My thanks to Dave! David (Escher 3d) Crocker and the folks at Duet for making a great product. Today (and last evening) I calibrated my Very Large Delta build. It has a theoretical printable radius of 480mm and a printable height of 240mm or so with the current 590mm arms.

Some of the fun I ran into during the build.

1. I designed ordered some machined "cheese plates" with holes for building a printer or any radius. The holes are big enough for M5 bolts which tap into 4040 at four places. I ordered them with 5.5mm holes. It would have been better to go smaller, but 5.5 is the standard hole for an M5 bolt. The errors seem to be magnified in a large build.

2. When I installed the linear rails there was a little bit of an angle on one of them resting against the extrusion. I am using 4040 extrusions with two channels mounted off-centered on the cheese plate, the linear rail is on the center line. I shimmed underneath one side of one MGN slider to compensate.

3. I imported a Seemecnc slider carriage from open source for 3d printing, I am using Seemecnc ball joints and a 713 maker Duet Smart Effector Adapter.
   I reverse engineered a ball cup holder for 3d printing for construction of custom arms. I used stock lengths of 500mm 10od 8id rods and the result was a very lengthy 590mm arm. I have 400mm carbon fiber tube on order for a more reasonable arm size, which should give me more height. It would be better to have injection molded arm ends and carriages. There is more friction than I want to have in the ball joints. Perhaps SLA ends would be preferable. I was able to get a good click in after 20 iterations of the arm end. It would be great to have a wider effector.

4. I have provisioned a 400mm square glass heated bed. I just completed a bed calibration and a map. The bed is spring loaded and the map indicated that I had not leveled it. This is a challenge of a big build. The machine has leveling feet and was leveled with a bubble level. The bed was also checked for level. The bed map was more precise and thus said otherwise. I am very happy with the calibration. I can push the effector outside the radius and stuff is wonky out of range.

5. Some small issues encountered and fixed. The bed I ordered was cheap at $120 with a silicon pad, glass, springs, and SRR. Unfortunately, I don't trust the "fotek" SRR which shipped with the bed so I purchased a UL listed Crydom SRR which was about $60. I can accommodate a 500mm bed when/if they get more available. I had some problems with Chinese faux Makerbot end stops so I used some good contact switches. I did not want the mass of a flying extruder so I was going to hang the extruder in the middle, this had the effect of having a pendulum in the middle of the printer. I am now hanging the extruder on one of the three carbon fiber "stiffening struts" that firm up the printer. The effector support can rock back and forth without transferring energy to the printer. The more rigid the better, perhaps c Channel might be more rigid than 4040. The struts are a great solution.

6. The effector bubble level is "in the bullseye zone" over the whole build; with the bubble perfectly center in most positions.. My one remaining question is whether or not this should be tuned in software via bed.g 8-factor calibration?

7. the build was not as expensive as anticipated. 1) I had allotted of spare parts including the rails 2) the frame was $250 3) cheese plates $400, 4) duet and effector, 5) Meanwell power supply (spent a day checking that all elements of the build were grounded, 6) other bed $120, SSR $60 wire, bolts carbon tubes. Much bigger than this the rails would be prohibitively expensive and, and much taller it would need C channel, struts, or more rigidity.

My two Seemecnc printers are running effector mounted bond-tech extruders. The results are phenomenal with that configuration. I would not even try to put that kind of mass on this large build. Also, I am planning to use a big .8 nozzle, I can go to a .6 if need be.

@Phaedrux Thanks

BTW I ditched the lead screws and am now using belts. I am using stuff in stock so two standard 6mm GT2 belts are being used. I have X and U synced up. Looks like I will have 1.2 meters x .5 meters printable area; with a z height of about 250mm.

Any ideas for the "final ideal" belt?

I would like to go with a 1mm or 1.2mm nozzle. Any ideas in terms of hot ends? Do I want to be using 1.75 or 2.85.. I have some bondtech 1.75mm in stock.

I have a Triangle Labs piezio board and sensor in stock I would like to try. The sensor sits on the hot end nozzle. The board has ao do and and Vcc pins. If I use digital output would I use an endstop header? As opposed to the z probe connector?

@Phaedrux
I took a pass on the E3d corexy because the build volume was limited and I don't have a use case for tool changing.
I am going to have a tiny amount of Z, a large nozzle and a massive build volume xy

In my case It might be more effective to lift the tool than it would be to lift the platform.

Mostly the design is driven by some of the parts I had on hand for a CNC build. I was going to build a CNC but decided to buy a Shapeoko. I have done a few good projects with the Shapeoko (designed under cabinet lights and tool storage shelves) but I am using 3d printers more.

Have sequential prints (one after another) or even think about concurrent prints. (two at a time)

For me the build is a month or two of fun. The printer will have three times the xy build volume as my very big delta;

https://www.thingiverse.com/thing:3416307

@Phaedrux

If I go with a belt I am going to have to have the stepper motor ride the x axis gantry; have a the belt lay flat and draw up into the stepper motor around idlers. If I go with a ball screw I will have a more direct way to drive. I might be able to have the belt loop around so flat ends oppose.

I will use a very big nozzle and a piezio leveling system.

The last printer I built was a very big delta with 400mm printing radius. I had large plates machined. The first task of the build was to construct carbon fiber arms.

I am going with a Cartesian so I can do 1000mm long by 400mm wide prints. It looks like a much easier build. The real task is going to be managing the weight on the gantry.

Thanks for the response. I was going to build a CNC but instead I am going to build another large 3d printer. The printer will have 1200mm x, 300mm y and 150mm z.. I am not quite sure if I want to use a belt for the X or a ball screw.

I reviewed the documentation and it looks like it might squeak by

I currently have this in stock;

Can I hook this up to the duet 2 directly
Nema 23 Stepper Motor 3.0 A 1.3 Nm 184 oz.in 76mm Body w/ 1m Cable for 3D Printer/CNC
Or would I use an interface board of some kind?

Alternately the following stepper might be more appropriate for a direct connect?
1 x Nema 23 Stepper Motor - 2.5 A, 1.3 N.m, 184 oz.in

If the board supports up to 2.4amps could either work if I use M906 and set the motor current below 2amps. M906 X2000

I am deciding whether or not I want to use a screw or belts on the X axis; or if I want a single large motor or two Nema 17s

My thanks to Dave! David (Escher 3d) Crocker and the folks at Duet for making a great product. Today (and last evening) I calibrated my Very Large Delta build. It has a theoretical printable radius of 480mm and a printable height of 240mm or so with the current 590mm arms.

Some of the fun I ran into during the build.

1. I designed ordered some machined "cheese plates" with holes for building a printer or any radius. The holes are big enough for M5 bolts which tap into 4040 at four places. I ordered them with 5.5mm holes. It would have been better to go smaller, but 5.5 is the standard hole for an M5 bolt. The errors seem to be magnified in a large build.

2. When I installed the linear rails there was a little bit of an angle on one of them resting against the extrusion. I am using 4040 extrusions with two channels mounted off-centered on the cheese plate, the linear rail is on the center line. I shimmed underneath one side of one MGN slider to compensate.

3. I imported a Seemecnc slider carriage from open source for 3d printing, I am using Seemecnc ball joints and a 713 maker Duet Smart Effector Adapter.
   I reverse engineered a ball cup holder for 3d printing for construction of custom arms. I used stock lengths of 500mm 10od 8id rods and the result was a very lengthy 590mm arm. I have 400mm carbon fiber tube on order for a more reasonable arm size, which should give me more height. It would be better to have injection molded arm ends and carriages. There is more friction than I want to have in the ball joints. Perhaps SLA ends would be preferable. I was able to get a good click in after 20 iterations of the arm end. It would be great to have a wider effector.

4. I have provisioned a 400mm square glass heated bed. I just completed a bed calibration and a map. The bed is spring loaded and the map indicated that I had not leveled it. This is a challenge of a big build. The machine has leveling feet and was leveled with a bubble level. The bed was also checked for level. The bed map was more precise and thus said otherwise. I am very happy with the calibration. I can push the effector outside the radius and stuff is wonky out of range.

5. Some small issues encountered and fixed. The bed I ordered was cheap at $120 with a silicon pad, glass, springs, and SRR. Unfortunately, I don't trust the "fotek" SRR which shipped with the bed so I purchased a UL listed Crydom SRR which was about $60. I can accommodate a 500mm bed when/if they get more available. I had some problems with Chinese faux Makerbot end stops so I used some good contact switches. I did not want the mass of a flying extruder so I was going to hang the extruder in the middle, this had the effect of having a pendulum in the middle of the printer. I am now hanging the extruder on one of the three carbon fiber "stiffening struts" that firm up the printer. The effector support can rock back and forth without transferring energy to the printer. The more rigid the better, perhaps c Channel might be more rigid than 4040. The struts are a great solution.

6. The effector bubble level is "in the bullseye zone" over the whole build; with the bubble perfectly center in most positions.. My one remaining question is whether or not this should be tuned in software via bed.g 8-factor calibration?

7. the build was not as expensive as anticipated. 1) I had allotted of spare parts including the rails 2) the frame was $250 3) cheese plates $400, 4) duet and effector, 5) Meanwell power supply (spent a day checking that all elements of the build were grounded, 6) other bed $120, SSR $60 wire, bolts carbon tubes. Much bigger than this the rails would be prohibitively expensive and, and much taller it would need C channel, struts, or more rigidity.

My two Seemecnc printers are running effector mounted bond-tech extruders. The results are phenomenal with that configuration. I would not even try to put that kind of mass on this large build. Also, I am planning to use a big .8 nozzle, I can go to a .6 if need be.

@alexander-mundy

The bed is "mostly" an afterthought at this time because I don't know what that the final geometry of the printer will be. I want to get a rigid platform first, then get the arms on, check if everything is level. I have two sets of arms in development.

(Building the printer is more fun to me that 3d printing. Totally fun! trying stuff.)

Once the geometry is solidified I will do initial testing w/o a heater.

I am considering a bunch of options - I don't know the final diameter of the printer
---(thinking about) separating the heater from the printer entirely with a keenovo solution
-- also thinking about http://www.zyltech.com/zyltech-3mm-400x400-heated-bed-kit/

--The Wago 221's are rated to 20 amps.
--I am not going to run a DC heated bed through them (low voltage is higher amps)
--I will run the heated bed through an SRR

@monster-delta

I checked out your monster delta.. WOW..
I plan to go wide not tall. Once I get wide where I want it I can then decide if I want to go tall.

It would be great if you could post more pics!

I see you have a hex design?

VBD (Very Big Delta) Questions and Request for input.

1. I designed a Delta Cheese Plate - it's going to let me adjust the delta radius based on performance. I can move the extrusions from 500 inward as needed. The delta cheese plate is rock solid with 40x40mm extrusions.

2. I wired up my power supply, and I noticed that DC is not grounded. Should I hook up negative DC to Green ground (which is mains ground)? I have the DC cables terminated in Wago connectors.. Should I jump ground from my negative wago to green?