The Prototyper: 4x8, 8020, cncrouterparts

[curriergroh]

This thread will be essential for the success of this machine, so thanks for reading and commenting.

The purpose of the machine is to facilitate rapid experimentation of full-scale architectural and sculptural ideas. It will be used for cutting soft sheet material: wood and plastic, and will be able to cut full 4x8 sheets fully beyond their edges. The machine will also be used to make molds and jigs to assist in other fabrication processes so a z range of 6 in. +/- is planned.

Research has been underway for most of the past year, however, I just recently started looking more closely at the bolt-together approach provided by the design efforts of cncrouterparts and finelineautomation.

My planned work flow using the machine will be Rhino to MadCam to Mach3 to electronics which aren't selected yet because I don't have weight or desired cut speed figured out.

I am designing the machine currently in Rhino using the supplied 3d models from cncrouterparts and 8020. I'm willing to convert and share pieces of model files if others request them. Currently looking for a PC892 model.

Ok so enough intro here are some renderings.


The High Z Plate allows for the use of the extended linear carriages, which I am using for extra stiffness and possibly longer travel down the road if mold making becomes more important.


The Gantry Interface Plate uses the extended linear carriages as well and rides on a 2.5 in rail. The cantilevered portion lets the gantry carriages clear the x-axis carriages, which leads to a tighter overall width. Is the .25 in steel plate (red) enough material to support a gantry this size? The piece of 3060 weighs 45 lb. and the 7” y-axis rail weighs 30 lb. So add in the z-axis and all the hardware and I'm estimating a gantry weight of about 150 lb. for now.


I'm using a piece of 1545 for the x-axis connected to the 2.5 in. rail. If I were spanning the full 111” without intermediate supports and did not take into consideration the steel rail acting as a flange the 1545 would resist deflection better than 1530 or 3030. But since it will be supported intermediately and also has a flange the choice, I think, comes down to other design considerations like attachment surface and compatibility with other hardware (gantry interface plate won't work with 3030). I checked with ahren at cncrouter parts and the R&P will clear the 1545 with the rack mounted in the center slot.

Use of the High Z Plate also makes the rail and extrusion assembly the moving piece. This part of the design is least clear to me, but it does seem to make a compact assembly. The parts that I do have modeled for the screw are lining up perfectly. Mostly the vertical arrangement and spacing is what I don't have a grasp of. Oh and I am looking for an accurate model of the router mount and a PC 892 if anyone has one to share.


The bed pictured here is just a quick attempt. I would like to design a table and bed that avoids being overbuilt. I think this is an opportunity for cost savings where 1530 is probably structurally sufficient if designed properly.

[DIYaholic]

When do you anticipate starting actual construction? Looking forward to seeing this come together. Good luck,

Randy,

[curriergroh]

Timeline is roughly like this:
March: Get feedback here and finish a full digital model.
April: Accumulate materials.
May: Construct and wire.
June: Calibration, dust collection, storage.

A friend who is finishing a MechMate build gave me some feedback tonight on the design, but I welcome any other input from others. I'll post images of design revisions.

Really interested in feedback on the 'gantry interface plate' whether to use or not.

[curriergroh]

Can I get some feedback on electronics and drive configuration?

X-Axis: R&P x2
Y-Axis: R&P
Z-Axis: Acme Screw

If the x-axis has two motors does the torque scale linearly? So if using 2 640 oz-in I could count that as 1280 oz-in on the x-axis?

Is there a reason to use identical motors? Or should I choose a motor that is right for the application? I'm thinking that the z-axis would not need a 640 oz-in Nema 34 motor on it.

If the gantry is 150 lb. how much torque do I need for 500 IPM rapids, for 800 IPM rapids?

[CarveOne]

Can I get some feedback on electronics and drive configuration?

X-Axis: R&P x2
Y-Axis: R&P
Z-Axis: Acme Screw

If the x-axis has two motors does the torque scale linearly? So if using 2 640 oz-in I could count that as 1280 oz-in on the x-axis?

I think yes, since they are both pushing and pulling at the same time.

Is there a reason to use identical motors? Or should I choose a motor that is right for the application? I'm thinking that the z-axis would not need a 640 oz-in Nema 34 motor on it.

Yes for the dual X axis. Not necessarily, for the Y or Z. There is a formula on the Gecko site for sizing motors.

If the gantry is 150 lb. how much torque do I need for 500 IPM rapids, for 800 IPM rapids?

My gantry weight is ~180 lbs. It can jog at 1,400 ipm with two 495 oz-in steppers being driven with G203V drivers and 72vdc. I worry more about wringing off the 1/4" NEMA23 motor shafts. I set the jog rate during cutting to not much more than the job needs. Usually around 250 ipm.

CarveOne

[curriergroh]

I've been reading about stepper motors and drivers at Gecko and here all day. I'm also running some calculations for kits with documentation at keling just to see if it will sink in faster.

From Gecko site: The driver will draw 2/3 of the rated current when in parallel.

So for the KL34H260-06-4A from Keling, rated current of 6Amps, 6 * 2/3 = 4Amps. And then 4 motors so 16 Amps. I guess I'm assuming this is a Bipolar motor in parallel.

The inductance on those motors is 1.7mH. So what Gecko calls Max Voltage would be 32 * sqrt(1.7) = 41.72 V.

What does Gecko mean by Max Voltage? I see on the forum people saying to run the motors at 4-20x Rated Voltage, but the spec sheet does not have Rated Voltage on it.

So if I use 4 of the motors above I think that the Keling unregulated power supply KL-5020 would be good. It is 50VDC and 20A. The Amps are ok because the motors will draw what they need. Is the 50VDC ok? If it is why are the kits from Keling using lower voltage supplies?

If all that is good then my setup would be:
4 x 465 oz-in Nema 34
4 x G203V
1 x KL-5020

I've seen many different types of calculations and approaches to figuring this out today. Can't say I understand it yet, but I'm trying to do the math.
Evaluations of the selections and or additional calculations I need to do?

[CarveOne]

The G203V driver has a maximum input of 80vdc. Anything above that will blow its internal SMT fuse.

72vdc is the highest input voltage you can use from a standard linear power supply and have enough margin to insure (so far) that an output voltage surge due to an input line surge won't blow the fuse. I'm using the Keling KL-7220 model PSU.

The step motor has a voltage rating that is low, like 5.1v for my 495 oz-in motor. The spec is here http://www.kelinginc.net/KL23H2100-30-4BM.pdf

32 * sqrt (5.1) = 72.3 vdc is within the less than 80 vdc input spec.

Other driver models from Gecko (and other vendors) may have different maximum input voltage requirements and will need to have the PSU output voltage calculated accordingly. In general, use the highest voltage PSU within the driver's capability, and the step motor will give the most torque and rpm it can under those circumstances. You won't get nearly the maximum capability out of a particular motor if you run it with a lower voltage driver and a 24vdc PSU.

CarveOne

[CarveOne]

- snip-

What does Gecko mean by Max Voltage? I see on the forum people saying to run the motors at 4-20x Rated Voltage, but the spec sheet does not have Rated Voltage on it.

So if I use 4 of the motors above I think that the Keling unregulated power supply KL-5020 would be good. It is 50VDC and 20A. The Amps are ok because the motors will draw what they need. Is the 50VDC ok? If it is why are the kits from Keling using lower voltage supplies?

Maximum input to their driver. (G203V, G540, etc.)

I dunno. It is only $10 lower than the 65 and 72v models. As long as all else meets the G203V input spec I would pay the $10 to get the 65v or 72v PSU. I bounced around for some time trying to decide on 65v or 72v and finally went with Leeway and a couple others who used 72v. I've been happy with it since. I have enough PSU current margin to add a couple more motors to my existing 4 motors if I wish to.

CarveOne

[louieatienza]

Pretty cool design...

I'm wondering, however, if you could mount the x rails on the vertical on the side of the table, so that the upper bearing assembly sits below the work surface? It may require beefing up the gantry uprights, but you could then decrease the overall width wihout sacrificing travel, or evevn utilize the savved space to extend the gantry past the upright for a rotary axis on the side of the machine?

I realize the r&p design allows any pulling forces to pull the gear into the rack, but if mounted horizontally, then those pulling forces would be absorbed by the carraiges? Could also use a larger plate then for more distance between the carriages...

[curriergroh]

Pretty cool design...

I'm wondering, however, if you could mount the x rails on the vertical on the side of the table, so that the upper bearing assembly sits below the work surface? It may require beefing up the gantry uprights, but you could then decrease the overall width wihout sacrificing travel, or evevn utilize the savved space to extend the gantry past the upright for a rotary axis on the side of the machine?

I realize the r&p design allows any pulling forces to pull the gear into the rack, but if mounted horizontally, then those pulling forces would be absorbed by the carraiges? Could also use a larger plate then for more distance between the carriages...

Is there any reason not to mount the x-axis rails vertically when using R&P? I haven't seen any build logs doing this with the R&P from cncrouterparts. I like the idea.

I keep wanting to design the gantry without risers/vertical sections. I think it will make the assembly stiffer, but maybe I am underestimating the 8020 connections.

[CarveOne]

Is there any reason not to mount the x-axis rails vertically when using R&P? I haven't seen any build logs doing this with the R&P from cncrouterparts. I like the idea.

I keep wanting to design the gantry without risers/vertical sections. I think it will make the assembly stiffer, but maybe I am underestimating the 8020 connections.

With the horizontal rail design there are four bearings on each side of the machine riding on top of the rails and supporting the gantry weight. The edge bearings act as guides and restrict the sideways forces while cutting. The bottom two bearings on each side are guides and restrict the plunging forces that are trying to lift the gantry. If the lifting forces never exceed the gantry weight the bearings are not doing much at all.

I used vertical rails in my 12' long machine in "CarveOne's Work Table CNC Machine" build log with Ahren's carriages. The entire 180 lb gantry weight is rolling on just two edge bearings on each side of the machine. The bearings on each side of the rails just restrict sideways forces. The bottom edge bearings restrict lifting forces from plunging forces. Unless I accidentally plunge through my table top into a steel box tube I doubt I even need these bearings.

The top bearings may need replacing more often, but that's the only downside I can think of. Many of the commercial R&P machines do it this way using right angle rails and V rollers instead of CRS bar stock rails and flat bearings.

The way my machine is built I have far less problems with dust getting onto the rails. Part of that is from adding a strip of vinyl siding material that extends over the rails. 180 lbs on the rollers effectively flattens the particles when it does. I just wipe it off with a finger. My gantry could be 3" lower than it is but I expected to cut 6" thick molds eventually. The tall riser plate could be shortened to the length needed to bolt the carriages together and tie the ends of the gantry to it for a rigid joint.

CarveOne

[curriergroh]

Maximum input to their driver. (G203V, G540, etc.)

I dunno. It is only $10 lower than the 65 and 72v models. As long as all else meets the G203V input spec I would pay the $10 to get the 65v or 72v PSU. I bounced around for some time trying to decide on 65v or 72v and finally went with Leeway and a couple others who used 72v. I've been happy with it since. I have enough PSU current margin to add a couple more motors to my existing 4 motors if I wish to.

CarveOne

Can you explain how a dual rectifier PS (KL-7220) would affect wiring vs. a single rectifier PS (KL-7212)? I searched the forums for "dual rectifier" and "KL-7220" and couldn't figure this out. Does each rectifier power two drivers for the KL-7220?

Thanks for the help. I really want to get Mach3 and the electronics running since I have space for that now. Space for the machine won't be ready until after April.