky, problematic, some even ridicule. The funniest, a microswitch mounted on a arm actuated by a servo.

Can you clarify if the "Seismic" sensor is good or bad? This sounds interesting, though I suppose it too requires a warm clean nozzle to work.

I'm interested though, this seems like a neat idea. I am going with a 9.5mm thick bed so if I can get it to stay in position well enough, I may consider something like that, although I just realized I would be back to a hot probe on PEI which wouldn't work I don't think.

Quote
Number_5

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

Now to set the Z home, a sensor makes thing easier. The best approach, and it is what I use is to detect the nozzle contacting the bed. "Seismic" sensor on the bed, sorry, more issues than worth it.

Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.
DC42 smart effector, deserves its "smart" label, but for a Delta. Piezo is fine too.
The others, the "proximity" are bulky, problematic, some even ridicule. The funniest, a microswitch mounted on a arm actuated by a servo.

Can you clarify if the "Seismic" sensor is good or bad? This sounds interesting, though I suppose it too requires a warm clean nozzle to work.

I'm interested though, this seems like a neat idea. I am going with a 9.5mm thick bed so if I can get it to stay in position well enough, I may consider something like that, although I just realized I would be back to a hot probe on PEI which wouldn't work I don't think.

What I call "seismic" is anything that will detect the head hitting the bed and mounted on the bed, like the piezo under the bed as shown in this forum. Add complexity, not worth it, even more so if there are many of them.
In any case, when using the nozzle to bed contact, the nozzle must be clean and all at T°. This is not a problem. I used it for more than a year. Fact, it is better to home when all is at T°which I do. I had to put a piece of Al shim at the Home to protect the bed PEI from being "drilled" by the hot nozzle, giving an incorrect Z0

---

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

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

I would be using ABL to compensate for thermal expansion / wonkey bed mounting. Sure it would be nice to not have it but its much easier and probably cheaper to have it. Id be interested to hear what designs dont need ABL though - sounds like there's some good design there.

Quote
MKSA
Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.

I am VERY familiar with the working of Renishaw probes, they are far to high tech for 3D printing though! (and also have far more degrees of freedom needed)

It is generally best to use the Z probing to tell you about errors while correcting for the smaller ones transparently. For the bigger errors it is a call to find and fix/redesign the bit that is causing the problem.

There are too many possible problems to address without knowing a lot about a particular proposed design but common sense goes a long way - e.g. two different metals connected together will at least try to act as a bimetallic strip. Possible resolutions for this one are to make one of the metal parts thick enough to stop the other moving or to have them only bound together on one end and allow mutual sliding at the other.

I was told of a Delta 3D printer which had been printing without adjustment all winter but which suddenly needed adjustment before every print as summer approached. The problem here was that one of the black anodized aluminium towers was in direct summer sunlight while the other two were in shade. The effect depended on time of day and cloud cover.

It is of course best not to have this sort of thing happen - but you know what they say about the unexpected.

Mike

Quote
Ed3D

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

I would be using ABL to compensate for thermal expansion / wonkey bed mounting. Sure it would be nice to not have it but its much easier and probably cheaper to have it. Id be interested to hear what designs dont need ABL though - sounds like there's some good design there.

Quote
MKSA
Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.

I am VERY familiar with the working of Renishaw probes, they are far to high tech for 3D printing though! (and also have far more degrees of freedom needed)

You don't need to buy one, just to give an idea and make something that fits your carriage/extruder assembly.

If you need to use ABL to compensate for thermal expansion that means your design is wrong.

---

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

Quote
Number_5
Right now I guess either the BLTouch or a piezo for a close second might fit most people's criteria as best, all-around.

The second reason is the BLTouch can start the auto level feature before the nozzle is warmed up. No point in getting a fancy sensor if I still have to feel like I need to babysit it. A Piezo is a good concept though and in reality it's probably not a big deal for some to wait for the nozzel to heat up, but then again can a hot nozzel probe a PEI bed? Maybe someone will make a Piezo sensor with a flip down leg that works like the BLTouch.

You don't need to wait for the nozzle to warm up with a piezo sensor, you just need a clean nozzle. One of the advantages of piezo sensors is that they use the nozzle as the probe so there are zero xy offsets, this makes bed mapping more accurate and is particularly useful for delta calibration. A BLT like approach would lose these advantages.

Idris

---

{Precision Piezo} Accurate, repeatable, versatile z-probe plus piezo discs, endstop cables, pt100, 50w heaters.

Quote
Moriquendi

Quote
Number_5
Right now I guess either the BLTouch or a piezo for a close second might fit most people's criteria as best, all-around.

The second reason is the BLTouch can start the auto level feature before the nozzle is warmed up. No point in getting a fancy sensor if I still have to feel like I need to babysit it. A Piezo is a good concept though and in reality it's probably not a big deal for some to wait for the nozzel to heat up, but then again can a hot nozzel probe a PEI bed? Maybe someone will make a Piezo sensor with a flip down leg that works like the BLTouch.

You don't need to wait for the nozzle to warm up with a piezo sensor, you just need a clean nozzle. One of the advantages of piezo sensors is that they use the nozzle as the probe so there are zero xy offsets, this makes bed mapping more accurate and is particularly useful for delta calibration. A BLT like approach would lose these advantages.

Idris

It is always better to have the printer at working T°. Besides to avoid errors, the nozzle being at T° can be cleaned easily and often not necessary as the melted plastic will be squeezed when homing. It is what I do, even home Z twice. When a job is finished, I remove the part, home and leave the still hot nozzle against the bed. No oozing.

Anyway, when I see a BLTouch on a machine, I know it is a clunker.
To put a piezo on a flip leg would make it a joke.
Ingenuity is in simplicity, not complexity.

---

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

Here's my own experience with Z probes on four printers.

Delta printer: A Z probe is needed to do auto calibration. I used to use the Mini Differential IR sensor, but I was only able to calibrate the printer to a standard deviation of 0.1mm. This was partly because the IR sensor doesn't work particularly well using transparent bed surfaces (PEI in this case), but mostly because the printer had small geometrical errors that made it impossible to get the printing plane completely flat. I upgraded the printer to magnetic rods, linear rails and Smart Effector. Now it calibrates to around 0.023mm deviation. I don't normally use mesh bed compensation because it is accurate enough without.

SCARA printer: I use the Z probe for two purposes: to check whether the printer is levelled and tell me how much to adjust the levelling screws in the feet by, and for mesh bed compensation. I use the Mini Differential IR sensor on this, and it works brilliantly on this printer. The bed is blue painter's tape, which is an ideal surface for the sensor. The design of the cheap SCARA arm means that it sags by 0.8mm at the extremities, so mesh bed compensation is essential.

Ormerod (Cartesian) printer: the printer isn't exactly solid, so the bed tends to go out of level if I move it, and it's handy to have the firmware probe the bed and tell me how much to adjust the screws by. Even though I rarely use mesh bed compensation, I do run mesh probing occasionally to check that the height map is flat enough.

Crane Quad (Cartesian) printer: this doesn't have a Z probe. I really miss having one to make adjusting the levelling screws easier and getting a height map, so I plan to add one.

Note, on many i3-type designs, even if the bed is perfectly flat and level the gentry may tend to sag a little when the carriage is near the centre. Bed compensation is useful to correct this too, although if you know the amount of sag it could in principle be corrected by a single firmware parameter, instead of by mesh bed compensation.

HTH David

---

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].

Great info in this thread.

"Best| really doesn't exist for all cases. It always depends on what you're looking for.

Engineering is a series of compromises. Ideal materials do not exist, so we decide what factors are important to us, what do we NEED, what do we want, and what would be nice to have. If we had an ultra lightweight material that could make perfectly straight frictionless surfaces that could withstand any temperature short of fusion, but be shaped and formed easily, well... I could go on with a wish list for an ideal material, but I think you get the idea.

Our printers are made from materials which are relatively inexpensively available. Some materials cost more than others, and are better than others. My leadscrew pritner frame is aluminum tube and acryllic, which is lightweight, reasonably stiff, easily machineable, and because I had access to a laser cutter, was inexpensive. I would likely not recommend this material to most people, because for the cost, there are better ways to make it. other people might use aluminum extrusion frames. These aren't cheap, but are stiff, and easily reinforced to make good rigid right angles. The extrusion has slots which can be used as a part of the motion control. Other people will use wood and 8mm rods for motion control, which is probably the least expensive option, but requires different skills and tools to build. Those rods also have more possibility of sagging or binding.

Similarly, Z probes are a compromise. We have to balance cost, precision, mass, and physical size constraints. An ideal Z probe would probably be something that could sense the hotend position in absolute space at all times, but we settle for "good enough."

I would say that most of the Z probes offered are good enough, when used properly. So much of what makes one better than another is how the printer itself is designed, and what you can do without losing the performance characteristics of the printer.

Anything on the hotend carriage, will increase the moving mass of the hotend carriage. This is going to mean either a loss of precision in movement changes, or slower movement changes. (IE, acceleration and jerk settings.) The further from the movement arm, the greater effect this will have. The further from the actual hotend that the probing position is, you will need greater compensation, possibly more movement on one or more axes to allow the probe to access all positions of the bed, or give up on being able to probe for part of the bed.

Anything not on the hotend (underbed piezo sensors, Z axis movement switches) has a chance of inconsistent sensing. This can be minimized with good design, but unless you're using something like a delta, these sensors are often also moving mass.

For extruders, these are also compromises. A Bowden extruder lets you offload the mass of the motor from the extruder, which keeps the mass on the hotend carriage down, but the tube itself means greater compensation (And sometimes motor power) is needed in order to get the extrusion timing right. A direct extruder has fewer problems that way, but in turn has greater mass on the hotend. I chose a direct drive extruder, but my printer is not made for high-speed printing. Even then, I've given some serious consideration to converting it to Bowden operation to reduce the moving mass on the hotend carriage.

So for "best" you need to define what's important to you. What do you want out of your printer? How are you planning to build it?

---

MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts

Quote
MKSA

Quote
Ed3D

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

I would be using ABL to compensate for thermal expansion / wonkey bed mounting. Sure it would be nice to not have it but its much easier and probably cheaper to have it. Id be interested to hear what designs dont need ABL though - sounds like there's some good design there.

Quote
MKSA
Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.

I am VERY familiar with the working of Renishaw probes, they are far to high tech for 3D printing though! (and also have far more degrees of freedom needed)

You don't need to buy one, just to give an idea and make something that fits your carriage/extruder assembly.

If you need to use ABL to compensate for thermal expansion that means your design is wrong.

Maybe off topic for this thread but what beds can I use that dont deform under heat but dont use mains voltage?

Quote
Ed3D

Quote
MKSA

Quote
Ed3D

Quote
MKSA
If you look here, you can see that the very few correctly built cartesian printers don't use ABL because they don't need it. If you need ABL, you made a clunker and no "sensor" will be good enough as it will stay a clunker.
Of course a Delta, SCARA or similar it is a must in order to define the origins.

I would be using ABL to compensate for thermal expansion / wonkey bed mounting. Sure it would be nice to not have it but its much easier and probably cheaper to have it. Id be interested to hear what designs dont need ABL though - sounds like there's some good design there.

Quote
MKSA
Can be done by simple high quality contact (see how Renishaw works for CNC), strain gauge, piezo.

I am VERY familiar with the working of Renishaw probes, they are far to high tech for 3D printing though! (and also have far more degrees of freedom needed)

You don't need to buy one, just to give an idea and make something that fits your carriage/extruder assembly.

If you need to use ABL to compensate for thermal expansion that means your design is wrong.

Maybe off topic for this thread but what beds can I use that dont deform under heat but dont use mains voltage?

Thermal expansion MUST always be accounted for in the design !
A thick cast Al tooling plate will not likely bend when heated at the T° used in 3D but will expand and if constrained will bend or bend the part it is fixed too. It is why tie wraps, flexible guides and frame allow poorly designed machine to still work and not bind. This lead to people believing their "design" is good ! This is a lucky example when two bad make one OK This has been discussed many times here and Digital Dentist explained how he did it.

Should I build a new printer, I would try a slab of granite. Of course due to its weight, must only move in Z.
A granite slab would be

---

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

Sure it should be accounted for, but you can only site one example on the internet of a "good" heat bed?

Are you planning on heating the granite too? If you do manage to get the granite up to temperature, it will move almost half as much as aluminum, in various directions dictated by the natural grain of the product.

You realize the granite will have to be at least 3/4" to 1 1/4" thick to not snap in half if you accidentally drop a screwdriver on it?

Thermal expansion worries you, how about a 15 to 20 pound heat bed, is the rest of your frame up to moving that thing consistently accurate to 0.050mm? I doubt it. I suspect the thermal expansion or flex in a frame that can support a 20 pound, 12" build plate is going to be a bigger concern.

Why don't you use Zerodur if you want to be cleaver? Granite?, its you haven't thought this through, sure it's flat, but it has pros and cons just like the other materials. Concrete is cheaper, easier to work with, just cast it on a sheet of glass. It has the same CTE as granite and cheaper, but lacks 3D print snob appeal. You may even add fiberglass to strengthen it's tensile strength. There is a reason they don't build with granite, other then looks, concrete is superior in most other ways.

I have Zerodur here to use on small printers, since the aspect ratio is so thick, it stays flat enough to be used in interferometry. "Optically flat". Like Granite though, it's tricky to heat since it's 1 1/8" thick. I selected cast aluminum for my new build as a pragmatic option to Zerodur since it is bigger, I was having a hard time sourcing a surplus 300mm piece of Zerodur for under $2,000. Retail is closer to $10,000 if I wanted to order it. It would also be over 2" thick, just like the proper thickness of granite would be for a 12" bed. 1:6 aspect ratio is ideal.

Zerodur has a CTE of 0.0100, much better then the granite CTE of 8.4000.

Digital Dentist here did a correct bed and explained how. He answered to you, so no need to repeat. Sad to say, it is one of the very few found on the internet.

I intend to steal the granite slab from a cemetery It is not that fragile and is in fact cheap and flat enough.

Fiber reinforced concrete, I thought about it too, mainly for the frame ! Makes a very good base for a regular machine.

A piece of Schott glass from a dead hot plate would be worth a try too.

Edited 1 time(s). Last edit at 11/13/2018 08:26AM by MKSA.

---

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

Granite used in high precision machines ( Not that used in surface plates though) is usually a granite epoxy composite. I would be quite worried about using either natural granite or a composite granite for a build stage.

MKSA does make a good point though - even if it was not the point he intended when he said

Quote

It is why tie wraps, flexible guides and frame allow poorly designed machine to still work and not bind. This lead to people believing their "design" is good ! This is a lucky example when two bad make one OK

In building a 3D printer, or any other machine for that matter, adding rigidity can easily go beyond the point where any improvement is visible or even measurable. Taking up internal vibrations by having a printer mounted on damping feet can give better results and, (dare I say it with bated breath) even flexibility in the frame can do the same. A careful analysis of why early Mendel Prusa printers printed so well - despite the horrible unbraced rectangle that made up the front view of the printer will give surprising results.

Sometimes though "Lucky" is like this [www.youtube.com] at 3:37, 4:53, 6:20 & 7:05

Mike

Edited 1 time(s). Last edit at 11/13/2018 10:18AM by leadinglights.

Quote
MKSA
Digital Dentist here did a correct bed and explained how. He answered to you, so no need to repeat. Sad to say, it is one of the very few found on the internet.

I intend to steal the granite slab from a cemetery It is not that fragile and is in fact cheap and flat enough.

Fiber reinforced concrete, I thought about it too, mainly for the frame ! Makes a very good base for a regular machine.

A piece of Schott glass from a dead hot plate would be worth a try too.

The only flaw, or perhaps I should say compromise with Digital Dentist's build plate, is that the plate is indexed on one side. This should not be a problem if you leave the bed heated for the whole print. However if you only want to heat the bed for the first couple of layers, the bed will shift. How bad will depend on the location on the bed, the side opposite the indexing pin will see the full expansion of 0.5mm or 0.8mm, or whatever it is.
It may even move more then a "regular" heat bed fixed at 4 corners or 3 points, at least that may stay relatively centered.

Credit due and all that because it is a very nice printer, don't get me wrong.

I understand that a tombstone may stay flat for you, it's just that the hardware needed to support such a stone will be huge and may present a whole new set of challenges of movement and expansion. I look forward to it nonetheless.

At this point I'm not overly concerned about it, but if I did decide to have expansion plates, I would index the heat bed in the center and have one of the Teflon blocks in a channel to keep the bed from rotating. I would have to order a custom heater with the appropriate holes though.

Edited 1 time(s). Last edit at 11/13/2018 06:59PM by Number_5.

Number_5: The issue with thermal expansion here is NOT accuracy but FORCE acting on an overconstrained set up leading to deformation, wear, binding ....

Speaking about accuracy, don't forget to readjust your "pound" as of this Friday the kilogram definition changes

Edited 1 time(s). Last edit at 11/15/2018 02:13AM by MKSA.

---

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

So is this where we start doing away with mains powered bed heaters, and go to propane or natural gas? 80,000 BTU or so should heat a granite slab to 100 deg C in a reasonable timeframe.

I'm willing to accept that the heated bed must remain heated for the duration of the print. I do that anyways for consistency.

---

MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts

I Love my steaks on a hot granite slab, and my £5 granite slab comes in handy.
So Handy it likes to help the other printers out and is now providing support to another over weight printer.

What are we looking at here, a bed-flinger with a granite slab?

If so, the next question would be why, but you already said you got it for 5 bucks. I guess you showed me. lol

No...it's the whole gantry that moves!
I love a Little Red Herring, and ppl's reaction to it....

First question was it cheap?
second, does it look sexy in the photo?
third, has it a real use?
fourth, do I know what it is...?
fifth, ok so whats it's current use...

it's a base for my other printer.

Edited 2 time(s). Last edit at 11/16/2018 02:45PM by MechaBits.

No, you aren't that clever, your picture was obviously framed to be misleading.

What do you mean the whole gantry moves? Red herring?, it still looks like a bed-flinger.