I read his description about this again & think I understand how he is doing it now.

I have used this little belt tensioner before & it works pretty good, although I would consider this a bandaid fix.
MicroWave – YouMagine.com

 

this may be a dumb question but I wanna know
how do you people level your frames?

 

Would changing the Vertex Idler Cap to be adjustable instead of using the bolts to pull each vertex, work just as well? Here is a photo of what I mean along with the openscad file. I would probably have to change the Top Vertex part a little also. The raised edge could point to the bolt head instead of against the plastic to make the part more rigid, or leave that out entirely. You could also cut this out of aluminum instead.

 

tbh, the way it is designed works really well. I tried a spring tensioner at the start but when I zeroed the arms they would hit the top due to the spring opening on deceleration.
now I just set up the top frame with the verticals poking out the top 2mm (measured with an allan key) and then put all three belts on, I try to make sure I have the same tension on them when I hook the teeth in, then adjust the tension as normal.

When I measured the bed, my corrections where .3, .6 -.7. (I performed a manual bed level using a piece of paper)
With a Smoothie / delta, I don't need to touch again until I remove a belt.

Simple works.

 

How do you like the Smoothie board? Is there anything else you need to buy electronics wise for the smoothie? What kind of LCD would you need to hook up to this board? I found this smoothieboard clone on ebay for $69 and free shipping that seems to have the 5 steppers connected. I wonder how good the electronics parts on it since it is less than half the cost of real smoothie.

Smoothieware compatible Board, MKS SBASE 32 BIT ARM , RepRap,CNC,Laser, New v1.1

 

I like the Smoothieboard but have never used anything else to compare with.
I have seen mention before that a lot of the other controllers will not allow a delta to be ran at full speed due to lack of computing power.

I did look at the clone version but there was problems with temperature mentioned due to noise from the 5v regulator so I decided to go with the real one.

just stick 24v into it and wire it up. adjust the config file and off it goes.

I am using IP connectivity and was having some problems with the host software, changed to serial(usb) and works great now.

 

Is there a good source for longer arms or the parts to make longer arms? The longest I have been able to find are 17.5" or 444mm from tridprinting.com. I am thinking about making the base with 500mm lengths & by using my calculator or the simulator, 515mm lengths looks about right. If I used 440mm, the design angle comes out around 54, instead of around the 60. I am planning to use 2060's for the vertical, mostly because I already have one of them, so would be less to buy. I did a test print of a vertex & carriage I modified to fit a 2060 & seems like it will work. Here is a screen capture of the calcs from the simulator on this. Pierre has more detailed calculations in his simulator than my calculator does.

I have also noticed that most people with this design are using a carriage & another backplate on the other side of the extrusion. Does that make this design a lot stronger, or could I just use a plastic carriage on the inside?

 

I used carbon fiber tube from ebay, I have not seen many longer than 500mm though.
What country are you in David?

 

I am in the U.S. I may end up making it smaller as there are probably other problems associated with making the arms longer.

I was also wondering if it is necessary to bolt thru to the ends on the horizontal extrusions. Seems like if I have 4 bolts into the sides with t-nuts that would make it rigid. I was trying to avoid tapping the ends. If I do have to tap the ends, does anyone have a good recommendation for a tap? I have not tapped holes before. I saw some this tap jig /holder for v-slot that seems like it would make tapping easier. Tap Jig / Holder for Aluminum Extrusion 20mm / any width Misumi, OpenBuilds v-slot, MakerSlide compatible by Protobuilds

 

I tried that one awhile ago. Was worthless.

Edit:
I see he added one with two screws on each side. That might work better than the original one hole version that kept slipping.

 

Here is another version of a M5 tap jig that was done back in 2012.
2040 Aluminum Extrusion M5 Tapping Guide by SystemsGuy
This one may have the same problem you mentioned.

 

Tapping the holes in the end should not be a problem, aluminium is easy to tap.
The secret is to use the correct tap, eg taper and HSS not carbon.

Buy a 5mm TAPER tap, use HSS not carbon.
Chamfer the start of the hole with a larger drill (say 10mm) to take of the burrs.
The holes are already 4.2mm.

Use a battery drill with a screw clutch, set on the lowest torque.
Use lube, crc or oil, dont need much.

Then just make like you are going to drill, once you feel a bit of weight on the tap or the clutch slips, back it out half a turn and go again.
I would stop and backup about 3 to 4 times for a hole if going in 15mm.
back it all the way out, clean it with a brush and do the next.

 

Spending the day tuning my printer, having problems with overhangs.
Reading on the internet suggest I need a cooling fan for PLA, does this sound right?

The attached photo is a pla temperature test, the temp was dropped 5c per segment, does not seem to make a big difference.
It was printed upright.

 

This is my Wolfstock/CherryPi/DeltaSix design that I combined the best of each designs. I started with a Wolfstock because I had some Makerslide. Then I like the spring loaded end effector of the CherryPi. I also changed how the effector does z-probing by using 3 limit switches in the base of the effector and loading it with springs. The only problem I have is the original base brackets of the Wolfstock allows the vertical tower to be not stable. This is why I am switching to the DetlaSix base design because I think it can be more rigid by doubling the height of the bracket and the way the bracket secures the horizontal beams with screw on the backside of the bracket. Also considering to print the brackets in PETG.


The Spring loaded design is using carbon fiber rods from Goodwinds and Ball ends from ebay
16pcs 10mm M4 Threaded Steel Ball Rod Ends For Kossel 3Dprinter Magnetic Joints
Ball ends where screwed to a 4mm thread rod and glued to the carbon rods. The springs are connected to 30lb fishing steel leaders.

My horizontal is 500mm Misumi 20x40 and my vertical are 1000mm Makerslide

To give credit where is credit due here some of the designs I combined
Cherry Pi III by AndyCart
Boo's Parameterised FSR Groove Effector for Delta Printers V3 by boofly
Kossel Mini spring-string effector and carriage – YouMagine.com
Githubiverse: Wolfstock a Mini Kossel derivative by Jim Morris
Hot End Effector Setup for Cherry Pi III with 10 mm ball bearings (Or 8 mm balls) by maso27

Still a work in progress let be know your thoughts

 

I really like the drilling with 3/16" idea. I will have to try that on one piece to see how it works. Did you use your tap jig to keep your drill bit perpendicular or does having the hole already at 4.2mm help to keep it centered? I just ordered some eSun blue PETG from Best quality 3D printer filament from eSUN to print these parts. I still have a couple of more test prints to do on the carriage & Vertex extrusions holes for the 2060 size extrusions. After 3 test prints, think I have the extrusion hole size fitting nicely with my PLA prints. I am going to do one more test print of the smaller top vertex in PLA before printing them in PETG next week. Here is what my current design for the bottom 2x 2020 horizontal & 1x 2060 vertical looks like. I made the holes where the endstop wires come out to be a little bigger than the original design. Is anyone bringing the wires down thru the middle of the extrusion or is everyone coming down one of the inside V-slot grooves? I am looking at redesigning this a little be able to get the wire down the middle from the top of the upper vertex & thru the bottom of base vertex. I would probably have to create something like a partial threshold to rest the verticals on at the bottom to get access to that opening.

Here is what my current bottom vertex looks like. I probably only need 4 holes for the outer verticals instead of the 6 I have.

 

Kyo thanks but I already started modeling the bracket. See picture below of old Wolfstock to new DeltaSix version. How big is your printer with Makerslide? Also is it rigid?

 

Many use Carbon Arrow shafts for rods with magnetic ball ends attached at each end. My CF shafts are 4.5mm X 825mm. Do a google search for magnetic effector.

Edit: I see that it was covered in post #464 above. Sorry for any duplication.

Thurmond

 

Progress!

 

I recently built the Mostly printed CNC router that uses 3/4" electrical conduit (23.5mmOD) for the structure and the movement. That pipe is about $3.50 per 10' length. Yesterday I started thinking about how to use electrical conduit pipe in a delta design. If this works, I would only need to use aluminum extrusion for the verticals. The parts I would need to redesign would be vertex & a way to mount the build plate on the conduit. This design would only use the EMT for Structure and would make the printer heavier.

Here is a screen capture of my thoughts on this. I probably need to make the vertex extend into the horizontal pipesThe 2nd photo shows a 2mm gap cut half way thru the vertex that would have a place above each side to put a small bolt in to tighten the plastic to the conduit pipe. Does anyone have some thoughts on this alteration?

 

Has anyone ever used the garden stake idea for the diagonals like the Griffin Delta build uses?

 

I took out the vertical conduits as it did not seem to need them since the original design did just fine with a 2040. I am not sure I will use this alternative, since I have most of the extrusions I need already for this build.

 

Here is what I came up with for the vertex using the electrical Conduit pipe for the horizontals. The conduit pipe goes in 1" into the Plastic. I did a test print of just the portion where the conduit goes into the plastic & it is a nice tight fit, then tighter when you put the small bolt on it. This uses #6-32x3/4" or 1" long bolts to tighten the Conduit Pipe in.

This would still need a way to mount the glass or heat bed to the pipes. I have a pipe drill jig on thingiverse that might help with adding a part there if you drill thru both sides of the pipe. Remix of Parametric Drill Guide for Z-axis Pipes Mostly Printed CNC by GeoDave

I put slotted holes in the plastic to use for belt tensioning instead of the vertical M3 bolts. I included the scad files in case anyone wants to try this idea. I will probably print at least 1 vertex out to see how well it actually does work.

 

I was playing with cutting a carbon fiber arrow I bought from WalMart using my rotary tool today. Since it was so easy, I made a parametric plastic jig to cut all 6 at the same time. I am still debating on the size to make it, but leaning towards using 500mm horizontals & 1000mm 2060 Verticals. That would give me an arm length in the 515mm range with a working dia. around 455mmm & working height 322 to 387mm. Here is a link to that cutting jig on thingiverse. Parametric Carbon Fiber Arrow Cutting Jig for 3d Printer Delta Arms by GeoDave

 

I am ordering my parts for the arms & planning to use magnet connections. I ordered 6x10mm magnets from ebay. 6x10 mm Super Strong Cylinder Round Disc Magnets N35 Rare Earth Neodymium 50PCS

I am also looking at these 10mm M4 threaded steel balls. https://www.jpson.com/product/10mm-...for-reprap-delta-kossel-3d-printer_p1226.html They are $5 for 16 of them & free shipping. These look like a lot better price than what I saw mentioned earlier in this thread. Do these look like they would be good to use?

 

I'm using these magnets: K&J Magnetics: D58-N52
They hold up to 7.66lbs, you might find that your N35's won't be strong enough.

If it puts it into perspective for you, my Rostock has been running on N52's that are weaker than the ones I linked above, and I have problems with my effector popping off with the slightest bump into a raised corner or stray chunk of support. My Delta-six build will have stronger magnets to counter this a bit. The fact that it pops off has saved me in the past as a designed failure point so nothing else breaks (slamming the nozzle into the glass), but it has also popped off when it shouldn't and ruined prints.

The steel balls seem pretty good, these are the ones I'm using on both my Rostock and my not quite put together Delta-six: 10mm Diameter Steel Balls | Deltaprintr

 

Those magnets seem kind of pricey. I see some 4mmx8mm N52's on ebay in 50 quantity that are only $8.45. I will wait & see how the N35's work. I should have looked at the strengths of those closer before ordering the N35's. Thanks for the info.

Maybe I could put 3 of those magnets together if they cooperate with each other & it does not add too much weight. It might actually give the magnets more surface area to grip the 10mm ball. That would use 36 magnets instead of 12.

Here is a screen shot of my thoughts on this idea. I initially tried 5 degs & 15 degs rotation for the magnets, but looks like 30 degs. touches the 10mm ball more. Might be some clearance problems here & I would obviously have to make a special plastic fitting for this. The magnets are shown in yellow, carbon arrow shaft is gray & the 10mm ball is silver.