The 1/4" aluminum is MIC6 cast aluminum, and surfaced flat on both sides. It's 1/4" thick because you can't get MIC6 any thinner than that. It's overly large because most of the extrusions are pre-cut at 1000mm lengths. I've chosen to limit the print area to 600x600mm (24"x24"), as to go over 24x24 with a PEI sheet increases the costs dramatically.
I got the base assembly put together, as well as the vertical rails for the Z-axis. I was a little disappointed to find out that the "1000mm" V-Slot I received wasn't accurately cut to 1000mm as the first batch I had ordered a year or so ago. All of the pieces I received were between 955mm and 957mm in length, compared to a previous order where they're all 1000mm on the dot. Not a problem - I can take the V-Slot I used for the Z-Axis rails, and use them for the top frame - they're all the same length at 1000mm. It's more important that the pieces used for the frame match - otherwise they won't form a good square. Also, I assembled the first carriage for one of the Z-Axis rails. It's sitting on the rail, defying gravity... but moves freely at a touch. If there's any play, I can't feel it.
Sorry, my numbers were off. I meant 995 to 997mm, 5mm to 3mm off the 1000mm length. I also need to get a better yardstick (meter stick), I have two that don't agree with each other.
My apologies - I re-measured with an accurate yardstick, and all the V-Slot I received were all 1000mm +/- 1mm. A few of the V-Slot pieces I received during a previous order were actually closer to 1005mm, and that threw me off. Before, I was using a flexible measuring tape that had probably stretched.
Thinking about adding some additional corner bracing to counteract any racking of the frame. I'll probably make the braces a little shorter, and butt them up against the corner plates.
Seems the price of cast aluminum dropped recently, as I was able to order the 1/4" aluminum tooling plate for about $30 less than what I had anticipated. The PEI print surface is 24"x24" (still to be ordered), and the aluminum heat bed underneath is 26" x 26". Going to drill a series of 5mm countersunk holes for mounting. Plus, I'm adding a layer of cork board between the aluminum and the V-Slot rails, for insulation - with the heater sandwiched in between.
Used one of those M5 drill taps to tap the ends of the V-Slot, and promptly snapped it off inside one end... So now I'm using a conventional M5 0.8 tap, and doing them by hand. You get more feedback by hand, and less likely to break the tap.
I was thinking of a similar thing you are doing with the cork for insulation. Are you worried at all that this may allow the bed to become uneven at all?
Nah, the bed will be supported around the edges by the V-Slot, and the cork is fairly uniform in thickness. The tooling plate being cast aluminum and surfaced is what makes it flat. That, and the cork will prevent the heat from transferring to the V-Slot.
Designed up some belt clamps for the CoreXY belt ends, for 3GT 9mm wide belt. Just some 3mm aluminum bars with 5mm holes and some grooves cut in 3mm spacing to match the belt. Also on top you can see the plate with cutouts for the hot-end wiring connectors - 12 position Molex Mini-Fit Jr connectors: http://www.molexkits.com/76650-0079 The idea is that the wires will come out of the cable chain on top, and then go down through the plate. The plug housing latches into the cutouts in the top plate: http://www.molex.com/molex/products...0039012121_CRIMP_HOUSINGS.xml&channel=Product The connectors are large, but cheap and easy to assemble. 12 pins: 4 Stepper Motor pins. 2 Heater Cartridge pins. 2 PT100 Sensor pins. 2 Hot End Fan pins. 2 Print Cooling Fan pins.
Redesigned the connector plates and the mounting system for the cable chain end. The cable chain end is held up via 2 x 40mm posts, a flat bar, and a center 20mm post. The plates for the connectors will be bolted above the V-Slot face, allowing for an easier routing of the cables.
You don't quite realize how BIG the build area is going to be, until you get the 26" x 26" aluminum plate... It's HUGE. The surface is protected by the plastic sheeting you see here... underneath it's ground flat, to a near-mirror finish.
I noticed in your parts list that you are using all smooth idler pulleys. Aren't you worried about belt life and mushing the teeth? It's generally advised to use toothed idlers for the tooth side of the belts. Or are you twisting your belts so the smooth side is always against the pulleys?
I'm using smooth idler wheels because OpenBuilds doesn't offer toothed wheels in GT3, unless I want to use aluminum pulleys. I suppose I could print GT3 toothed wheels once I get up and running...
Gotcha. I was actually looking at it because I was hoping you found some GT3 pulley's since I couldn't find any either and was thinking of using 3GT belt. Between that and that (according to some things I read) 3GT has more backlash than 2GT (though I'm guessing with enough belt tension it may be negligible), I'm probably going to stick with 2GT-9mm despite having a bit longer belts than you. Still trying to decide. PS your aluminum plate looks huge...and awesome! I bet you can't wait to get that thing mounted with the heater
It wouldn't be too difficult to generate an OpenScad-based idler pulley/wheel with a 3GT/GT3 profile: Parametric pulley - lots of tooth profiles by droftarts Kind of the same as the T5 model shown, but with the 3GT profile... Print with PET or some PC/ABS mix... And yes, the heat bed plate is GINORMOUS... the silicone heater is 24"x24", and is rated for 2000W @ 110V. You can see all the heater wires traced through the silicone... they cover nearly the entire mat, and are pretty consistent... so it should heat well.
An enclosure so big, it looks like a gazebo... Sketched out a shallow pyramid top, to be made out of 1/4" plywood. The side panels are 48"x48", and the back panel is 48" tall and somewhat narrower than 48" across. I will probably put some 24x48 acrylic panels on the front as doors.
I'm reconsidering this option now... 12mm lead screws, 3mm lead, single start, 1500mm length. The lead nuts with flanges aren't available as anti-backlash, I wonder how much of a difference that would make... 12mm (3mm Lead) x 1500mm Trapezoidal Lead Screws - A4 Stainless Steel: Accu.co.uk: Precision Lead Screws & Nuts 12mm Lead Screw Diameter (3mm Lead) Flanged Trapezoidal Lead Screw Nuts - Bronze: Accu.co.uk: Precision Lead Screws & Nuts
You shouldn't need anti backlash nuts for z axis. I'm not planning on using them either. Just don't fully counterbalance the bed, leave a pound so it loads downward. Also I fully support your decision, for the 0 amount my word matters haha!
I'm also looking at 16mm ball screws, 1500mm long. eBay specials priced around $115 each, but that's with ball bearing nuts, machined ends, and ball bearing end blocks. 5mm lead, so a little higher than 2 but much less than 8. I'm kind of leaning towards the ball screws, actually.
Yea it does. If budget isn't a big deal for you, maybe go with 4040 uprights. I'm on a much tighter budget than you do I'm going with 2040 and am pretty sure it will still work well, and if not I'll grab some hunks of metal off Craigslist or a local junkyard and brace it with that. Still think it will be fine tho. Also, while I don't know much about the difference about using a ball screw, but 16mm will definitely be stronger than 8. I'm going with 12 because I can easily find cheap pillow blocks for them. There are actually arguments for going with skinnier screws and letting them flex but I'm not a fan of that philosophy. I'm planning to run my screws on a 1:3ish ratio to get even more, probably unnecessary, precision.
Could probably use 40x40 uprights. I'd have to figure out an alternative Z axis rail system with the larger gantry plates, or just go with bolt on linear ball rails. I'll noodle on it for a while...
I naturally couldn't sleep because I was thinking... Some 20mm linear rail and carriage matches up well with 16mm ballscrews 1500mm long. I was thankfully able to find some Sketchup models of both, and extended to suit. I'm envisioning some simple 3mm aluminum plates to bolt the carriage and ballscrew blocks to the V-Slot, easy enough to fabricate.
Investigating ballscrews and rails on AliExpress has gotten me quotes ranging from $520 to $840 for 4 sets of rails and 4 sets of ballscrews in 1500mm lengths. Also rather than a counterbalance system, I'm looking into stepper motors with brakes: free shipping 4-lead Nema 17 Stepper Motor nema 17 step motor with brake CNC Laser and 3D printer free shipping 4 lead Nema 17 Stepper Motor nema 17 step motor with brake CNC Laser and 3D printer-in Stepper Motor from Home Improvement on Aliexpress.com | Alibaba Group The part on the top of the stepper is the brake - apply power, and the brake releases. No power, and the brake keeps the stepper from turning. I can probably wire the brake up to the 24v power, or a mosfet control on the Duet.
I feel like lead screws, esp 4, wouldn't allow the bed to fall at all like belt driven z axis would do. Is this a problem if ball screws?Another advantage of that counterweight is to keep less weight off the screws, to keep them from wanting to bow and impart forces into the frame, though beefy 16mm screws may be enough to prevent this, not sure. Downside is counterweights are ugly, but I'm still planning on going that route. Brakes won't hurt though and you could give it a try. *EDIT* looked it up and ball screws do need brakes... ive never used them before. Learn something new every day.
Yeah, ball screws are anti-backlash to begin with, and have low resistance between the screw and the nut.. The ball bearings in the nut actually roll down the threads like a corkscrew. Based on the weight I'm expecting, the calculations indicate that the stepper motors won't need too much torque to lift the plate - so NEMA 17 steppers will do. There's one or two NEMA 17 Brake Steppers on AliExpress, and the Alibaba site has more available...
Appreciate the info, I have heard of melting HDPE into acme nuts. And I've seen others use belts to drive multiple Z screws - but that's not the method I want to use. I want 4 independent Z motors, on 4 independent stepper drivers - that way the electronics can level the bed by itself before each print. The Duet + Duex5 electronics already has the capability of creating a Z-homing macro that can do that - home Z in center, then probe the 4 corners one at a time, leveling each screw separately.