PDA

View Full Version : BUILD LOG: Router Build Mk 1



hoppo
15-05-2017, 07:45 PM
After a fair few years of 'thinking' about it, and a few months learning a CAD package to make sure it'll stand a chance of working I've finally started work on my Mk 1 CNC router. I should have started it a few years ago but living in a flat I don't think it would have gone down to well with the neighbours. Anyhow since then I've only gone and got married to the good lady wife, moved house twice and generally spent the last two years decorating and doing house renovations. It's all been worth it though as I now have a bit of spare time, and the house we bought has a shed. My shed! It's my little world where I can do what I like and the Mrs is more than happy to leave me be in there.

So I've been ordering some profile and some ally, set to work with a bench drill and made some headway with my build. As it was you lot here that inspired me to get on with it all those years ago I though I'd let you know how I've been getting on and share my findings and exasperations! So here's some photo's to tell the story so far.

21649
A bit of space to spread out in the shed. Ally profile supplied and cut to specified size from KJN in Leicestershire. Two 60 x 60 profiles to support the supported rails and Five 30 x 60 profiles as cross beams to create the base.

21650
Profile nuts and angle brackets again supplied by KJN. Captive hex nuts from eBay. If you're going to build your own router, get loads of varying sizes. I kept buying what I thought I needed to save a few pounds here and there. In the end I decided if I needed 1 I might as well buy 100 as you will use them.

21651
First couple of cross beams attached. This is all very easy, just like playing with posh (expensive) meccano.

21652
Second piece of 60 x 60 profile attached. Not yet tightened up.

21653
All cross members are now attached, still not tightened up.

21654
Placed the whole thing on the flattest thing I own. A brand new piece of 18mm MDF on a steel bench that I've welded together. I have photo's of this construction if anyone's particularly interested. It's not that interesting though. Just a few mitred bits of steel angle welded together in the form of a cube which my spirit level tells me is flat. Everything is now firmed up and all seems pretty rigid.

21655
The 20mm supported linear rails that I bought from Chai about 4 years ago attached using profile nuts and 12mm captive hex bolts. These rails are 1150mm in length with ball screws that I bought as a package from Chai. My first major bit of advice is design your machine to the nth degree before buying anything. I didn't and bought a package of rails, bearings, ball screws etc thinking it would spur me on. It did to some extent but I think it would have been easier to get the supplier to make things to my requirements, rather than me building my machine to the constraints placed on me by the lengths of the ballscrews. Just because it's a package doesn't mean that they are necessarily the best dimensions to work together. For example as we'll see later on my Z axis I have a ball screw of 350mm with rails of 300mm. After spacing the bearings sensibly to maximise the strength, I had travel on the Z Axis of just over 100mm, which is fine for what I need but there's effectively 200mm of ball screw there that will never be used.

I hope you find this build log useful. More to follow....

Clive S
15-05-2017, 08:16 PM
Its always nice to see a new build log, keep it going and don't be afraid to ask questions. Good luck with the build.

routercnc
15-05-2017, 09:37 PM
Nice work hoppo. Any pictures or sketches of the whole design?

Desertboy
16-05-2017, 04:12 PM
KJN are excellent, I live right next to them and they have been invaluable to me ;)


Good look with the build

hoppo
16-05-2017, 05:08 PM
Here are some rendered drawings of my design. Some bits are missing like the motor and pulleys on the second X axis ball screw but they are identical to the other side only mirrored so you should get the idea.

21664
Overall view.


21662
Front view.


21663
Side view.

The hardest part of anything so far was learning the CAD software to ensure my design would work. I've been using Autodesk Fusion 360. Which is very good but I haven't really got anything to reference that against. There may well be better software out there, but Fusion 360 is free and seems to have done the job even if it runs a bit slow on my computer. Apologies for the state of the pictures in this post, they don't really show the construction too well.

Desertboy
17-05-2017, 06:46 AM
Hi,
Good luck with your build, I'm no expert but looking at your cad models 2 things springs to mind.
21670

I am going to do things differently to you and not have the 4 corner plates, instead I'm going to bolt the supported rail to the sides of the extrusion and the ballscrew to the top and have vertical plate mounts.21671
Of course the disadvantage of mine is it's direct drive via couplings.

A little late now but KJN offer a drilling service as well so you can do away with the corners and bolt through the extrusion I'm going to have mine drilled AND use the corners.

A_Camera
17-05-2017, 09:05 AM
Of course the disadvantage of mine is it's direct drive via couplings.

That's not just disadvantage... it can be an advantage as well since you won't have issues with backlash. The way I see it is that the only disadvantage direct drive has is the maximum speed limitation. On the other hand, it is giving very good torque and the best possible accuracy.

hoppo
17-05-2017, 10:00 PM
Hi,
Good luck with your build, I'm no expert but looking at your cad models 2 things springs to mind.

I am going to do things differently to you and not have the 4 corner plates, instead I'm going to bolt the supported rail to the sides of the extrusion and the ballscrew to the top and have vertical plate mounts.
Of course the disadvantage of mine is it's direct drive via couplings.

A little late now but KJN offer a drilling service as well so you can do away with the corners and bolt through the extrusion I'm going to have mine drilled AND use the corners.

Hi Desertboy,

Thanks for the comments. I did think about placing the supported rails on the side of the extrusion. The one reason that I didn't was because of the tolerances of the cutting of the aluminium profile for the Y axis. KJN quote, I think a tolerance of 0 to +2 mm. Which I thought if this was the worst case and was 2 mm over I would need to use shims. Whereas bolting to the top of the extrusion gives me a certain amount of wiggle room with the bolts in the slots to remove this tolerance if required. I'm not sure if that is best practice, it just seemed to make sense at the time I ordered the profile.

I did also consider direct driving the ballscrews. In hindsight the best thing I could have done is got 1610 instead of 1605 ballscrews and then I would have had speed as well as torque by direct driving them, but at the time of designing I think I was more interested in the speed and hadn't really considered the potential backlash issues. I could I suppose easily convert to direct drive later if backlash becomes an issue.

Cheers

Desertboy
17-05-2017, 10:36 PM
Hi Desertboy,

Thanks for the comments. I did think about placing the supported rails on the side of the extrusion. The one reason that I didn't was because of the tolerances of the cutting of the aluminium profile for the Y axis. KJN quote, I think a tolerance of 0 to +2 mm. Which I thought if this was the worst case and was 2 mm over I would need to use shims. Whereas bolting to the top of the extrusion gives me a certain amount of wiggle room with the bolts in the slots to remove this tolerance if required. I'm not sure if that is best practice, it just seemed to make sense at the time I ordered the profile.

I did also consider direct driving the ballscrews. In hindsight the best thing I could have done is got 1610 instead of 1605 ballscrews and then I would have had speed as well as torque by direct driving them, but at the time of designing I think I was more interested in the speed and hadn't really considered the potential backlash issues. I could I suppose easily convert to direct drive later if backlash becomes an issue.

Cheers


Hi I'm waiting for my ballscrews to arrive from China they will come tomorrow hopefully they are in the UK been tracking them ;)

like you I should have bought 1610 at least I bought 2 1604 1m and 1 1605 1.4m

I've reversed conventional design with my short and long axis because I have 25mm supported rail fror the short axis and 15mm hiwin for the long one.
21729

I will hammer a lot more down when the ballscrews come but I was intending to just build it and only cad out the bits I need to.

I have a few choices for a base that's why I haven't worried too much at moment I thought get the axis moving first.

hoppo
17-05-2017, 10:53 PM
Started construction of the Z axis front and rear plates. Armed with a hammer a center punch, a set of hss drill bits and a few counterbore bits off ebay and my trusty Clarke Bench drill this is the outcome.

21703
Holes all marked out on the Z axis rear panel. This is the part that will attach to two ball screws and 8 bearings.

21704
Drilling starts with earnest. Amazingly way easier than I thought it would be. I did use some cutting oil just in case and got a very good finish on the holes. Again as a newbie and not being too sure about my layout being particularly accurate I oversized the holes by 0.5mm to give me a little bit of wiggle room in case of alignment issues due to my possible dodgy setting out.

21705
Finished drilling the holes on this plate. The larger holes are to take 10mm bolts to fit to the ballscrew block. Drilled at 11mm diameter.

21706
The process begins again for the front plate that will hold the ball screw bearing blocks, spindle mounts and the supported rails.

21707
More drilling.

21709
The two plates with all the holes drilled. All look to be in the right place. I'm fairly happy with how fairly straightforward this was. It was working with aluminum that was most unsure about as I have no real experience working with it. It's just like drilling wood but smoother!

21716
Countersinking of the holes to take the captive bolts that will hold the whole thing together.

21717
The two finished plates, with all the holes in the right place and more importantly all the conterbores on the correct sides. There was a moment of panic when I thought I'd countersunk the larger holes on the wrong side but was relieved when checking the drawings that all was in fact correct. I did get some quite serious chatter whilst countersinking the larger holes. These are 18mm from memory and I had the bench drill on the slowest speed which is about 500rpm. I reckon it would have been better if it could have been slowed down a bit more. Anyway with plenty of cutting oil and only a little ringing in my ears afterwards I think I got away with it.

I hope I didn't 'bore' you all to death with a post that mainly 'revolves' around drilling holes. My wife thinks I'm mental having spent a day in the cellar doing nothing but drilling. Personally I loved it and am feeling quite chuffed.

I hope you found this post useful.

Cheers.

Neale
17-05-2017, 11:42 PM
I could I suppose easily convert to direct drive later if backlash becomes an issue.

There are plenty of us using belt drive without backlash issues. Properly set up amd with the proper belt/pulley combination, backlash isn't a problem with belts. I use two drive motors for the master and slave X axis ballscrews on my router with short belts but there are quite a few builders who have gone for one motor with a long belt which works fine. I think that you should make the direct/belt drive choice based on other factors. I went belt drive as it meant that my motors are tucked in inside the overall build envelope, I can change drive ratios if I need to, and I get a bit of isolation of motor stepping from the ballscrew. It's also easier to build and adjust mechanically. Direct drive means very careful alignment (even if you are using Oldham/Lovejoy/flexible couplings). Both are viable methods but be aware of all the trade-offs before deciding.

Desertboy
18-05-2017, 07:29 AM
Started construction of the Z axis front and rear plates. Armed with a hammer a center punch, a set of hss drill bits and a few counterbore bits off ebay and my trusty Clarke Bench drill this is the outcome.

21703
Holes all marked out on the Z axis rear panel. This is the part that will attach to two ball screws and 8 bearings.

21704
Drilling starts with earnest. Amazingly way easier than I thought it would be. I did use some cutting oil just in case and got a very good finish on the holes. Again as a newbie and not being too sure about my layout being particularly accurate I oversized the holes by 0.5mm to give me a little bit of wiggle room in case of alignment issues due to my possible dodgy setting out.

21705
Finished drilling the holes on this plate. The larger holes are to take 10mm bolts to fit to the ballscrew block. Drilled at 11mm diameter.

21706
The process begins again for the front plate that will hold the ball screw bearing blocks, spindle mounts and the supported rails.

21707
More drilling.

21709
The two plates with all the holes drilled. All look to be in the right place. I'm fairly happy with how fairly straightforward this was. It was working with aluminum that was most unsure about as I have no real experience working with it. It's just like drilling wood but smoother!

21716
Countersinking of the holes to take the captive bolts that will hold the whole thing together.

21717
The two finished plates, with all the holes in the right place and more importantly all the conterbores on the correct sides. There was a moment of panic when I thought I'd countersunk the larger holes on the wrong side but was relieved when checking the drawings that all was in fact correct. I did get some quite serious chatter whilst countersinking the larger holes. These are 18mm from memory and I had the bench drill on the slowest speed which is about 500rpm. I reckon it would have been better if it could have been slowed down a bit more. Anyway with plenty of cutting oil and only a little ringing in my ears afterwards I think I got away with it.

I hope I didn't 'bore' you all to death with a post that mainly 'revolves' around drilling holes. My wife thinks I'm mental having spent a day in the cellar doing nothing but drilling. Personally I loved it and am feeling quite chuffed.

I hope you found this post useful.

Cheers.

Nice work, I'm shitting making the mounting plates the plan was to pay to have them made but I have considered trying to make them myself by printing to scale on A4 clear sticky plastic sheets. Then sticking the plastic onto the plates and use a drill press. You're inspiring me ;)

If I have them made I will have the mounts to the SBR made in steel and the mount for the z axis (Which will also be the z axis) from aluminium. Whatever happens I need to have the aluminium plate made as it will have 2 bearings pressure fitted fro the ballscrew so needs to be super accurate.

Desertboy
18-05-2017, 07:33 AM
There are plenty of us using belt drive without backlash issues. Properly set up amd with the proper belt/pulley combination, backlash isn't a problem with belts. I use two drive motors for the master and slave X axis ballscrews on my router with short belts but there are quite a few builders who have gone for one motor with a long belt which works fine. I think that you should make the direct/belt drive choice based on other factors. I went belt drive as it meant that my motors are tucked in inside the overall build envelope, I can change drive ratios if I need to, and I get a bit of isolation of motor stepping from the ballscrew. It's also easier to build and adjust mechanically. Direct drive means very careful alignment (even if you are using Oldham/Lovejoy/flexible couplings). Both are viable methods but be aware of all the trade-offs before deciding.

I've scrapped quite a few commercial printers over the last 5 years and I keep seeing the same thing, the main drives are direct drive by either ballscrews or often leadscrews. The ink is always delivered by nema 23 using belt drive with tensioners gearing to ~4 to 1. With a properly setup up belt tensioner I can see backlash is so slight as to be negligible. Thing is though making any system backlash free is a major hassle and even ballscrews exhibit backlash over time as the bearings wear unless using a 2 nut style system. The reality is though most people won't even notice minor backlash and if you can notice it then you have more than minor backlash lol.

What put me off was the belt tensioner, it's easier for me to direct drive then belt drive.

hoppo
18-05-2017, 09:33 AM
Nice work, I'm shitting making the mounting plates the plan was to pay to have them made but I have considered trying to make them myself by printing to scale on A4 clear sticky plastic sheets. Then sticking the plastic onto the plates and use a drill press. You're inspiring me ;)

Ha, that's more or less exactly what I did. The quotes to have them made were coming in at silly money. I think one quote for the two plates was over 600. This forced my hand. I figured that two pieces of aluminium cut to size would be cheap enough. I think the postage almost doubled the price. I then printed the layout to scale on the wife's cheapo laser printer. Used parcel tape to stick it to the aluminium and centre punched through the paper. I was also shitting it about this but it was pretty straightforward looking back. I thought I had a photo somewhere of the paper stuck to the aluminium, I'll have another look see f I can find it.

21733
The marking out process.

I reckon I could almost as easily have made them in steel if I had access to a better pillar drill and took my time with plenty of cutting oil.

Cheers

Desertboy
18-05-2017, 09:37 AM
Ha, that's more or less exactly what I did. The quotes to have them made were coming in at silly money. I think one quote for the two plates was over 600. This forced my hand. I figured that two pieces of aluminium cut to size would be cheap enough. I think the postage almost doubled the price. I then printed the layout to scale on the wife's cheapo laser printer. Used parcel tape to stick it to the aluminium and centre punched through the paper. I was also shitting it about this but it was pretty straightforward looking back. I thought I had a photo somewhere of the paper stuck to the aluminium, I'll have another look see f I can find it.

I reckon I could almost as easily have made them in steel if I had access to a better pillar drill and took my time with plenty of cutting oil.

Cheers

The only reason I would have them made in steel is my mate has the equipment to do it and told me it's actually easier for him to do steel than aluminium.

600 robbing bastards I was quoted 100 cash if I supply my material but I have less plates.

The good bit is once we get our machine working if we ever want to make another one we can make our own plates on our machines ;)

hoppo
18-05-2017, 09:43 AM
There are plenty of us using belt drive without backlash issues. Properly set up amd with the proper belt/pulley combination, backlash isn't a problem with belts. I use two drive motors for the master and slave X axis ballscrews on my router with short belts but there are quite a few builders who have gone for one motor with a long belt which works fine. I think that you should make the direct/belt drive choice based on other factors. I went belt drive as it meant that my motors are tucked in inside the overall build envelope, I can change drive ratios if I need to, and I get a bit of isolation of motor stepping from the ballscrew. It's also easier to build and adjust mechanically. Direct drive means very careful alignment (even if you are using Oldham/Lovejoy/flexible couplings). Both are viable methods but be aware of all the trade-offs before deciding.

Thanks Neale, that's reassuring. My main reason for going belt drive was to increase speed with the 1605 ballscrews. I've ordered HTD5 pulleys with 9mm width and belts. I figure I'll just cut slots in the vertical mounts to be able to adjust by sliding the motor to increase belt tension rather than using a separate belt tensioner.

Cheers

hoppo
13-08-2017, 04:19 PM
It's been a while since my last post, but things have moved on considerably. Here is the next installment of my build I hope you enjoy.

22555
Fashioning a bracket to hold the stepper motor at 90 degrees to the ballscrew. This is where I drilled the holes. The aluminium was precut from the supplier.

22556
After a bit of countersinking and filing I managed to get it to look like this.

22557
For some reason, after assembling it became apparent that it wasn't necessary, but after a lot more countersinking and some dodgy radial loads being applied with my pilllar drill, brute strength and plenty of ignorance I achieved my aim of countersinking the slots to sink the stepper motor holding bolts into. It's not the best looking piece of metal work, but I guess that's why I'm building the thing anyway! It should do the job.

22558
After marking through the holes on the stepper bracket mount with a transfer punch, I drilled them with a tap drill (4.2mm) for an M5 tap. This was the first of many taps. My advice to anyone attempting their own build is to buy a decent tap wrench and set of taps. I started off with the one seen here which came as a set with a set of taps and dies. It was bloody hard work, the handle kept falling off and eventually later on a cheap tap snapped off and remains in a hole on my build. I paid out for a snap on M5 and M6 tap and tap wrench. Probably cost the best part of 30 but one of the best purchases in hindsight and made the build a hell of a lot easier. As you'll see later there are plenty more holes to be tapped yet.

22559
Attaching the spindle brackets to the front plate of the z axis. Again in hindsight this isn't the best design as the bolts holding them on are inaccessible from behind once the rails are attached. Eventually I'll get round to turning the bolts around and using captive nuts on the back side. I would have liked to have drilled the holes and tapped them but cannot get the accuracy right without the use of a CNC machine. Chicken and Egg comes to mind.

22560
The front of the spindle brackets attached. Just because I can.

22561
Linear bearings attached to the rear plate of the z axis.

22562
The linear rails attached to the front plate of the z axis.

22563
And the ballscrew also attached to the front plate of the z axis.

22564
Rear plate and front plate of z axis married together and working well.

22565
As above but from a different angle.

22566
Attaching the stepper motor holding bracket to the top of the front plate.

22567
Final assembly of all components of the z axis. All in all fairly straightforward. As mentioned above the hardest parts were 'milling' (repetitive drilling and filing) out the slots for the stepper motor and tapping the threads with the worlds cheapest tap wrench and taps.

Plenty more to come. I hope you find it useful and or interesting.

Clive S
13-08-2017, 05:56 PM
How much Z travel have you aloud for? Its unusual to fit the motor on the spindle plate

Desertboy
14-08-2017, 08:32 AM
How much Z travel have you aloud for? Its unusual to fit the motor on the spindle plate

I had to take a 2nd look at his cad model at the beginning of the thread to see what was going on, the nema is a dual shaft nema sticking through the end of the plate and then belt driven.

I'm sure you saw this straight away but it confused me lol.

http://www.mycncuk.com/attachment.php?attachmentid=21664&d=1494946947

Looks like the travel is quite small.

hoppo
14-08-2017, 09:31 AM
How much Z travel have you aloud for? Its unusual to fit the motor on the spindle plate

Hi Clive, total travel on the z axis is just over 3 inches. Or 8 cm. this was the most I could realistically squeeze out of it due to the length of the z axis rails, also I wanted to keep the bearing blocks as far apart as possible to maintain rigidity of the z axis in both the x and y planes. I mentioned in an earlier post that one of the regrets I had was buying the rails and ballscrews and designing my build around those, rather than designing my build and ordering bespoke lengths to suit. Hindsight hey? Still 8cm is more than adequate for my needs at the moment.
I have loads of spare travel on the z axis ballscrew so I figure I could expand the z axis travel if needs be by adding longer rails which would hopefully not be too much of an expensive upgrade.
I can't remember the exact reason I chose to fit the stepper motor on the moving face of the z axis. I have been working on this design and build for best part of 5 years I reckon now. I'm sure there was a valid reason but can't for the life of me remember why. I put it down to age. I'm off to rack my brain and figure out what I was thinking at the time and what my reasoning was!