How to upgrade large (1500x2000mm) router from GT2 belt? Maybe 15mm HTD5 belt?

Hey CNC forum :)

I have a large (1500x2000mm) belt driven CNC router inspired from Shapeoko 3.
It uses Megarail80 a linear aluminum profile similar to the ones on a shapeoko 3 in a combination with Openbuilds v-wheels.

This cnc build aimed to build a large format cnc on a budget. The cnc is now done and its performing okay. But now I would like to upgrade it a bit to make it able to do more “agressive” cuts. At the moment im able to cut plywood with a 6mm carbide cutter at 2200 mm/min with a DOC of 4mm. Although the cuts are fine it produces a lot of chatter!

As I see it the most weakest link in this build, is the long GT2 9mm belts (Correct me if im wrong)

So my main question is:

Would replacing the GT2 9mm wide belt with HTD5 15 or 20mm wide belt make an improvement?
This would be the cheapest fix due to an easy replacement of pulleys and belt and not to mention the price.

Or is there a different type of belt that would be better?

The other thing would be using a rack and pinion system but this would be far more expensive and I would have to also redesign and build my plates.

Best Regards

Morten


Attachment 372478

Attachment 372478Attachment 372480Attachment 372480

Attachment 372482

Attachment 372484

Attachment 372484

Attachment 372486

The only way to get good performance from a belt drive is to use the "servo belt" design.
http://www.cnczone.com/forums/linear...570-forum.html

Hey Ger21

I have been looking at the "servo belt" setup and also the dual belt used on some Openbuilds CNC.

It would be possible (at least on the Y-axis) to easily glue/fasten Gt2 9mm wide belt on to the Y rail profiles to make a "servo belt" / Dual belt system. Would that make a difference or would that still be to weak/narrow belts?

The most simple thing would be changing the current belts with HTD5-15mm wide belt - would that be better than dual Gt2 9m belt and would it make a noticeable difference from my current setup?
Can I use HTD5- 15mm belt I the current system I use now.. without any gear reduction?. Just a pulley directly on the steppermotor and on to the belt, like now?

Quote:

---

Originally Posted by morten_denmark

Hey CNC forum :)

But now I would like to upgrade it a bit to make it able to do more “agressive” cuts. At the moment im able to cut plywood with a 6mm carbide cutter at 2200 mm/min with a DOC of 4mm. Although the cuts are fine it produces a lot of chatter!

---

I'm not an expert or even a novice at designing belt systems on CNC machines, but In my humble opinion, this chatter is a result of the overall design, which has large spans and isn't very stiff relative to CNC machines that are capable of making aggressive cuts. Also, V groove bearings, excellent for a plasma machine, and acceptable for a wood machine, but not the kind of machine that can do things very aggressively. The thin cross sections over large unsupported spans make the problem worse.

Quote:

---

Originally Posted by morten_denmark

As I see it the most weakest link in this build, is the long GT2 9mm belts (Correct me if im wrong)

So my main question is:

Would replacing the GT2 9mm wide belt with HTD5 15 or 20mm wide belt make an improvement?
This would be the cheapest fix due to an easy replacement of pulleys and belt and not to mention the price.

---

I think you are wrong. That's just my opinion. If it's not much money, you could always try the larger belts and let us know, but my gut feeling is that the maximum speeds and depths of cut you can achieve are a result of the overall design, and the belt type won't fix it.

Perhaps you could try a few different types of cutting bits to try for some better performance?

If you do change the belts and notice a difference, please let us know, I'd be curious to hear about it. The belts that you have do look quite skinny.

Is any of the "chatter" noise coming from the belt system? Have you ever noticed any slippage of the belt system? Like I said, I know nothing about CNC belt systems, they aren't very common.

Hey NIC77

Thanks for you reply :)

Yes It is a very large router compared to the lack of strength :(. It could be the overall design that Is to large for a design like that.
I thought I might be the belt that was the problem, due to all the threads I read about belt driven cnc's and people always say there is to much stretch when using them on a longer span.

My inspiration for this builds, was this cnc on youtube https://www.youtube.com/watch?v=NuEN79cQLD8. I know it has a more beefy x-axis, a large spindle (2.2 kw vs. 0.8 kw), and also metal v-wheels on the z-axis. But in many way the design of the cnc is similar. This router is also running on v-wheels and only 8 on the x-axis where mine has 16 and also my Y-rails are a lot beefier. The cutting speed of this cnc would be very satisfying for my needs. I dont. believe that I would help with a stronger spindle since 0.8kw should be enought for this type of cnc. Maybe the metal v-wheels on the Z-axis is making the difference? My rails should be strong enough... at least when cutting along the Y-axis.

It sound a lot like the chatter is coming from the bit and makes vibrations in the spindle itself and to the whole machine. There is no slipping in the belts and all my cut are within tolerance?? :-/

- - - Updated - - -

Here is one more video of his machine:

https://www.youtube.com/watch?v=xm03YK9yt5Q&t=67s

Your Y rails are beefier, but if you look at the Shapeoko, he has gussets on the Y Rails joining them to the frame at regular intervals. From some of your pics it looks like your Y rails are unsupported. But from the other pics it looks like they are sitting on the frame. I can't really tell what you've done there. Is there a better way to join your Y rails to your frame?

Your gantry beam, it looks like your carriage is on both sides. Otherwise you could bolt a piece of square steel tube to it, perhaps 4" x 1.5" x 1/8". Not sure how you could stiffen it at this point seeing as you don't have a free side to bolt a stiffener to.

It looks like your Z axis has alot of travel. Do you use it to carve foam, or is most of your work on plywood and wood pieces an inch thick or less?

Probably the easiest things to start with

1. Look at adding more gussets to the frame and connecting pieces where you can.

2. Move your spindle up higher in the spindle mount and put down a bunch of sheets of plywood or mdf as a spacer effectively moving your table top up several inches. Raise your table up as high as you can. With the spindle mounted higher in the mount, this will allow for space for a cutting bit above the work piece when the work piece sits just below the gantry. You could remove this is you want to cut thick foam. If you have some spare sheets of wood lying around, this is the easiest to try. This will also reduce the moment that is seen by the rubber? wheels on your Z axis as they will be closer to the bit when it is cutting.

3. Use the shortest bit that you can in the spindle for the depth of cut that you are doing and ask around about what are good bits to use. I can't really tell you. Ger21 could.

4. Look at replacing the Z axis with a stiffer one or modify that one to use linear bearings. You could probably bolt linear rails on the sides and mount two linear bearings on each side with the bearing block at 90 degrees. What I mean is that you could use the T slot of your existing Z to mount the rail to. Or you could do something entirely better.

5. Look at ways to stiffen the gantry. Not sure how you would do this.

6. Perhaps after those things look at the belts. What size belts does the Shapeoko use?

7. A more powerful spindle may help you also. Not sure what other people are achieving as far as depth of cut or speeds with that spindle. Ask around to find other people who have the same spindle and find out what results they are getting before making a decision to upgrade it.

My gut feeling is that you will see some improvement with #1 and #2, and those don't require much work. #3 doesn't require much work either, but it will take a bit of money.

I hope that was helpful. Also, its just my opinion.

If you read the comment on that second video, he says the belts are the weakest link, and he's planning on switching to rack and pinion.
I still see that other machine as being painfully slow. You'll get poor tool life using compression bits at those slow feedrates,

If you do go with a larger belt, I'd go to at least 25mm, and use some type of gear reduction (or belt reduction) Direct drive is not a good choice for belt or rack and pinion drive with steppers.

If the belts do make a difference, what you will find is that the next weakest link of the machine will be exposed. There's no substitute for rigidity, and that goes for every part of the machine. As you go down the upgrade path, each thing you replace, will always point out the next weakest part.

First let me say after looking at your pictures I don't see a machine with stiffness to be worth upgrading. That is the frame will be a significant weak link even if you improve your motion capabilities.

Quote:

---

Originally Posted by morten_denmark

Hey CNC forum :)

I have a large (1500x2000mm) belt driven CNC router inspired from Shapeoko 3.
It uses Megarail80 a linear aluminum profile similar to the ones on a shapeoko 3 in a combination with Openbuilds v-wheels.

---

You should seriously consider beefing up the machines mechanical design which in this case amounts to building a new machine. You can get some improvements with a properly designed belt drive but the mechanical design is still going to lead to a lot of vibration, shaking and deflection.

Quote:

---

This cnc build aimed to build a large format cnc on a budget. The cnc is now done and its performing okay. But now I would like to upgrade it a bit to make it able to do more “agressive” cuts. At the moment im able to cut plywood with a 6mm carbide cutter at 2200 mm/min with a DOC of 4mm. Although the cuts are fine it produces a lot of chatter!

---

The chatter is no surprise at all. What portion of the chatter is directly related to the belts is unknown.

Quote:

---

As I see it the most weakest link in this build, is the long GT2 9mm belts (Correct me if im wrong)

---

I would have to say there are other issues that are equally a problem if you want more aggressive machining.

Quote:

---

So my main question is:

Would replacing the GT2 9mm wide belt with HTD5 15 or 20mm wide belt make an improvement?

---

Assuming you can do it in a way that implements a solid belt drive solution, that you buy high strength belts that can be properly tensioned, and that you implement gear motor drives you would see some improvements. However even 15 mm or 20 mm is pretty narrow. You need belts that can be pretensioned to avoid the common pitfalls of belt drives and don't deform rapidly under load.

Quote:

---

This would be the cheapest fix due to an easy replacement of pulleys and belt and not to mention the price.

---

Going cheap isn't a way to move forward. As for belt drives I really don't see them as being any better than than anything else if they meet your needs. However since you really didn't express your needs I can only say that in a machine this size I'd go with ball screws. Note I said a machine this size, I'd have to think long and hard about putting money into this specific machine.

The point I'm trying to get to here is that belt drives are perfectly fine when used in the right application. In the automation industry we use a great many linear slides, actuators and leadscrew drives that are belt driven. Properly implemented they are good solutions that are highly serviceable. However just as many positioning systems make use of ball screws as they better fit the applications need. So You have to consider what your needs are to determine if a belt drive is suitable.

Quote:

---

Or is there a different type of belt that would be better?

---

There are a whole bunch of tooth profiles that have been standardized. Personally I don't buy into the idea that one is a fantastic improvement over another. In your case you need to consider if belt drive improvements are cost effective compared to a ball screw, leadscrew or rack and pinion upgrade.

Quote:

---

The other thing would be using a rack and pinion system but this would be far more expensive and I would have to also redesign and build my plates.

---

You would need to do that with a reasonable belt drive upgrade. A well designed belt drive has the driving pinion supported by fairly heavy bearings due to the tension the belt is under, for example. The drive needs to lay the belt down on a frame member to deaden slap and oscillation. To get reasonable resolution you need a high quality backlash free gear box to drive the belt or another reduction method that gives you reasonable linear resolution. I really see a lot of work having to go into this machine to produce a reasonable belt drive solution.

Quote:

---

Best Regards

Morten

---

Obviously this is my opinion but I wouldn't even consider a drive upgrade of any sort until a more robust frame is had. Even if it is just the table frame, I'd beef that up along with the linear rails first. A solid, heavy, frame gives you something to improve upon. A suitably heavy frame will help with vibration and better linear rails would stiffen up the X axis considerably. Going this route would allow work on the rest of the machine at your leisure.

Thanks for all the replies! I really appreciate all of you to have taken the time and answered all my questions in detail :)

wizard: As I can understand from you reply the whole design it way to weak and with my budget I might not be worth the (Huge) amount of redesign this inc would need. And the main problem would be the strength of the cnc. So before anything try to beef it up in any way possible. From all the answers I for now will not take the chance of upgrading to wider/stronger belt, due to the complexity of adding gearing, an all this work and cost without knowing if it will help.

This cnc is first of all a hobby project and having fun and learning from the design and building of this machine :) The use of it would mainly be to cutting in wood and plastic for prototypes and in a really small scale production. An example of this kind of jobs would for example plywood desktop with linoleum (See picture). Last time I made six of those at a cutting time around 40 min each. I would be nice to be able to cut does a little bit faster or at least with the inc sounding like it is fallen apart ;)

Attachment 372770 Attachment 372772 Attachment 372774

Ger21: From you response can can also understand that adding wider and stronger belt would require some kind of reduction or gearing :(. I did read the comments in on the youtube video and there is also a video of the machine upgraded with rack and pinion. I did comment on one of his videos about the belt type his was using but I couldn't remember... but it was about 12-13mm wide. The speed that his cnc is cutting (The setup using belt and not the rack and pinion), would be a very satisfying goal for my needs. By the way - the cutters I use the most is 6mm carbide 2 flute straight cutter to get a more clean edge on the plywood. Those cutters had a fair amount of stick-out to be able to cut trough 20mm plywood with linoleum.

NIC 77: The Y-rails is only fixed by a 40x40 extrusion in each end of the cnc (See picture) and then 2000mm is probably a long span with no support at all!
The z-axis was the smallest I could by from Openbuilds and I believe the extrusion is 250mm.
Most jobs would be in plywood or plastic so to give up on some of the z-travel might be necessary.

Attachment 372776Attachment 372778

It would might help to beef of the connection of the to plates on the X-axis. At first I didn't have 5mm (Indside the aluminum spacers connections the two plates) bolt long enough, so I hade to use threaded rod. This rod was very weak and replacing them with some steel ones did help a lot. The same by adding some spacers between the v-wheels on the X-axis. (See pictures). Their might still be more to gain by adding spacers between the rest of the v-wheels or something else between the plates

Attachment 372780Attachment 372782Attachment 372786

1: To add gussets might be the next thing to do :)

2: This I will also try

The Shapeoko 3 uses the same size of belt and rails.

This is a link to my Instagram profile where there is a lot more pictures in details of the inc build:

https://www.instagram.com/twodanesonevan/


Thanks :)

Quote:

---

Originally Posted by morten_denmark

By the way - the cutters I use the most is 6mm carbide 2 flute straight cutter to get a more clean edge on the plywood. Those cutters had a fair amount of stick-out to be able to cut trough 20mm plywood with linoleum.

---

Perhaps try some 9mm or 3/8 cutters and see if you have less chatter with them.

Quote:

---

Originally Posted by morten_denmark

NIC 77: The Y-rails is only fixed by a 40x40 extrusion in each end of the cnc (See picture) and then 2000mm is probably a long span with no support at all!

---

Deflection is a function of length cubed, where length is the unsupported span. So in that sense, even though the Shapeoko rails are not as beefy, they are stiffer than your current setup. It looks like you have room to add some inside gussets without interfering with your travel. Obviously you can't add them on the outside as that would interfere with your rollers. So you could start there, add a couple of gussets on each side.

Also, for your side plates that mount to the gantry, the easiest thing there is to bolt on some steel angle iron (vertically) to the existing plates. Perhaps two pieces per side.

Quote:

---

Originally Posted by morten_denmark

The z-axis was the smallest I could by from Openbuilds and I believe the extrusion is 250mm.

---

The z axis is problematic. This may be where most of the issue lies. So that extrusion is 9.8" wide? That's 250mm. It looks smaller to me. Regardless, that's not important. The rubber wheels riding in the T slot is the weak link. At the risk of sounding repetitive, you can bolt some linear rails, perhaps 20mm size on each side of your existing z axis, with two linear blocks per side. So two rails, 4 blocks. If you look around, you could probably find some that aren't too expensive and are good quality. Then make two adapter plates out of aluminum to join to your existing Z plate.

Also, the way the table is low and the spindle is mounted at the top makes the problem worse, because those rubber wheels see a much higher moment of force x distance.

Quote:

---

Originally Posted by morten_denmark

Most jobs would be in plywood or plastic so to give up on some of the z-travel might be necessary.

---

With the table top sitting as high as you can get it, and the spindle mounted up further in the mount, you should notice a difference right away. Test it out. Find a spacer, perhaps on a small area of the table and try it perhaps on a two foot by two foot piece with the spindle mounted further up. Then decide if you want to move the entire table up.

Quote:

---

Originally Posted by morten_denmark

It would might help to beef of the connection of the to plates on the X-axis. At first I didn't have 5mm (Indside the aluminum spacers connections the two plates) bolt long enough, so I hade to use threaded rod. This rod was very weak and replacing them with some steel ones did help a lot. The same by adding some spacers between the v-wheels on the X-axis. (See pictures). Their might still be more to gain by adding spacers between the rest of the v-wheels or something else between the plates

---

I think I understand what you mean. You could add spacers to the back plate of your gantry so that the plate is extended further out, and bolt on some square steel tube, perhaps 4" x 1.5" x 1/8" on the back of it (your gantry). Yes, this might help. I'd try the other things first and save this for last if you're still not happy.

Quote:

---

Originally Posted by morten_denmark

The Shapeoko 3 uses the same size of belt and rails.

---

Ok. So you should be able to get the same performance. I'm not saying you shouldn't upgrade the belts, I know nothing of belts on CNC machines, I avoid them on purpose, but it's not the priority IMO.

If you do #1 and #2 and #4, IMO, you will see an improvement. This machine will never be a speed demon, but I don't see why you can't get this up to what the Shapeoko is doing, which for most people is still quite slow. Might cost a few hundred dollars, and none of these mods are really difficult. Then consider the belts.

Good luck. Of course that's just my opinion. Let us know how it turns out.

Hey NIC

Sorry for my late response.

The largest cutter I can use is 7mm due to the small diameter of my spindle, 65mm and not the normal 80mm.
The next thing I will due is definitely to ad the gussets :) If I cut of a little bit of the X-axis profile, i will be able to put the v-wheels on the inside of the plates to make room for the gussets on the outside.

I also might try to upgrade the belt from the 9mm gt2 to 15mm gt2 and hopefully not need to use any gearing.

After the above suggestions, if this has no effect I will rebuild the z-axis. Just for info this is my current Z-axis: C-Beam® Linear Actuator Bundle - OpenBuilds Part Store I made a custom plate with 3 polycarbonate v-wheels on each side.

I will give and update on these upgrades when done :) It might take a while due to a baby on the way :)