8x4 router project

[Cynic-al]

Hi all

So I'm not a complete novice but a long way off knowing what I'm talking about! So this thread is picking the brains of people who have done this sort of thing before!

I want to build an 8x4 router, mainly for wood, maybe plastic, maybe the odd hole in long metal stock but I have a small CNC mill so I'm not looking to replace that. Initally I was going to do a plasma based on the Simon R design and got a lot of the parts made for that but then the material below came my way and I've decided a router would be more useful. At some point I will try move on what I've got for the plasma and start work on this.

I work for an injection moulding company, we used to have several Cartesian robots to load / unload the machines but as machines have been replaced we have moved to 6 axis robots, mainly ABB. This is the last one to be pulled out after 20 years of service. Rather than throwing it in the scrap like the others I decided it would be a great base for this project.

The main part of the robot was this 6m ali extrusion. Originally they have the half round guides with bearings on but the front one that takes all the weight of the robot tended to fail so we replaced those with linear bearings. I've cut these in half and I think, stood on end they will make a great base for the machine. I even went scrounging and found a spare pair of linear bearings that we saved off one of the other robots. This gives me two for each side which should make a pretty stable base for the gantry.

30725

30726

Then I stripped the Y axis, this is long enough for the 4ft and has a solid lump to hold the spindle assembly. I think the current bearing setup will be fine too.

30727

30728

30729

Plus I have a pile of other bits which I will keep incase they are useful. The servo motors are good, the electrician offered me the drives and software but after several reminders I still don't have them. To be honest I'm not sure how useful they will be for a home CNC setup. If I can get them I will as they will only end up dumped in a heap somewhere but I'm expecting they will be of no use.

My next step is to get a spindle, I can't really design it until I know how high that needs to be off the bed etc.





So my questions to those who are experienced;

1. What spindle would you use? I was thinking one of the rectangular air cooled types with a variable speed drive and a collet?

2. Can I used the belts that are already on the robot? They are the steel wire reinforced timing belts. Or will they have too much stretch to be accurate? Do I need to swap out to a lead screw?

Thanks all!

[Colin Barron]

Have a look on youtube at other peoples projects before you spend any time and money. Have you considered a rebuild?

[Cynic-al]

Hi thanks for the reply.

I have been looking at photos of other machines to see what others do.

For the z axis I will use a ball screw.

For the 2.5m axis I'm concerned the belts will be too stretchy, I have enough belt for both sides but only one set of pulleys so I don't have the parts to just try it without spending money. Don't want to throw good money after bad so am thinking of rack and pinion.

The belt drive is already mounted and setup on the 1.3m axis so am tempted to just try that and see what happens.

I've been looking at spindles. I have an spare inverter here that will run a 2.2kw spindle but from what I've read a 4 kW would be more useful so I think I will order a vevor air cooled then then I can start work building the frames up and take it from there.

[Cynic-al]

Not sure what you mean by a rebuild?

[Colin Barron]

A rebuild is making a useful machine from a one which has out of date or failed complicated electronic/ drive systems which would otherwise be spares or scrap. For a business they are beyond economic repair, for the DIY person who does not factor in labour cost and is willing to use Chinese drives/electronics they are a viable alternative.
The main problems with industrial machines are they are very big, heavy and are often 3 phase.

[Cynic-al]

I'm with you. I guess its right place at the right time for that sort of thing

[JAZZCNC]

To be honest, out of all those bits the only parts I'd consider usable for a decent Router at this size is the Long length of profile and rails. The shorter profile with belt attached is not really strong enough for a Gantry, it would make a great plasma gantry just as it is but not really very good for a router at this size.

A belt drive system is Ok and in some ways better than R&P because it's much more efficient with less backlash but as you know they do start getting a bit flappy when you start going past 2mtr's so it's not something I'd use at this size unless was looking to do it on the cheap reusing what you have. However, the work involved in building such a large machine is high so I'd always try to talk someone out of doing it on the cheap because it never works out that way and often leads to disappointment or failure.
In fact, I'd try to talk anyone who is new to building CNC machines out of building such a large machine for their first project, it's common for people to think that it can't be any harder than building a small CNC as it's just scaled up but believe me it's very different and much harder if you want anything worthy of the BIG effort required.

My advice is to make a plan don't try to wing it and build/design as you go along! It will take you much longer and cost you more money as you'll scrap half of it in cock-ups or design changes. Taking a few days or even weeks to sit down and make a plan/design will pay back BIG time and you'll be cutting sooner with less frustration.

Here's the best and proven way to successfully build a great machine.

Do the research, make a plan, bit more research, change the plan, start a build thread, ask questions, do bit more research, few more Questions, finalize the plan, Start the build, THEN AND ONLY THEN BUY ELECTRICS

Good luck

[John McNamara]

Hi Cynic

The link below points to a build I worked on with a mate a while back 2013. The machine is still being used almost daily by the men's shed where it is set up. the X and Y axis uses timing belts. the only changes made have been two Makita routers, and now it is on its second 2..2KW Asian spindle. Yes it is used a lot.

If the timing pulleys and belts fitted to your robot bits are in good condition I would give them a try.
It is worth remembering that the total accuracy of a machine is the sum of all the parts and how they are aligned. Timing belts do work as this machine has proved.

Regards
John


Regards
John


https://www.machsupport.com/forum/index.php?topic=23730.0

[Kitwn]

Hi Cynic

The link below points to a build I worked on with a mate a while back 2013. The machine is still being used almost daily by the men's shed where it is set up. the X and Y axis uses timing belts. the only changes made have been two Makita routers, and now it is on its second 2..2KW Asian spindle. Yes it is used a lot.

If the timing pulleys and belts fitted to your robot bits are in good condition I would give them a try.
It is worth remembering that the total accuracy of a machine is the sum of all the parts and how they are aligned. Timing belts do work as this machine has proved.

Regards
John

https://www.machsupport.com/forum/index.php?topic=23730.0

Cynic-al,
If you've been reading this forum for a while you will know the sort of precision that John McNamara works to and so an endorsement of using belts from him is worth having.

John,
I notice your machine uses fully supported rails (as does mine) rather than Hi-Win or similar. With the high usage you say the machine gets have you had many problems with wear and play in the linear bearings?

Kit

[John McNamara]

G'Day Kitwin

Happy New Year :glee:

The Men's shed machine was built on a Budget, The rails and belts remain untouched to this day.
The only changes made are the spindle updates, and recently it is mounted on a new stand.

The machine was designed for woodworking and plastics, and light metals. 99% of the work is wood. Fully supported round rails are fine for this sort of application. The forces created by wood routing are fairly light. If I were designing a metal focused machine I would consider profile, Hiwin type, rails and ball screws.

The local Council has supported the men's shed with work, making chunky road signs out of timber sleepers. The machine was used to cut the deep lettering moving the wood along in stages as the signs are very long.

It was also used to cut deep pockets in about 2 metres long 100x100mm recycled plastic posts, there were hundreds of them! This was to make a kind of post and rail fence.

Then there is all the members projects.

If you study the video the long axis timing belt's (Steel stiffened) and Joined to be in tandem by a cross shaft (Note the torsion tube). are fully supported on the bed to avoid vibrating if suspended, rising up only at the drive point on the carriage. The belts are stationary. note the tensioner arrangements. These belts were were properly tensioned at build time, so far they have not required re tensioning.
The Z axis uses a 16mm ball screw.

It should be noted that all rotating parts were supported by Flange or snap ring type ball bearings clamped by laser cut retainers into the laser cut frame. Study the images. This was a big contributor to the overall accuracy. Also note the timing pulleys were carefully bored.

All in all a low cost design that works. It met and excelled the design criteria.

Regards
John

[Cynic-al]

"Winging it" and "on the cheap". It's like your writing my obituary! 🤣

I like to make things out of what I can find which I have found does sometimes lead to doing things twice or 'learning from your mistakes'

The shorter drive belt on the robot is stationary, going around the moving driven pulley so I think that would be worth trying.

The long belt was on the full 6m of ali. I'm struggling to remember if that was stationary or moving. I think probably moving as the motor was on the end of the rail not the moving carriage. I think if I alter it to be stationary I could put one either side. If your doing this can you put a motor each side and run them as a pair or is it better to physically link them with a shaft?

Your eight, I need to do some planning, I will look at the links and play on cad

[Kitwn]

G'Day Kitwin

Happy New Year :glee:

The Men's shed machine was built on a Budget, The rails and belts remain untouched to this day.
The only changes made are the spindle updates, and recently it is mounted on a new stand.

The machine was designed for woodworking and plastics, and light metals. 99% of the work is wood. Fully supported round rails are fine for this sort of application. The forces created by wood routing are fairly light. If I were designing a metal focused machine I would consider profile, Hiwin type, rails and ball screws.

The local Council has supported the men's shed with work, making chunky road signs out of timber sleepers. The machine was used to cut the deep lettering moving the wood along in stages as the signs are very long.

It was also used to cut deep pockets in about 2 metres long 100x100mm recycled plastic posts, there were hundreds of them! This was to make a kind of post and rail fence.

Then there is all the members projects.

If you study the video the long axis timing belt's (Steel stiffened) and Joined to be in tandem by a cross shaft (Note the torsion tube). are fully supported on the bed to avoid vibrating if suspended, rising up only at the drive point on the carriage. The belts are stationary. note the tensioner arrangements. These belts were were properly tensioned at build time, so far they have not required re tensioning.
The Z axis uses a 16mm ball screw.

It should be noted that all rotating parts were supported by Flange or snap ring type ball bearings clamped by laser cut retainers into the laser cut frame. Study the images. This was a big contributor to the overall accuracy. Also note the timing pulleys were carefully bored.

All in all a low cost design that works. It met and excelled the design criteria.

Regards
John

Nice to hear from you John, a very happy new year to you and yours.

It's good to have feedback about the performance of such a budget machine over several years of use, it has obviously proved well worth the investment of both time and money. Maybe not fast enough for time-critical commercial use but that's not what most newcomers to this forum are looking for. You have clearly managed the balance between cost and performance by producing an excellent design. My signature used to be a quote from the duke of Wellington: "An engineer can do for ten shillings what any fool can do for a pound".

Re the belts: I think this has come up before on the forum but one option to help avoid stretching of long belts is to glue a piece of pre-tentioned belt onto the frame so that the teeth of it and the driven belt are engaged. This locks most of the belt in place and only the length rising to the pulley can stretch. There is a name fro this idea, but I can't recall it at present.

I asked about wear on the linear bearings as I had to replace a couple of bearings on my machine after not that much use, though I suspect this may have been due to excessive strain from poor alignment.

Regards
Kit

[John McNamara]

Hi Kit

I guess you are referring to the Everman Belt system.
An interesting design Idea, Patented.

Keeping the join between the belts clean is important. Wood chips will have to be kept out.

It is also worth noting that timing belts do not like bending backwards over a tight radius. The reason I used a pair of largish ball bearings. I cant remember the exact size! Maybe 60mm. We used steel cored European belting Breco? Not no name Asian, You can look up the minimum bending specs on their website.

https://www.google.com/search?q=Everman+Belt+Drive&source=lnms&tbm=isch&sa=X&ved=2ahUKEwjaxdqbkpP1AhVp7nMBHT3HCtAQ_AUoAnoECAEQB A&cshid=1641129155934043&biw=1234&bih=720&dpr=1.25

Regards
John

[Kitwn]

Hi Kit

I guess you are referring to the Everman Belt system.
An interesting design Idea, Patented.

Keeping the join between the belts clean is important. Wood chips will have to be kept out.

It is also worth noting that timing belts do not like bending backwards over a tight radius. The reason I used a pair of largish ball bearings. I cant remember the exact size! Maybe 60mm. We used steel cored European belting Breco? Not no name Asian, You can look up the minimum bending specs on their website.

https://www.google.com/search?q=Everman+Belt+Drive&source=lnms&tbm=isch&sa=X&ved=2ahUKEwjaxdqbkpP1AhVp7nMBHT3HCtAQ_AUoAnoECAEQB A&cshid=1641129155934043&biw=1234&bih=720&dpr=1.25

Regards
John

Everman, that's the one.

Like every other option you can try, whether belts work well or not will depend on the details of the design and the precision of your requirements.

[John McNamara]

Hmmm this post got me thinkin!

Re Belt Drives....

This design, link below, will allow the drive to be preloaded using 1 full length fixed open ended belt and a conventional belt with teeth facing out.

If you study the design you will see that the two sections in contact with the fixed belt can be moved towards or apart from each other thus changing the effective pitch of the small belt sections. This will enable a small amount of preload to be set.

Easy enough to do with a few ball bearings.

https://www.youtube.com/watch?v=NJZ7YVHRwAM

Regards
John

[Cynic-al]

Out of interest do people consider a 2.2kw spindle enough?

I already have a spare inverter that will run that and an er20 collet set so there would be a cost saving there.

If its not up to it I will plan for a 4kw now and buy a new inverter. I dont want to waste money on the 2.2kw spindle if its going to be limiting.

[JAZZCNC]

Out of interest do people consider a 2.2kw spindle enough?

I already have a spare inverter that will run that and an er20 collet set so there would be a cost saving there.

If its not up to it I will plan for a 4kw now and buy a new inverter. I dont want to waste money on the 2.2kw spindle if its going to be limiting.

It Depends on the size of cutters you want to use and how aggressive you want to cut.? they can be pushed more than people realize and are very reliable considering the low cost.
Also, remember everything needs to scale accordingly so if you're using a larger spindle it will be heavy and your z-axis, motors, and gantry, etc will need to match.

John said it before and he's perfectly correct the machine is only as good as the sum of its parts or its leakest link so the Z-axis and spindle are IMO THE most important area, get this wrong, and doesn't matter how good the rest of the machine the quality of parts that come off it will always be shit.!