Desktop 1000 mm x 600 mm Gantry Router

My aim is to build a desktop router utilising 43 mm Kress spindle, it will mainly be used for cutting wood and aluminium plate.

It will be a hobby machine but I would still like it reliable and accurate as I intend to use it mainly for cutting birch ply for RC model aeroplanes and would like them to 'click' together. I will also use it for all the other fun DIY projects.

I don't have a workshop, my only tool is a Stepcraft D600 and 3d Printer, so the design has been tweaked so I can machine most of the parts myself. The only two pieces I will have to subcontract are the top and bottom Z axis plates as I cannot get the vertical accuracy.

I am trying to balance the complexity / weight / accuracy

All the axis are direct drive so there are no pulleys etc

The design utilises :-
1000 mm 1605 C5 twin ballscrews on the Y axis
750 mm 1605 C5 ballscrew for the X axis
200 mm 1605 C7 ballscrew for the z axis
NEMA 23 74 mm stepper motors on all axis
CW5054 stepper drivers
400 Watt PSU
KK01 Breakout board
All the extrusion is standard off the shelf profiles (Oozenest and KJN)
HGR15 linear rails and HiWin carriages on all axis
Low profile t Slot plate (vacuum tables)

Cutting Area should be 850 mm x 550 mm with 70 mm vertical clearance

Attachment 29967Attachment 29968Attachment 29969Attachment 29970Attachment 29971Attachment 29972Attachment 29973

Try to imagine the cutting forces on the tool and the movement you will get on the sides. There is a gantry force calculator somewhere on this site as well as lots of ideas for gantry designs that work well .....don't be disheartened though, I got to design 14b before I got any real positive feedback [emoji51]

Charliehttps://uploads.tapatalk-cdn.com/202...4e72d9162f.jpg

Sent from my M2003J15SC using Tapatalk

Hi Charlie, thanks for the feedback. When designing the machine I considered various issues with possible flex in the various axis. Like you the ability of the gantry to move due to cutting forces concerned me. Although I have not done any complex calcs working through the axis, my logic is......
Z axis - 4 Hiwin carriages on 15 mm rail will easily support 5Kg of the spindle, carriages and plate
X Azis - 4 Hiwin carriages on 15 mm rail, I am using closed sided aluminium extrusion plus the rail, as long as I don't compress the axis when setting up there will be virtually no flex, it needs to support 12kg (see picture of extrusion)
Y axis - again 4 carriages on 15 mm rail, in compression this shouldn't be a problem, the only issue is the sideways cutting forces as you pointed out. I am using 8 mm bolts to attach the gantry plates and they will be embedded in 2 mm pockets as will the carriages and ballnuts. This leaves about 150 mm by 750 mm 'box' shape that has the potential to deflect. I am using 10 mm cast tooling plate. My gut feeling is this will not flex since the cutting forces for a 1 kw spindle will be comparatively light

Attachment 29979

If there is too much flex I have a couple of get out of jail options
1) add a cross beam, I would prefer not to do this as I am planning to use the space under the machine to house the electronics
2) Add another rail on each side with more carriages, again more complexity, weight and cost

Attachment 29981

Attachment 29982

40 mm x 40 mm KJN aluminium extrusion

I am starting to cut soon so I will have to fix the design

Thanks

Ian

Slight change to design, working on my KISS principle (Keep it simple stupid) I have removed the longitudinal extrusion between the Y axis plates, the 10 mm plate supporting the machine and Y axis ballscrews will be of sufficient strength.

Attachment 29983

8 fewer holes, reduced cutting and plate, removed 2 pieces of extrusion.

I was hoping to have two Y axis plates across the whole width of the machine but I cannot cut this length and to buy them will be too expensive so I have opted for a '4 feet' design.

Quote:

---

Originally Posted by Soyb

Hi Charlie, thanks for the feedback. When designing the machine I considered various issues with possible flex in the various axis. Like you the ability of the gantry to move due to cutting forces concerned me. Although I have not done any complex calcs working through the axis, my logic is......
Z axis - 4 Hiwin carriages on 15 mm rail will easily support 5Kg of the spindle, carriages and plate
X Azis - 4 Hiwin carriages on 15 mm rail, I am using closed sided aluminium extrusion plus the rail, as long as I don't compress the axis when setting up there will be virtually no flex, it needs to support 12kg (see picture of extrusion)
Y axis - again 4 carriages on 15 mm rail, in compression this shouldn't be a problem, the only issue is the sideways cutting forces as you pointed out. I am using 8 mm bolts to attach the gantry plates and they will be embedded in 2 mm pockets as will the carriages and ballnuts. This leaves about 150 mm by 750 mm 'box' shape that has the potential to deflect. I am using 10 mm cast tooling plate. My gut feeling is this will not flex since the cutting forces for a 1 kw spindle will be comparatively light

Attachment 29979

If there is too much flex I have a couple of get out of jail options
1) add a cross beam, I would prefer not to do this as I am planning to use the space under the machine to house the electronics
2) Add another rail on each side with more carriages, again more complexity, weight and cost

Attachment 29981

Attachment 29982

40 mm x 40 mm KJN aluminium extrusion

I am starting to cut soon so I will have to fix the design

Thanks

Ian

---

It's not so much the deflection as it is the lack of rigidity which will lead to chatter, premature tool wear and tool breakage ect, probably ok for woods but I would imagine you'll struggle cutting aluminium.

Charlie

Sent from my M2003J15SC using Tapatalk

I would consider 10mm pitch screws on X and Y and 5mm on Z

I have to admit I had not thought about vibration / resonance, with an open x axis there is a potential for chatter

Hi Clive, why larger pitch ballscrews? from a newbie I assumed that 5 mm would give better control, less backlash etc

Quote:

---

Originally Posted by Soyb

Hi Clive, why larger pitch ballscrews? from a newbie I assumed that 5 mm would give better control, less backlash etc

---

Steppers have higher torque at lower speed . Also you will get less whip with a lower speed on the screw

Quote:

---

Originally Posted by Soyb

Hi Clive, why larger pitch ballscrews? from a newbie I assumed that 5 mm would give better control, less backlash etc

---

The pitch has nothing to do with how much backlash you get and everything to do with the speed and resolution. Clive is correct that for a router that will cut a range of materials you are better with 10mm pitch on X & Y as it covers a much wider range of feed rates for the various materials correctly. Also, like Clive says it allows you to keep the rpm down which gives higher torque where you need it, which is when cutting.

If you are planning on cutting mostly Aluminium then a 5mm pitch would be better as it allows higher resolution, that said 10mm pitch still gives a decent resolution. However, if you are mostly cutting aluminium then you'll need a stronger machine than what you are planning now.
The design you have now is ok for an all-around machine, though it could use a little beefier gantry sides, it would also benefit from using my "L" shape gantry design which will be much stronger than what you have shown. If you do stay with your gantry design I would drop the plate at the rear on the x-axis carriage as you will find it a pain to set up and doesn't add very much strength.

I have built lots of machines using a similar design with higher sides and the design is stronger than people realise, so don't let the naysayers put you off because as an all-round material cutting machine they work very well. I also have customers who run very very successful businesses with my machines which are very similar in design.

Some advice would be to use 20mm rails rather than 15mm as the bearing size is easier to work with and just suits a machine like this better, the cost difference isn't much more and worth the extra. 15mm bearings are fiddly little things.

Use 15mm plate as a minimum for gantry sides and Z-axis. Remember the Z-axis is the most important area as it holds the cutting tool so if it's weak and vibrates then doesn't matter how strong the rest of the machine is you'll get rubbish finish and excess tool wear.

DUMP the KRESS, they are toys compared to water-cooled spindles.

DUMP the KK01 Bob and buy a decent Ethernet motion controller like the AXBB-E and run it using UCCNC, this will be the best investment you'll make.

Lastly but not as important as the controller but still does make a big difference, I would use higher power drives than 50Vdc and if you can afford it go with closed-loop steppers as they are SOOOO much better than standard steppers.

Building a good machine is all about the balance of electronics and components along with a strong design, if anyone is weak the machine will suffer.

Good luck.

Hi Jazzcnc, thanks for the detailed reply

More things to consider!

I keep thinking of different ways of doing things and then over complicate the design trying to do everything, I keep having to pull myself back to the original concepts and need. The machine will be for cutting mainly wood accurately with the occasional plastic, CF and aluminium. I have based the design on what I currently use which is a Stepcraft D600, its a nice little machine but definitely for the hobbyist, leadscrews and rollers on extrusion. This has been a great tool to get in to CNC / machining but its maintenance and lack of accuracy is now an issue.This machine is the next step and instead of just buying a new machine I thought it would be more interesting to design and make my own. (and cheaper).

I am using my Stepcraft to make the parts so I am limited in what I can do hence the design attempts to use as many off the shelf components and the use of a Kress and KK01 BOB as I already have a UCCNC UC100 controller and license.

I will upgrade to 15 mm gantry sides and z axis and look at the spindle / controller options / closed loop steppers although these might be a version 2 upgrade once I can sell the Stepcraft.

Thanks for the feedback

Ian

Added design for housing electronics under the machine, I am going to put a drag chain from the gantry ballscrew to the electronics.

Attachment 29988Attachment 29989

Some minor updates, I have decided to use proximity sensors instead of mechanical switches for the limits / homing so Slight change to each axis to have mounting holes
Attachment 30013Attachment 30014Attachment 30015

I have also increased the length of the z axis rails to allow a accommodate 4 blocks and give me ~ 110 mm travel, and redesigned the spindle mounting plate to allow access to the grease nipple on z axis

Attachment 30016

In picture 1 if the is the hole for the home switch it might be better to not hit it end on but have the target passing across the face.

They tend to get crunched the way you have it.

Started the build today, small steps, 3d printed the feet and stuck rubber pads on the bottom

Attachment 30017

Cheers for the info, I have only ever used mechanical limit / home switches, I will have a think how I can mount them

I have proximity switches mounted inline - can confirm crunching....

Definitely mount the switches at 90 degrees / not in the path so a crash / run away cannot damage them.

Use physical stops to prevent running off the end of axis.

I have the option of using mechanical switches that have enough travel to react in time to a crash / runaway so I will use them but have a play with the proximity switches to learn how they work, every day is a school day!

Cross Beams finished, starting on feet. Once I have the base frame assembled I will start on a test setup of the electronics so I can drive the twin Y axis and limit switchesAttachment 30019

Quote:

---

Originally Posted by Soyb

I have the option of using mechanical switches that have enough travel to react in time to a crash / runaway so I will use them but have a play with the proximity switches to learn how they work, every day is a school day!

---

Proximity switches are much better, even the cheap ones, than mechanical switches unless you buy expensive high-quality mechanical switches.
They react faster and are more reliable, a long time ago I did a test showing the repeatability of the cheap proximity switches to show how repeatable they are, it's on youtube and this forum somewhere. But since that time I have literally fit 1000's to machines I build and I have very few issues, probably 2-3 in every 100 will arrive faulty and I rarely have any fail when in use on the machine, I can't remember the last time I had one fail on a machine.
However, everyone I've fit slides by the target rather than face on. They sense and repeat just fine when sliding past the target and you don't risk crunching them if crashing the machine at high feeds where inertia will cause overtravel even though the switch reacted quick enough.

First stage finished. nice and solid using standard C beam extrusion and 4 end plates.
Attachment 30025

I decided to go with 1605 ball screws for accuracy as speed it less important for a small machine, next stage is to fit the 2 y axis stepper motors and test connect the electronics, limits etc

Attachment 30026Attachment 30027Attachment 30028

Very kind of you to show all the pictures upside-down for the benefit of us Australian readers :highly_amused:

All you need to do now is take a hacksaw and a drill to those lovely end brackets and modify them to take the proxy sensors the right way on.

Kit

The pictures were the right way up on my computer.....but glad to make our antipodean friends happy,

I have cut end plates with and without the lugs so I have options, I am going to test proximity sensors and mechanical switches and have a play with the electronics before going any further as I want to get the the speeds, limits, stops etc all working and sorted

My immediate problem is that the frame is 1 mm wider at one end, I am going to try and measure carriage to carriage to get an accurate dimension and shim the cross beams to try and remove as much inaccuracy as possible.


Attachment 30036Attachment 30037Attachment 30038Attachment 30039Attachment 30040

I put some temporary plates on the Y carriages and used the cross beam to check the difference between each end, using some feeler gauges the gap is virtually identical at ~ 0.4 mm, the gap seems to be with the edges of the extrusion and the T slot plate so I am happy that the rails are square. I am going to put 1 mm pockets in the gantry side plates for the ball screw nuts, carriages and cross beams, so I can accommodate .2 mm either side for the X axis cross beams. Phew, first issue wasn't.
Attachment 30045

Started to 'play' with electronics, all axis working, slaved C to Y axis, tested proximity sensors and they will work with the feeds I have setup without crashing. Relays setup.
Having a play with steps/ feeds etc Realistically the machine does not need to be super accurate so I am going to set it up to have a resolution of 0.01 mm with an expected tolerance of +/- 0.05 mm ?

Pleased with the results so far, I am using Stepperonline 2Nm Nema 23 Closed Loop Stepper Motor 5A 76mm & Servo Driver CL57Y, the movement is nice and smooth

Attachment 30046Attachment 30047

Waiting for some electrical components, once they arrive I will assemble the electronics properly

Latest update, started cutting the plate for the , spindle, z axis. I have subcontracted the top and bottom of the z axis as I cannot get the squareness required and these are critical for the smooth running. My Stepcraft is good enough for plate Y/X but isn't square on the Z, anything over 10 mm thick and you can see a noticeable slant.

Attachment 30077Attachment 30078

Also start finalising the design for the dust shoe

Attachment 30079

And doing test layouts for the electronics, I am mounting them upside down under the machine on rubber shock mounts

Attachment 30080

Finished Z axis and done a dry assembly, using a standard Stepcraft dust shoe, next job is machining the gantry plates

Attachment 30109Attachment 30110Attachment 30111

Very nice, keep up witht he good work :thumsup:

Cut gantry sides, I had to use 10 mm cast plate as it is the maximum thickness I can cut with any reasonable accuracy. Very happy with fit, Ball screw support nuts and HGR15 carriages fitted in to 1.5 mm pockets in gantry plates and x axis cross beams 'sat' in their pockets with just a very small amount of play, about 0.8 mm. The gap can be closed with light finger pressure. I am going to thread the holes and tighten up x axis cross members and loosen off the ball screw support nuts and it there is still a gap I will make some shims.

Attachment 30127Attachment 30123Attachment 30128Attachment 30129

Attachment 30225Attachment 30226Attachment 30227Attachment 30228

Latest stage, 'dry' assembled all the mechanics and have wired up the main electronics, they are going to be installed upside down under the bed supported by 8 anti vibration mounts. I have moved all the axis and they are all very solid (its only a 1kw kress spindle so the forces will be low compared to most of the machines on this forum), the axis are very quiet and smooth running very pleased so far. Next stage is to 3d print drag chain mounts for x axis and route all the cables and fix main drag chain under the table to link Z/X to electronics.

(no idea why the pictures are upside down, cannot seem to sort it out)

It's looking very smart.

Quote:

---

Originally Posted by Soyb

(no idea why the pictures are upside down, cannot seem to sort it out)

---

That's funny, they're the right way up down here in Tasmania :cool:

Finally worked out how to get the photos up the right way! the website doesn't read the photo orientation meta data so I have to open them in good old fashion paint in windows 10, rotate to the correct way and then save and upload, sorry to the antipodean forum members but they should be up the right way from now on!

Latest up date, just finished the electronics, tested limit switches, working on feed rates.Will have it the right way up tomorrow.
Attachment 30325

Finished (albeit a bit of fettling). Waiting for a optilaser to clear cutoms, need to re cut the rear panel for power connectors, need to decide if I am going to use the AXBB-E to power spindle, need to to replace 1 proximity sensor on the Y axis after a issue homing with two sensors. I have now set it up to use use the same sensor to home the Y axis and slave C axis and manually square the Y axis. I have closed loop steppers so there should not be an issue with alignment. I have added the USCB so I can add a 4th axis if required.

I think the machine has cost approximately £2000, and for that I have a CNC machine that is capable of cutting all the materials I use with an accuracy / reliability far greater than my Stepcraft. Compared to most machines on this forum it is in toy category but it is ideal for what I want it to do, I can use a laser attachment, vinyl cutter etc.

Thanks to the forum for all the help, now to actually start doing some cutting!

Attachment 30330Attachment 30331Attachment 30332Attachment 30333

It looks like a nice clean machine, well done.

If you can I would put a WC spindle with vfd on it.

I am making similar machine 610x1220mm working area (1/4 of full size board, but with wide gantry so half size board can be supported by bed.
I like your design - got very similar on 30x60 IR profiles from KJN
Wonder if your side gates gives enough stiffness/rigidity?

Hi Tom, I was using 10mm plate for side plates but replaced them with 15mm. What size spindle are you thinking of using? My machine was designed for making RC planes so I only cut wood, and thin aluminium. I use a 1kw AMB (Kress) spindle. This adequate for my needs. I have been using the machine in a small workshop and haven't had any issues with it, accuracy is about 0.01 mm.

If you are going to use a bigger spindle then you might want to redesign the side plates by doubling them up. I only had a Stepcraft D800 to make parts for my machine so I was limited to what I could cut and accuracy.

Have you thought of using

https://vallder.com/product/cnc-x-ax...rofile-200x96/

I used C-Beam but only one rail, if you are going for a beefier machine then you could add another rail and 2 x 10 mm side plate

I hope this helps

Cheers


Ian

Hi Ian

Thanks for reply.
Spindle 1.5kW or 2.2Kw depends on what profiles I use mainly 30x60 or 40x80 both IR.
Normally I would weld the frame, but time is factor here so will use extruded aluminium IR as quick and cost effective solution.
IMO Vallder profile is ok for 6040 size machine and light Kress.
I like your workmanship, attention for details. Please share your rc models job.
I do make mold for sailplane avionic panels.

Cheers
Tom

Hi Tom,

My design ethos was to build a desktop machine with reasonable accuracy, it’s not a production machine but has been used regularly. It is a toy compare to many machines in this forum but it fits my criteria and works well. For a more robust machine I would add extra rails to the Y axis, and re design the X axis so the rails are forward facing and use stronger extrusion, probably a larger C channel. A more robust machine = heavier components = stronger frame = heavier machine etc etc, I had to decide at what point in the design process I had achieved my goals and stop developing / adding complexity. I only had to buy 2 machined parts that I was unable to make, but with a bit of redesign I could remove them to use just flat plate. 2 time around I would make improvesment but still end up with a very similar machine in size / capabilities. Sound like your machine will be similar but a little more robust.