First Printer, 225mm^3 printer / mill combo build

Well, I think it is time to solicit some feedback on the printer I've been working on. With that said, the design is mostly complete, I just have a few more parts to fab up.

Background:

I received some 'freebie' industrial linear rails from a co-worker. They are 10 years old, but were still in the original packaging. They have a slight bit of surface rust, but nothing that affects the performance. They are strong enough to carry me around, so they are far over-built for this project. Free is free, I'm not complaining .
The motors were also a freebie!, 4x NEMA 23 Triple stacks at 286oz/in holding torque. 0.6 ohms, 5.6 amps, 1.8 degree per step by Automation Direct. (Company surplus and a generous boss)

So, with my two freebies in hand, I could start designing and building my "almost free" 3d printer!

Or so I thought...

Anyway, here is the design so far. I've used 40mm misumi extrusion, and 3/16" cold rolled steel for all the brackets, waterjetted by a friend of mine.

The build volume is 9x9x9", or 225 mm^3, set largely by the travel of the linear rails.





As it sits currently:


Plans:
Electronics: Azteeg X3 Pro with 8x SureStepper ~2A drivers. [On Order, one drive per Z Axis, with room to expand]
Heated Bed: Glass over QU-BD 225mm x 225mm 24v silicon bed [On Order]
Power Supply: 24vdc 20A variable lab supply. [Temporary until I determine current draw, I can run up to 35vdc if Azteeg will handle it. (TBD based on stepper performance)]
Firmware: Marlin
Extruder: Geared Hinged Wades Extruder with Nema 17 5kg*cm motor [On Order]
Hot End: J-Head [Source TBD], Bowden setup. Redundant thermistors planned

I've left room for a second J-Head in case I get crazy with it.

I've also included space on the tool plate, and offset the tools such that the tip of each tool can reach the entire 9x9x9 build volume.

The spindle in the CAD is a rough mock up of a Dremel Fortiflex (1/5)HP, 5/32" shank flexible shaft setup with a 3 jaw chuck. This is a future add-on, for which I'll probably print some brackets.
In the event that I do go for the mill setup, I'll be using a sacrificial MDF base on top of the 1/2" Lexan shown in the CAD.

So, I guess I'm looking for some feedback, things to watch out for, etc. Like I said, some of this is set in stone since it is already built, but then again, with extrusion almost nothing is set in stone. I have already been warned that flex couplers on a z supported axis can be troublesome, so I'll watch for that.

For a medium-size printer looks very nice using good linear ways and aluminum construction you might have to add a couple more structural pieces but that'll be a piece of cake. You're definitely on the right track their here is a picture of mine a little bit larger with ballscrews

Edited 1 time(s). Last edit at 02/13/2014 08:00PM by cnc dick.

Thanks! I've seen yours on other posts. Looks pretty impressive. I can tell you've done some industrial work.

If you look at my original photos, you'll see the four 8mm holes off the top of the mast. Those are for gusstets and/or turnbuckles like you have. I've left them off for a few reasons...

A ) I don't think they will be required, as the black gusset is pretty darn rigid, remember this is 40mm, not 20mm extrusion, so it has a contact patch at the bottom with an M12 threaded from below. Same with members aligned with the Y axis... 8 more M12s threaded in.
B ) I felt like they would be in the way and I'd catch my hands/wrists on them.
C ) I don't have them, and would have to fab them. When I see the right size stainless go through the scrap bin I'll secure them for future upgrades

Here is an old screenshot where they were "installed". I would either slot one side or replace with turnbuckles if applicable:


Edited 2 time(s). Last edit at 02/14/2014 10:36AM by Tired2.

Something else I provisioned for and hope to add once the printer is printing is a lower and possibly upper thrust bearing on the Z threaded rod.

I checked the thrust rating on my steppers, and luckily since they are the high torque model, they can handle 13lbs of thrust load each! So, as long as my moving Z stage is under 26lbs [11.79kG], I should be okay in the short term without them.

Either way, i just need a bracket that a bearing can seat in. A few Nyloc nuts and a 608zz should lock the thrust in a bit and stabilize the runout a bit. Even if the thrust load is not in need of improvement, the 'sag' of the flex couplers carrying the load will be minimized. I'm just realizing they will have to handle the full weight of the moving Z stage, which might be a bit much for them

Axial contact bearings would be nice on the screws the reason I posted a pic is you just registered yesterday I thought you were brand-new in these areas obviously you've built things before but didn't know how much you had researched 3-D printing do you have access to machine tools. I personally am more interested in your electrical hookup with the azteeg x3 and drivers. I have some Nema 23 motors linear rails and a bunch of 8020 laying around and I've been thinking about building a small portable machine like yours but I was never a fan of the reprap electronics it will be nice to see the results of the azteeg x3 on a correctly built machine

Edited 1 time(s). Last edit at 02/15/2014 03:19AM by cnc dick.

I am new here, and to "diy" 3d printing, but I've been researching and lurking for the last 3 months or so. I also have an electronics / programming background in the industrial automation industry, so that helped a lot starting out. Things like PID loops and axis homing, CNC motion are all things I deal with regularly, so the parts I needed brought up to speed on were the 'state of the scene' sort of stuff. There is a lot of info out there to weed through.
I also maintained 4 pro grade machines for a while as part of my job duties at a former employer... Stratasys FDM, 3d Systems, and ZCorp.

Anyway, I do have access to a manual mill, and a lathe, and various other tools.

I started filming a youtube series on my printer, so as I get those uploaded I'll post them in this thread. I mostly skipped the build up until now because it was a very "bootstrapped" project that I don't expect others to really benefit from, but I'll be doing video documentation on the process to bring the printer online from here on out. Essentially what I want to convey in the video series is how to rep-strap a machine, starting with basically what I have here so far.

I've found that doing some videos on a particular topic forces me to really think things through and I don't mess up as much. Of course, being new to this I also stand to risk putting information out there that is untested, but I make that clear in my videos.

As for the Azteeg X3, and drivers and stuff, I looked through the Marlin code a bit, and nothing stands out as particularly challenging. It seems like it is really well done. I'm very curious to find out how well it can handle running G-code from mastercam and various other CAM packages. I did consider using EMC2 or Mach3 to drive the steppers, but I really like that the Azteeg can run without a computer attached, and it handles all the heating on its own. I'm also space limited a bit at home, and I hate having to set up that old 600MHz beast computer when an 8bit AVR will do most of the work.

I think Marlin, with only slight tweaks should handle everything I need, and with the auto-bed leveling feature, it seems awesome. I can also probably leverage that to be an auto-tool length detection for milling if/when I get around to the mill aspect of this machine. (Think dremel wants to sponsor a machine?)

Alright, back to work. Let me know if you want the solidworks models, I'd be happy to share. PM me if you want and I'll share the whole dropbox folder if you indeed plan to build something similar and can benefit from my work.

I thought I mentioned this, but maybe not, so I will here...

I am a bit worried about how much torque I'll actually need. My steppers can do 5.6A, but I'll only be able to provide about 2.5A (continuous) using the surestepper drivers. I really want to stick with the small pololu package for cost reasons, so hopefully that is enough. I that assumes that for a given motor at a given RPM, the torque is proportional to the current.

I'm basically leaving over half the motor's rated torque unused, because I can't supply the current, which sucks. I'm also running the DRV8825 chips at max current, which is asking a lot of them, whereas most reprappers run them at reduced current to keep the motors from burning up.

Additionally, these particular motors have almost double the torque output at higher voltages.

Here is the response graph. As the speed goes up, the torque available is almost cut in half at 1350 RPM on 32v vs 70v. The Max the drivers will handle is 35v, and I'm not sure the components on the Azteeg are up to the job of 35v. (It only mentions 24v everywhere I've seen, and Panucatt has not been that responsive in related inquires about obtaining documentation for the Azteeg X3 Pro :-/ )



But, if 24v works, I'll just leave it there because it makes other things easier. I think the Z Axis will be okay, since it is going to be pretty low RPM, where most of the torque is available at any voltage.

Maybe it is time to break down and do the math.

Like I said I'm not crazy about reprap electronics I was hoping you could do better than what was out there I will keep an eye out and see how it goes keep us up to date. I use Mach 3 and it works fantastic anything what a circular path or radius is nice and smooth everything with reprap electronics and their firmware seems to be segmented. Most of the slicers seem to do a good job it's just the electronics and the machine operating software firmware of reprap. Could also be a course resolution with the belts

Edited 1 time(s). Last edit at 02/14/2014 02:37PM by cnc dick.

From what I have read there have been some pretty big improvements in arc segmentation in Marlin. I have no experience yet though, but I saw some prints that look good.

These curves look good: [thingiverse-production.s3.amazonaws.com]

They are from sailfish firmware, but I think everything in sailfish with respect to curves is the same in Marlin? I'm still a bit fuzzy on all that, it seems sailfish only works on makerbot, or something like that.

I think the belts should work out just fine, but we will see.

I have a plan to re-purpose the Azteeg for a few other potential motion control projects, so if I have bad luck I'll move it to that where the motion coordination is not near as critical.

I would like to investigate if there is an easy way to feed step/dir pulses over the USB serial connection to the Azteeg and just have them be passed through... basically, use the serial data to send a 20khz or whatever pulse rate and have the AVR parse that stream and output it to the pins. That way you could still have it run PID and all the other goodies, and house and current control the stepper drivers, but let Mach3 or EMC handle the mill / print step generation for nice pretty curves and mill-centric interface (spindle control, offsets, etc, etc).

I dunno, sounds like a lot of work for added convenience.

I think I'm going to wait and see how yours prints because I know you have a nice sturdy accurate machine and bad mechanics won't be in question. I've never operated a professional machine but have worked on plenty of prints from them the FDM and SLA I used to be a prototype technician for a large Pro audio company so I had to finish and make temporary silicone molds from Masters and sometimes get parts out to a trade show before molds were made. And I have never been happy with the reprap print results if you can get the electronics and software to a decent point that would be great to use on a small machine

Edited 3 time(s). Last edit at 02/15/2014 07:39PM by cnc dick.

Do you have the file and settings you used to print that bottle opener? If you have the settings and the stl and such I can see how well I can print the same object, and we will have a decent comparison. There are lots of other factors. Ideally you could run your current machine with an adruino running marlin generating step and pulse to minimize any differences.

Have you considered running Marlin on a cheap $20 arduino board and letting your machine do all the temperature stuff external like you do now, just replace your parallel port's step/dir functionality with Marlin's?

It would be a good test I think unless there is something I am not thinking of that you are doing our of your PC for control. I think that comparison is worthy of a thread

It was a sample file in skeinforge slicer I don't think it really matters that we compare against the same print you want to be able to print something so it comes out decent no matter what it is I haven't really seen too many reprap prints that I would call real good. There are some but the majority to me look terrible like I say a lot of segmentation and rough finishes and corners that aren't square and have quite a radius . The electrical on my is quite large Nema 34 motors so to try to get the drive hooked up to some kind a ramps running individual wires and then changing the timing so that my drivers would operate the standard pulse is too short is a big deal for me. Basically I'm still wondering if it's just crappy design mechanicals or electronics and firmware. I like to build a small machine like I say I think I'll just wait and see if somebody like you build a good sturdy machine and see how it prints with the standard electronics

Edited 2 time(s). Last edit at 02/17/2014 06:27PM by cnc dick.

I guess that the bottle opener design is this one [www.thingiverse.com]. It doesn't look like a difficult object to print.

---

What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license

Still waiting on my Azteeg X3 Pro. shipping info I ordered it 5 working days ago... hopefully soon.

I have plenty of machining left to go still, but I modified a few parts and made drawings of some others.

Black delrin nuts are cut and ready to go on the Z axis, but I need to wait to put them on.

I found that the aluminum angle (2" x 2") was actually far from square... by over 1mm per 2". I figured it would be better than that, maybe my stock was damaged, but everything I checked was pretty bad. I found a nice aluminum chunk in the scrap bin that was already milled down nice and square, in an "L" shape.

I need to mill it to size, cut the slots, and drill the holes, but otherwise it should work just fine:





The fan is temporary, so I'm just going to bolt it on. I imagine I'll know more about how I want it after a while of messing around, and I'll hopefully print a custom shroud. I've read mixed reviews on blowing air across the part, that if you have a heated bed (which I will), you don't need/want it. I have room, but if anyone has opinions, please enlighten me

I like how this new end worked out... and I have the ability to move the hot end 9mm toward the back of the mount and fit another head in there. I'll just remove the press fit connector and put a set in the outer holes instead. I may nor may not actually go for dual heads ever.

Edited 2 time(s). Last edit at 02/17/2014 11:40PM by Tired2.

Oh! And my NEMA 17 motor came in for my extruder:



That is the difference between about 70 oz-in and 286 oz-in... NEMA17 VS NEMA23 frame.

I got that guy from QU-BD, who got it to me in about 2 days, I was impressed. Extruder is on the way, though I found out I need all sorts of parts for it, I thought it was complete when I ordered it :-/ Oh well. Anyone wants to donate a hobbed bolt let me know

Very nice build looks very good I personally don't like the remote extruder waiting to see how it prints. Looked through my parts collection yesterday I have some linear rails smaller than yours and plenty of 8020 various sizes and Nema 23's and a planetary gearbox Nema 17 laying around and a couple of high lead ballscrews and Acme screws and flange nuts I could use belts to I have some timing pulleys and belt stock laying around

Edited 1 time(s). Last edit at 02/18/2014 04:14PM by cnc dick.

Tired2 I've been looking at some electronics from a company called pibot supposed to work with repetier they have a Kit with drivers up to 4 1/2 Amp there are out of stock at the moment and there isn't any one out there with experience with these but it looks looks to be nice quality. Seems to be a fairly new company located in China obviously on their forum which is very small they have a picture of one print supposedly that they did and looks fantastic but this could be a fancy Photoshop I guess I'm a little skeptical

I'm not sure if this has made all the rounds yet, but for us extrusion builders...

[www.reddit.com]

$150 off if you have never ordered from misumi before, just pay shipping. No minimum.

I thought "I'll get some spare extrusion", and now I've got $300 worth of parts picked out that I have to weed down to a reasonable value.

Edited 1 time(s). Last edit at 02/21/2014 03:08PM by Tired2.

Quote
Tired2
I'm not sure if this has made all the rounds yet, but for us extrusion builders...

[www.reddit.com]

$150 off if you have never ordered from misumi before, just pay shipping. No minimum.

I thought "I'll get some spare extrusion", and now I've got $300 worth of parts that I have to weed down to a reasonable value.

What does this shipping to the US typically run?

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

It is actually pretty reasonable unless you get into freight stuff. I'm not sure what the extrusion length is that they will ship ground, you can probably get away with 4ft or so, but not 10.

The idler pulley mounting design is one of my biggest issues on this machine... I have a 12 tooth drive pulley and a 14 tooth idler, so I have to offset my idler toward the back of the machine a few mm to keep the belt tracking parallel with the X Axis (same on Y). All my pulleys were free though, so I didn't complain.

For just under $100 I can get 4 more pulleys, 2 x 14 tooth drive pulleys and 2x14 tooth idlers with the bearing style I need (countersunk). That will allow me to mount on a M6 Bolt and T-Nut and be able to slide the pulley for tensioning and all sorts of stuff. It was what I had planned initially, but I was floored by the cost of the pulleys, so I was going to use what I had.

This deal may let me get them, some gussets for the Z mast to really strengthen things up, some nice rubber feet, and some 608zz bearings for my extruder and future thrust bearing setup. I'll be ordering tomorrow or some time soon, and I'll report back what shipping comes out to.

I've just realized that with new pulleys, I'm no longer constrained to the belt type, which got me thinking, maybe GT2 or GT3 belt would be better, it has less backlash and is made for linear motion, as opposed to timing like XL belt is.

I also realized I need to do the speed/torque calculations as well to make sure I don't want to go smaller or bigger at this point. Going from a 14 tooth XL to a 24 tooth GT3 is a very small change in pulley diameter, they are both around 23mm... Pretty small. I can go bigger and get more speed, but I think maximizing torque is ideal. More calculations to follow.

Okay, maths done:

I made sure I was not maxing out the RPM of the motor, nor the maximum step generation capable by Marlin on a 328p chip (40kHz I believe).
This is with the smallest GT3 pulley I could get that had a matching sized idler appropriate for my mounting.

32 microstepping:


16 microstepping:


I don't think the torque math is needed, since I'm already on the smallest pulley I can really get away with. Going smaller would give me less top speed, but more torque, and conversely going to a bigger pulley would give me more speed, but would require more torque of the motors for each move.

So, my conclusion is that I can go minimum size on the pulley to maximize available torque, and at 1/32 microstepping on X and Y, I'll still have up to 450mm/s top speed, limited by the step/pulse generation.

That is about 3x what I need, so that is good, not to mention I doubt I can accelerate the mass I'll have up to 450mm/s in ~250mm of travel with the torque I have available.

If Anyone sees problems with my math, please let me know.

Edited 1 time(s). Last edit at 02/21/2014 04:52PM by Tired2.

Okay! Busy night for me.

The realization that I might be able to use different pulleys made for quite a bit of improvements on the machine. Fewer machined parts as well, which is great for me because I have to race home from work for the evening routine with the kiddo, and I don't have much time for before or after work machining in our shop right now.

So... changes:

Going for 3GT belts now instead of XL. They supposedly have less backlash, which will be good. Probably the XL would have been fine, but these should work better.

I've redesigned the X axis to use lower profile "A Type" pulleys since I am not limited to the freebie I had. This will let both shafts be shorter, getting the moment down closer to the shaft bearing on the stepper and closer to the mounting point for the idler.

Another Idea I had was to tap the side of the T-Nut with an M3 hole, and drill an M3 hole in the end cap. This will let me put a ~35mm bolt in and pull tension on the T-Nut before I tighten it down to tension the belt. That opened up the world of using a closed belt instead of my previous plan of mounting both ends at the bearing block / tool plate. Again, less machine work, which is nice. Maybe I can get away with a single peice of 30mm extrusion now instead of a machined block. Luckily I thought ahead enough to slot the motor mount plate in case I was able to work something like this out.





I finally gave some attention to the Y axis as well. This pulley change made it a lot easier. Unfortunately I could not find the pulley I needed in the "B Type" that extends out to clear, since I have a 1/4" shaft and selection is a bit limited on "Inch" parts with Misumi. I can just chuck it up in the lathe and ream out the last .35mm to go from a 6mm to 6.35mm shaft.





I've also added some gussets that will hopefully keep my Z masts a bit more square. I think I could measure about 0.5mm gap at the top when I squared it with a level.



Here is my shopping list, which comes in at about $161.67.

Part			Desc									Qty	Ea	Total
AGTFW24-GT3090-6	3GT 24 tooth double bearing 9mm belt 6mm shaft	                        2	$30.82	$61.64
GPA24GT3090-B-P6	3GT 24 tooth high torque drive, 9mm belt, 6mm bore - NEEDS DRILLED	1	$16.71	$16.71
GPA24GT3090-A-P6.35	3GT 24 tooth high torque drive, 9mm belt, 1/4" bore	                1	$14.84	$14.84
GBN11133GT-90	        3GT Belt , 9mm x 1117mm, X Axis closed	                                1	$13.78	$13.78
GBN8103GT-90	        3GT Belt , 9mm x 810mm, Y Axis closed	                                1	$11.40	$11.40
NFBG32-6-25-B	        bendy feet	                                                        4	$5.97	$23.88
HBLFSSW8	        z axis gussets, cast, curved	                                	2	$4.25	$8.50
SC608ZZ 	        skate bearing	                                                	7	$1.56	$10.92

Edited 2 time(s). Last edit at 02/22/2014 02:01AM by Tired2.

In other news, I've started my next YouTube video series on this printer.

Yes, for some reason I think that since I've read a bit and made half a printer I'm enough of an authority to make a video about it.... But seriously, I just try to make the videos I wish I saw when I first got into the hobby.

Evan's DIY 3D Printer / Mill - EP1 - Intro / Motivation

This series is my attempt to share some knowledge to those new to the hobby.

Future videos will be a review of the Azteeg X3 Pro, the Wade's extruder I just got, the qu-bd silicon heated bed, some software stuff, and hopefully a lot of cool application videos once this thing is up and running. Keep in mind the target audience is not necessarily people who have already built and run 3d printers before.

I've placed my Misumi order. $156.

I ended up only getting the pulleys, belts, and right angle brackets.

I realized at the very last minute that the belt I selected for the X axis was supposed to be a 1113mm, but was actually coming up as a 113mm. They did not have the belt length I needed in the 9mm belt width, so I had to increase the belt size to 15mm, which meant going to the next size up on pulleys. They are still 24 tooth 3GT, but they are taller. EDIT: They may have the 1113 after all, I'll hopefully be able to cancel my order tomorrow, switch it back, and go with 9mm since that makes my belt blocks fit better.

In other news, I received my Azteeg X3 Pro on Saturday, and I was blessed with enough time to get it to spin a motor, and make a new Youtube episode on it:

Evan's DIY 3D Printer / Mill - EP2 - Spinning Motors (Azteeg X3 Pro)

I hope the next ones are shorter, but no promises.

Edited 2 time(s). Last edit at 02/24/2014 11:58PM by Tired2.

Hey! So happy we could help with the belts. Sorry it was not a perfect solution, but hey.. $150 of free components.

Yes, I got it all figured out. There was a mistake on the Misumi site with the 1113mm belt. It is all worked out now, I'm back to 9mm belt width, which mounts nicely on all my belt blocks, much less machining.

I've started wiring the motors. I wired a big triple stack last night and my SD8825's got blazing hot, so I installed a temporary fan and all is good. Don't try to run them at 2.5A with no fan... ouch. (They are thermally protected, but my fingers are not.)

Hello tired2 how is the build going do you seem to have enough power with the small drivers and 12 Volt on those large motors. Keep us up to date like I said I'm in the the middle of possibly building a smaller portable machine. Ordered a few odds and ends and have been poking through my supply of pieces and parts.

Hi CNC_Dick.

I got some machining done this evening, but I still have some more to go. I got the groove mount adapter part made, which is nice to have behind me since that was the most difficult milling job. I think I got it close enough.

My misumi order was placed last week on Tuesday, and I just got a confirmation of shipment, which seems a bit long. I'm not sure if all my items were stocked though. So, those will be here friday, so hopefully I'll get the rest of the machining done before then so I can get the thing moving under its own power.

Did you check out EP2 of my youtube videos? I don't think I showed running the big motor, which I have done since that video.

It gets really damn hot at 2.5A. I've put a huge fan over top, and they still get warm on 2.5A. I turned them down to about 1.75 or 2.0A for now to see how things go. No need to cook the drives if I don't end up needing the power. They do have overtemp protection though, so I doubt I'll damage them. It will need an enclosure with a fan or two. I'll probably go for 3x60mm fans or so once I print a case for it all. Maybe I'll eventually work out temperature based PID fan control so I can keep the noise down.

I'll be running 24v I'm sure, and I have done so with an unloaded motor as well. No difference that I could tell with no load.

The motors do make quite a whaling noise on certain microstepping settings. I'm hoping the oscillation will stop a bit when they have a load, but I'm not really sure what to do about that just yet. Strangely, the NEMA17 I have is near silent when stopped, yet the big motors make some noise at certain positions.

Update...

I think I'm switching to the E3D all metal print head. I'll be ordering it next week unless anyone has any advice.

I put it in the model. The mount did not change, so it will still accept a J Head.



I finally found some time on the mill after work yesterday...



I also got the tool plate drilled and countersunk, it is in the painting phase. Rustoleum textured black is my all time favorite

3 more mounts to machine and I can get all the pulleys / belts on and jog it around, hopefully this weekend.