I'm Thinking Of Building My Own Large Scale 3D Printer, And I'm Looking For Some Advice On Part Choices, And Advice On Overstraining Familier Parts From Smaller Printers

Waw, firstly, I’m so glad to see that my thread has attracted so much interest, and I’m glad to see that so many people are contributing, and only a very small percentage of the group has shown negativity.

I’ve not actually read some of the more recent posts prior to this, so I’m actually going to work through the messages and respond to each a bit at a time. Hopefully squeeze in some other concerns of mine at the same time. And as for people telling me to google stuff, well this forum is integrated into google, and all our discussions will be mirrored on there for others to later search for and find. So let’s help others in the future, and discuss as the forum intends us to do so. Since sending people away from this site will ultimately reduce the sites popularity. I found the site via google, and I intend to use it as intended by the people who started it…

Stepper Trouble:
Firstly, I did check out the Pololu Blacks on the site, and they seem quite cheap, so I’ll probable get some of those on my next spend out. Also, talking of drivers, some of you may remember me mentioning about how I had turned the potentiometer anti-clockwise while powered off to find the lowest position to keep things safe. Well, after buying a new NEMA 17 0.9 degree 42N.cm (http://www.ebay.co.uk/itm/121466521423) to test out with, I found that the stepper drivers I have work on a 12 hour clock face type system. I base this on the fact that using my ceramic screwdriver I could turn the power dial with power on. I managed to get the NEMA 17 to turn successfully, and I videoed it to send to my welder to show progress. (http://youtu.be/bEu3Qj3JUxU) However the next day, I tried to do the same again in front of a visiting friend, however the motor was now just moving a fraction back and forth, with a slight buzzing sound. I noticed that the potentiometer was at the 9’oclock position for the motor to work yesterday and clearly I haven’t adjusted it since. Conscious of the temperature of the environment and potentially the circuits having now cooled down, I tried to turn only a few degrees to attempt to achieve the correct setting. As I continued to turn, the bussing went to silence, and it remained that way until I turned it all the way back around to the 9’oclock position again. Swapping out for my other stepper drivers that came in my kit, the same behaviour followed but at different positions on the clock face. I’ve since given up trying to get the green steppers to work. But how can the same fault appear across all the steppers at one go?
What could be wrong here? My Mega is powered via usb, and I got 12v coming in via the black and yellow 2x2 (4-pin) connector from an atx psu. Don’t worry, I do intend to switch to a dedicated supply at some stage. But first, I want some advice on…..

Choosing a Voltage:
I’ve since read (in the posts above) that using a higher voltage can be very beneficial to performance, and with no setbacks? Correct me if I’m wrong there.!! As mentioned above, my mega is currently on my bench powered via USB, and the RAMPS gets its 12v from the Black & Yellow 4 pin from an ATX PSU. I think I had best not install D1 (The diode that allows the RAMPS to pass power down to the Arduino) since I will be using as high a voltage that I can to optimise performance. So my question is, how do I decide what voltage to use? As far as I can see, the higher the better? But there must be an upper limit, or perhaps an optimum? As far as my motors are concerned, well, I’ve not settled for any particular ones just yet. But I am basing my design on the NEMA 23 form factor. And again, I’m open to the input discussed on this thread.

Quote
cozmicray
But sounds like you are RF engineer and will use any wire that is available?

Radio Frequency engineers would tend to use shielded cables due to their heightened awareness of these issues, no? While flattened by your complement, I am in fact not an RF engineer. So I would keep to the electronics and stay away from the crystal ball.

Quote
cozmicray
Again I'll say testing showed 12mm diameter 36" in length solid rods flexes a lot. up/down, side to side, even under tension!!!!!

Ok, no problem. What size diameter do you recommend I should use to ensure I get the precision I need based on the 1500x800x700mm specification? And in the interest in keeping the weight down, would there be any harm in using the thinner rods on the shorter axis; in my case the 700mm Y-axis?

Quote
cosmicray
If you are going to use a Bowden --- my condolences and be prepared for MAX frustration!!

Yeah, I think the Bowden tube is probably going to be a pain due to the oozing, and extra retraction; I can just imagine. And I kinda want a 2 extruder head to print different colours, so that’s extra weight again if I don’t use a Bowden design. I’m still undecided about the use of a Bowden design due to the retraction and oozing issues. Maybe I could use smaller motors on a duel & direct setup? Make use of 3d printed gears maybe? I’d like to hear some thoughts on that.

Quote
cosmicray
The pololu drivers can handle 8-35 V and 1 amp per phase.
Voltage
The higher the output voltage from the driver, the higher is the level of torque versus speed.
You can think of the voltage as the driver of the current. The higher the voltage,
the faster will the current in the windings reach its new target value from one step to the next.
Therefore it is conceivable why a higher voltage will result in better speed performance.

Ok great… I cant see why people don’t just jump in with 35V at all times?

Quote
cosmicray
The extruder hotend and heated bed will pull more on your RAMPS mosfets than the motors!
With your tight RAMPS case --- the components will be toasted real quick.

Yeah, I’ve seen a video where a man burns his finger on a faulty mosfet that was so hot it had worked loose from its 3 solder points. And the one he touched had a decent heatsink! So I think I’ll be getting a heatsink for mine.
I won’t be using a heated bed because of the associated issues, but I may change that out later if I am absolutely stuck. I’m going to be rating the Cuboid for PLA only, so I’m hoping with Kapton or Blue 3M Scotch tape I won’t get much (if any) warping. I’m also going to redesign the RAMPS box to have better air flow since my first design was made before I realised cooling was going to be a big factor.

Quote
cosmicray
Just keep bumbling around -- NOT doing some study from info on the web

Cosmic Ray, your input has been invaluable, but these snide comments throughout your posts are wearing thin, and you’re making a bit of a dick of yourself, and none of the other contributors make these type of comments. Other viewers will see you clearly know your stuff, and will be left disheartened to find that the good information they have just read has come from someone who enjoys putting others down; others who don’t have the skill set you have. For your information, I do indeed use google, and have done a lot of research in to this project. However, I don’t take what I read as gospel because the internet is full of Cosmicray’s and I will always continue to confirm information from several sources before putting my family at risk by powering up a ‘reprap style’ system of this magnitude with full confidence that my family will not suffer as a result of my projects. If you don’t like repeating yourself, or if you don’t like typing information that ‘you’ could find on google in just moments, then please keep it to yourself if it troubles you so much that you have to make a childish close to your messages, since I would rather go without the information than trouble you that much, and continue to allow you to make a dick of yourself on public viewable forums. However, if you can possibly find it within yourself to continue providing this priceless information as you have been, then please do so in a polite and decent fashion.

Quote
jaguarking11
One thing I would like to add. Printing with a large nozzle certainly helps a bunch. I accidentally bored my jhead to 1.54mm and tried to print with it anyway. After some fiddling it will work. You need a geared extruder with a ratio of 2:1 or so to keep the motor happy and that spool spinning fast.
…
Next up I am replacing the hot end with a all metal design. I will be boring one of my nozzles to 1mm or so and use that for higher speed large prints, and keep the variety of nozzles I have at stock sizes ranging from .3mm to .5mm..... I think having a large printer you need a larger nozzle unless you want to print small things in high detail ~1mm nozzle should allow for layer thickness in the .5-.8mm range. If you want to use 1.5mm nozzle or even 2mm nozzle I would suggest 3mm filament. You also need to specify the wall thickness in a 1:1 with the nozzle size, I found it helped preserve the dimensions, which also means you should keep an eye out on designs using parts smaller than your nozzle.

Hi Jaguarking, This is good information! I’m hoping to choose two hot ends that have an easy to change nozzle design so that on larger prints that require less precision. I’m interested to hear about the all-metal hot ends you mentioned if you have some more information that would be great!

Quote
cosmicray
Your huge prints will require days of printing
Will your electronics, motors last for days without melting?
…
ABS printers a usually enclosed -- keeping production chamber warm
--- are you going to put your printer in a huge powder coating oven?

Good points raised here. I won’t be printing ABS at all, I just don’t see the point; I hate the stuff. So the unit will be open on all sides allowing the motors to be a little cooler, and the ramps box will therefore be pulling in fresh air to cool itself down.
And again, please let me know what rod diameter you would recommend for each axis on a printer of this scale.

Quote
jaguarking11
Not necessarily, he could use geared motors and vslot grooves to build something bigger. Or even better he could buy some large and small gt2 pulleys and create his own belt driven gearbox, giving him proper ratios to double the torque of the motors used, at the expense of some speed, but certainly not 1/2 the speed. Nema 23's are available, so are more powerful stepper drivers. Engineering is about problem solving. Sometimes you wont know how it works until you try. My printer is not in the grand scale his is, but It is a large 4.5 feet off the ground, working with nema 17's.... rather well may I add. Certain things need to get larger, other things can stay the same.

I really don’t want to use any gearing made by myself, I’d rather get the correct motor for the task than engineer it to work; it feels a little “make do”. Even though, that probably engineering at its finest

Quote
mariechuria
If you are printing ABS, there will most likely be a need to print big parts in a heated build chamber and the design of the whole system should be taking that into account
…
[T]here is a high current, affordable, Pololu driver-compatible stepper motor driver called Powerlolu.
…
Also, if you have not, check out the large format printer from grassrootsengineering: [www.grassrootsengineering.com]
And look at this video of it in action, printing a kayak in a heated chamber: [www.youtube.com]
…
Can it really be built cheaply while still having appropriate safety standards and a build quality that is making the machine truly reliable?
What about resuming prints? I have not read a lot about this and I wonder what sort of options might be existing to print single large parts with pausing.
So really... I think building such a printer is doable... when you think smart even with a reasonable budget. The real question is- Is it really usable? And will it really be what you are looking for?

Hi Mariechuria, welcome to the discussion. For someone new you’ve provided some great input!
Ok. Firstly, and I know I didn’t cover this before, but it will be PLA only (for now/working progress), so making a heated chamber is not something I’m gonna be doing on this build.
I will defo check out the Powerlolu you mentioned since it sounds like it could be something useful if I have to do something crazy to lift my glass bed on Z with a tidy speed; since I am a fan of Z hopping.
Thanks for providing links to a large printer, im looking forward to checking that out. Will do that as soon as I finish this post. I’m glad you provided the links as opposed to telling me to google ‘grassrootsengineering’, its really helpful to me, and others who will make use of the information in this thread.

Quote
Paul Wanamaker
For larger jobs there needs to be a filament-out sensor and firmware that will detect it, or you can set something in your slicer to pause at a certain height (Cura has this under Extensions).
I know Repetier is planning on supporting a filament out sensor in future, not sure if others have it! I am planning on some jobs that will take more than 1kg of filament, I may combine rolls.

Hi there Paul, thanks for joining the cause.
I was hoping to use an optical sensor to detect the absence of the filament when the spool is empty. But I had no idea how to implement this in the firmware? I’m currently running the Marlin V1 R3 to make use of the Large Graphical LCD you may have seen pictures of in my previous posts in this tread. As part of my early design, I decided for this LCD to allow PC independent printing directly from gcode on the SD card. Surly there must be a way to get the firmware to react to a micro switch or optical sensor and pause execution of the gcode, move the carriage to a position on the x and y axis where it isn’t going to, or has already printed at that location (to prevent ooze from contaminating the print). Then sound a pezzio alarm (on the GFX LCD) to change the filament reel. Press the button once to acknowledge the alarm and silence, then again to continue print.
If you can point me in the right direction for this, that would be great. Even if it means changing out the firmware for other available ones, I’d be happy to try it out…

Quote
dclarkm
I'll return to my lurking and learning....for now. Again, sorry to intrude....please, continue.

You have not intruded, you’ve brought some valuable information to the table in a more polite format than some others have.

Quote
jaguarking11
One thing worth noting. As long as the wires for the end stops are not moving much, you can use cat5 cable to eliminate the cross talk.

OMFG, why the hell didn’t I think of this…. Phenomenal idea…. AND it looks nice!!!

-------------------------------

Well, there ya go… I’m up to date! Lets here some responses!
P.s any rude, disrespectful, insulting, childish or otherwise unaccepting of the different levels of knowledge or research styles from this point forward will be reported to an admin in the interests of keeping the reprap forum an inviting, fun and safe environment for anyone who may wish to join the discussion and/or build their own printer without fear of being made a fool of by perhaps more experienced engineers who feel it necessary to declare their qualifications in an effort to gain credence to online strangers who couldn’t care less, and simply want the support of those who will unconditionally share their accumulated knowledge in an attempt to accelerate the growth of the human race.

I put in my metal hot end today. They are marketed as all metal J-Heads. I bought two of them from different suppliers. I can say that certainly they are not all metal for the 1.75mm version. The thermal barrier uses teflon tubing in it. However even not actively cooled they do run very well with very little disipation to the holder, compared to the j-head. They are far superior, and may I add much cheaper.

Here is a link. [www.aliexpress.com] I am personaly very sadisfied with these. The heater block is larger as well, which means more weight, but better thermal stability when I bore the nozzle out to 1mm.

I will continue to experiment in a few days as I have some things to take care of before then. I will be checking in on your progress. As for the gearboxes I mentioned, the pulleys are bought and so is the belt. The only engineering you have to do is to create mounts for them. Not excessively difficult.....

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Quote
munchit1
hi all, new to the forum but very intrested so far, i'm also finding similar 'all about it bumf' that actualy tells zilch about actualy doing it.

ohh.. and i'm at a similar point on my 1000mm x 500x500mm print area lol.

now so far i was under the impresion the nema 17's would run a horizontal axis weather it was 10mm or 1000mm? theres only the time diference of running the distance?

the Z axis, if it drops the print surface down then so much power isnt needed surely? especialy when the forces are spread over turning 2 or 4 screw threads? (just needs enough to raise the empty bed).

multi motor systems, isnt there ways of adding aditional systems run by the controller? even with seperate power supply's per group maybe? a seperate heater bed sytstem would be very easily acheaved by simpley purchasing temperature control circuits and running them as a seperate system.(theres no real timing to the other motors going on is there?)


it's made me stop and think tbh...as in stop dead, i need it to go in one hit,and so far its sounding like nehhh it wont.

side issue reguarding vibrasions, whys no one useing wood? i came across the wood verses the metals in quodcopter building, the wood helps absorb vibration as well as sound, so am i missing the point on 'vibrasions'? is it the actual stepper acceleration/deceleration 'vibrasions' were talking about? (incorperating missed steps perhaps?).

and my final note is, the extruder system is as fast as it will go any ways....so thats where to 'time to' for the rest of the gear..... i'm a noob guys patients please lol.

Hi there Munchit, Glad to have you here with us.
Firstly, I'm glad you have found this threat a benefit to you already in these early stages of development.

I said myself a few posts back about the time difference in traveling along the axis, since I thought much the same as yourself.
However, I think the point that was made to me is that its how long one axis is to the other.
Your quite right about the length of its travel adding to the time, but if the X axis is pulling the Y gantry along with it across its length, it would appear to need more strength since per mm that it moves along its own axis (X), it also has to carry the Y axis along for the ride.
At least that's how I saw it when originally explained to me. Either way, it is indeed true that more powerful motors are required for the job.

As for the Z axis, my own determination is that the Z axis will need just as much consideration for the power for the motors. Remember that the Z will need to raise and lower in prints that use Z hopping; a feature to stop the brass nozzle drag gin across the top layer of the print as the head crossed perimeters.
Not to mention, once the job is complete the bed will need to return home after the object is removed, and as such the motors will be directly fighting gravity.

I didn't understand the 2nd half of your post, starting "It's made me stop and thing tbh..."
Hope what I have provided helps though!

Quote
Paul Wanamaker

Heating the chamber is no problem, the heated bed will do that all by itself nicely.

Hi and thanks a lot for your input, Paul! Indeed a wonderful delta you have built! Beautiful! ^-^ love it.
You mentioned one point that I myself was wondering quite a lot about... but haven't really found a satisfying answer to yet... You say you don't use dedicated "space heaters" to heat up the air volume in the enclosed chamber, but rely on the heated bed to be enough to heat up the air volume. I totally agree, that makes sense if it works. However, in my case, it is the other way around: because of the way I will construct the bed, it will be very difficult and/or expensive to implement a bed heater that evenly heats up the bed surface. However, I will build my printer like an oven and the internal space will be heated up and the air circulated by means of some sort of fan system as seen in many ovens. Now I am thinking about the option of having JUST this oven heating system that would then over time heat up everything contained in the oven chamber, including the (aluminium) build platform. After bringing everything up to temperature, the printing could be started. I wonder if a "sweet spot" temperature could be found to have proper adhesion to the bed and still have the chamber not at an insane temperature. I have seen successful heated chamber projects with beds at 70°C for ABS, but I am still concerned about this. Does anyone have any input on this?

Quote
munchit1

now so far i was under the impresion the nema 17's would run a horizontal axis weather it was 10mm or 1000mm? theres only the time diference of running the distance?

Well, of course, in theory, you are right. The motor does not care how long the axis travel is. However, when the axis gets longer, everything has to be dimensioned according to that. It is a matter of inertia. You see, many smaller printers use 8mm or 10mm linear bearing rod systems, where the rods just hang "in free air". That might be a valid option for short spans of a couple of hundred millimeters and will work just fine, but as you go larger and larger, the span will be so wide that the rod will noticeably flex / sag under static and dynamic forces, and in certain scenarios, it will oscillate because of that. That means you'll have to go with larger rods or, in most cases, with supported rails. These rails don't have a bearing reaching all the way around them but one that has an opening, so that the rod/rail can be mounted along its entire length onto a stiffer structure like an aluminium extrusion. This enables you to create a very stiff axis at any desired length, which is what you want. But an axis like that weighs a lot more than a free-floating 10mm steel rod of the same length. You can either gear down your NEMA17 motors by means of using belt reduction gearing or by using ACME lead screw drives, but be aware of the backlash and the maximum speed achieveable. I guess the easier and more effective way is to use NEMA23- these motors are not super-expensive. For the axis that just moves the extruder back and forth, you can use a NEMA17 since not too much changes compared to a smaller printer, except for maybe a beefier linear bearing block. Always think of the cascade of axis' that one motor moves.


On this topic: For large axis spans and high stiffness, I have been thinking of teflon (PTFE) linear glide systems. Imagine a relatively thin-walled, but large outer dimension aluminium square tubing extrusion with all 4 or just 3 sides covered lengthwise in PTFE tape. Then build a carriage that reaches around this beam with PTFE pads making contact with the tape. The only problem I see is that the bearing play would need to be adjustable, which complicates things a bit... but this option is going to be a lot lighter and cheaper than linear bearings on chrome-plated, hardened supported rails.

Edited 1 time(s). Last edit at 11/06/2014 04:24PM by mariechuria.

the 'thing' was a typo..sorry, should have been 'think'.

and ahh yes bigger bars weighing more.. good point.

i'm thinking 16mm bars for longest axis btw.. and thinking 12mm will be a bit saggy for the smaller axis. (and yup i see the weight diference a bit now lol...16mm...), or maybe pairing 10mm..

round and round it goes, it's going to be a suck it and see what flavour isnt it? lol.

You can also get hollow hardened, chrome plated rods I think from 16mm and up. Saves a bit of weight.

Using two smaller rods in place of one larger one is not a very effective use of material as there is no sort of connection between the rods (as there can't be) and they will sag and flex still. The key to making something stiffer is to increase the diameter / profile size. A hollow rod using the same amount of material is going to be stiffer than a solid one.

Edited 2 time(s). Last edit at 11/06/2014 04:37PM by mariechuria.

the hollow stuff goes right out of my price range unfortunatly..(by the time i get the rest of the stuff).

You can use precision pipe. Or if you have someone with a mill, you can have some aluminum pipe machined true.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

does the aloy pipe need delrin bareings? or plastick lined as aposed to balls.

Alloy pipe would need brass bushings.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

ahh yup.

Rods
What size diameter do you recommend
Hollow precision hardened rods? There is something that will be exceedingly expensive


If your brain was larger than a pea
you would have seen my recommendation in

[forums.reprap.org]
Vee grove rails on hefty stock

If you must use rods -- perhaps fully supported linear rail
like:
[www.cncshop.com.au]


As for the Z axis, my own determination is that the Z axis will need just as much consideration for the power for the motors. Remember that the Z will need to raise and lower in prints that use Z hopping; a feature to stop the brass nozzle drag gin across the top layer of the print as the head crossed perimeters.
Not to mention, once the job is complete the bed will need to return home after the object is removed, and as such the motors will be directly fighting gravity.

What is "Z hopping"? What 3D software provides for it?
I don't know how you are implementing your Z axis
but tremendous mechanical advantage of lead screws?
The pix you showed in first post used lead screws on z axis?
Get some acme leadscrews or ball screws.
A block of PLA smaller than 1500 x 800 x 700 should only weigh 500 kilos?

It doesn't appear that you read and consider info provided?
That's why I poke at you.
Since I am just a DICK? and the internet is full of me with priceless information,
I guess I can't recommend anything acceptable to you,
I'll let the other X spurts recommend stuff you will ignore.


I'll not post any more to this thread
Just lurk to see how much you break your budget and how big a dumpster is required for your printer

Quote
cozmicray



If your brain was larger than a pea
you would have seen my recommendation in

This thread has attracted the attention of us Admins , please keep the language family friendly and insult free, consider this a warning.

may i add a little about the hightening in the tensioons amoungst us children, one is asuming the other aught to know something... well.. if 'one is that psychic the lottery is the only goal is it not' lol

regaurding a print bed of the sizes mentioned, i doubt the full area will be in full plasticks, but yes, the sizes within those perameters will indeed be a lot heavier in general.

is it a meter cubed of water is a metric ton? lol hell of screw thread to sort that babey at speed lol.

hmmm V slot, cheers for that.

Hmm wow... and I thought this was a nice place...
The V-groove solution is a good and cost-effective choice. However, with just two rails and four bearings forming a linear track, they will not be quite comparable to proper linear bearings on supported rails since radial bearings are then subjected to torsional and axial forces. If they use two bearings per wheel, there'll still be the axial load and probably axial play. I doubt they'll use tapered or shoulder bearings. But they'll certainly be fine for many 3D printers. The BigRep One is using V-slot linear track systems as well. However, note the slotted holes on the tracks that allow the play of the system to be adjustable. This is crucial for a solid track. If the rail for the V-slot bearings doesn't allow for that, one would have to have the top or the bottom row of grooved wheels mounted in an adjustable way, which complicates things again.

If you want to go really cheap and gigantic, why not use extra hard (99a or even harder) skateboard wheels on anodized aluminium extrusion? You could get a circular aluminium tube and have three skateboard wheels riding at 120° spacing. Like they did in the old days with radial bearings riding on hardened steel rod. Look into the systems people build as camera dollys and abstract from that. If you are not familiar with the concept, you can just watch this video.

You could try something like Supported Rail or Slider for extrusion or even Rails on Extrusion with Wheels

There are lots of good stuff out there it is just what you are willing to pay?

as for nice light weight multi extruder hot ends you can't go far wrong with The Kraken This does need to be Bowden fed of course and can have nozzles of upto 0.8 mm as standard (They do a spares selection ranging from 0.25 to 0.8mm in both Brass and Stainless) see Nozzles

HTH

Doug

(I am in the process of building a Core XY machine with a build of 400-450 mm in both X and Y and 550 in Z (This could easily be increased by replacing my Z Rods and shuffling it all upwards (My verticals are 1000mm tall). can prob get to 850 mm by doing this).

Edited 1 time(s). Last edit at 11/07/2014 08:22AM by dougal1957.

hi stumbled in this thread looking for big 3d printer, because i also plan to built one, still in the drawing board, 500(X)x800(Y)x1400(Z) mm build volume, 1066x1066x1950mm footprint... fun to read thread... so 3 months lapsed since the first post, 2 months lapsed without update. any update on the big printer progress or photo?

Things you should consider also are the gantry style and frame rigidity. I have been successful in building a 16"x16"x16" 400mm x 400mm x 300mm, however it was slow(40-50 mm/s movement speed and had issues with the bowden setup) Your best bet is to use a strong frame made out of metal. Precision is a must for the frame and it must be extremely accurate otherwise your printer will have difficulty with higher resolutions.

Using a gantry similar to the ultimaker like i did which will give you decent results however the inner rods are hard to keep parallel to the driving (Spinning) rods on the frame.

looking back at my project, a gantry style like the Makerbot and clones, will likely yield better success because of less moving parts.

Your bed will be heavy if you are making a cube. There is two options:
A) You can raise the whole gantry with a fixed bed(possible but a pain when it comes to Z wobble)
You can lower the bed as it prints(as most do and I did) This will require alot of design into the Z axis to ensure that each side raises and lowers without sagging. you will likly need two sides supporting the z axis with 4 LM8uu's per side.

It is possible to go big, I went through about 7 iterations before being successful with my printer. Like others said printing big objects takes a ton of time. It wasnt worth it to me to keep printing with such a huge printer, I would rather have a army of mini's printing each segment of the model than one big printer making stuff that if it failed would be a huge waste of filament.

Over 1.5 ft, size multiply drasticaly and exponentialy cost, problems and the needed overall precision. A CNC mill can be large because you don't need an overall precision of 0.2mm, but will be pretty expensive anyways. For FDM technology, It looks better, quicker and cheaper to organize your manufacturing process for a bunch of small machines printing at the same time. Assembly time of several parts can be realy quick if you consider that at the start of your process. IMHO.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Concerning a filament out sensor as discussed above:
This is now easily doable. I've just upgraded to a Smoothieboard, and there is a feature for this, and also a pause/resume button, if you want to switch filament. It can move the print head out of the way, retract the filament, etc (whatever g-codes you tell it to do). It can then resume where it left off. I've not done it yet, but the functionality is there. I suppose you'd either need to tell the printer not to turn off the motors, or to home afterwards... I'm loving this board, so many problems gone.

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!

Quote
Paul Wanamaker
Concerning a filament out sensor as discussed above:
This is now easily doable. I've just upgraded to a Smoothieboard, and there is a feature for this, and also a pause/resume button, if you want to switch filament. It can move the print head out of the way, retract the filament, etc (whatever g-codes you tell it to do). It can then resume where it left off. I've not done it yet, but the functionality is there. I suppose you'd either need to tell the printer not to turn off the motors, or to home afterwards... I'm loving this board, so many problems gone.

Funny you mention this feature Paul. When doing the initial Delta motion tests using a pen to draw a design I was able to pause, move the head up and out of the way using pronterface axis controls and upon pressing resume the printer effector dropped back down and continued where is left off. Not sure if other boards allow this but certainly a fantastic feature to have.

---

My delta build blogspot [d3delta3d.blogspot.com.au]

Custom Delta printer
300x500 build volume
magnetic effector
Smoothieboard controlled