New Member

[01012f]

Hi all
I'm attempting to build myself a 1200x600 cnc router. I'm making up the plans as I go along, so probably making loads of mistakes. I've so far got the Y rails mounted and built the gantry with rails ready to be fitted. I hope to end up with something that will mostly cut wood but be capable of milling aluminium and possibly adding a 4th axis at a later date.

[01012f]

Hi all

I'm struggling with the linear rails I bought for my project.
How easily should the carriages move on a HGH20 rail?

My 3d printer has 12mm linear rails and the carriages fall freely when the rail is held vertically. I was expecting similar for the HGH20 rails but they just sit there. Even removing the wiper plates there is no roll although they do move with a light tap but then stop immediately. Is this normal?

As a test, I replaced the HGH20 rails with some Sbr16 rails and they work fine, even when I attach my end plates and gantry and my stepper motors easily move it all.

Have I bought the wrong rails? Am I expecting too much of them?

Thanks

[JAZZCNC]

This is quite normal, it's the preload that is holding them. The preload is one of the details which makes them so much better than the others. You don't want loose sloppy rails on your machine and any bearing that slides easily without any effort is basically sloppy.

Don't hesitate to fit them and forget, they will outperform and outlast all the others.

[Neale]

And they move more easily when loaded, surprisingly enough. My carriages seemed a bit sticky when I tested them on the rails as I was building. Then I leant on them and they moved much more smoothly and easily.

[01012f]

Thanks guys
I guess I'm now going to have to source dual lead screws for it, my steppers can't cope with it as it currently is (3:1 reduction & GT2 belt)

[JAZZCNC]

Thanks guys
I guess I'm now going to have to source dual lead screws for it, my steppers can't cope with it as it currently is (3:1 reduction & GT2 belt)

Why would you source new lead screws if the steppers are too small.? You might be throwing good money away on the wrong things.? I suggest you post up the full spec of what you have and a few picks of the machine so we can take a look for you.

[01012f]

Hi JazzCnc

I think I over-spec'd the rails, I could probably get away with smaller rails but I have them now.

I only have 50oz nema17 stepper motors and the way I had it all assembled, the motors can't cope. On each side of the gantry I have a 20 tooth pulley on the stepper driving 60 tooth pulley on a shaft with a 20 tooth pulley on the other end. A GT2 belt runs around 2 idler pulleys to that last 20 tooth pulley. I think my dual y-axis controller board is splitting 1 stepper output rather than having 2 full power outputs. It would probably work as planned if they were full power outputs.

So, my thoughts were to replace the belt drive with lead screws. That should give me a ~10:1 advantage rather than the belts 3:1.
And yes, lead screws are expensive, so I'm first going to test the princliple using m10 studding.

I have some (lack of) progress videos on my youtube channel

[dazp1976]

Hi JazzCnc

I only have 50oz nema17 stepper motors]

There's your problem.
Even cheap chinese machines that size have at least 425oz nema23.
Steppers are too small.
The rails are prob really good tbf.

Look at other machines that size and see what they have.
I suspect 425oz > DM542/DM556 >48vdc.

[JAZZCNC]

Hi JazzCnc

I think I over-spec'd the rails, I could probably get away with smaller rails but I have them now.

I only have 50oz nema17 stepper motors and the way I had it all assembled, the motors can't cope. On each side of the gantry I have a 20 tooth pulley on the stepper driving 60 tooth pulley on a shaft with a 20 tooth pulley on the other end. A GT2 belt runs around 2 idler pulleys to that last 20 tooth pulley. I think my dual y-axis controller board is splitting 1 stepper output rather than having 2 full power outputs. It would probably work as planned if they were full power outputs.

So, my thoughts were to replace the belt drive with lead screws. That should give me a ~10:1 advantage rather than the belts 3:1.
And yes, lead screws are expensive, so I'm first going to test the princliple using m10 studding.

I have some (lack of) progress videos on my youtube channel

Hi, Yes I went to your Youtube after posting what I did and saw what you are planning and what you are working with!

I understand testing the waters with what you have but what you have is just wrong in every department, but I think you get that by now.
My strong suggestion if you are serious about a router worth the effort is to forget all the electronics you have, right down to the PSU.? They will only frustrate the hell out of you and massively underperform.

NEMA 17 motors are ok for a small super lightweight machine ie: 3d printer using thin light belts etc but they just can't handle the inertia of larger lead screws and linear rails that you will need for a router this size. You'll need at least 2Nm Nema 23/24 and ideally 3Nm or more, however, the real key is the voltage you run them at and this requires the drives to match and have a little spare overhead.
This is why I said "right down to the PSU" because for a typical 3Nm stepper you want to be running them with around 50-55Vdc, smaller motors like 2Nm you can get away with a little less at 40-44Vdc. If you run them with less voltage than this you will be compromising performance and it's just not worth doing.

Regards the drives then don't buy Cheap analog drives, they are old tech and run very rough compared to digital drives, there is a night and day difference in performance and smoothness. You also want to leave a 10% margin of safety on the maximum voltage they can handle compared to the voltage you run the motors at.

Lastly the controller.? This is another KEY component to a good machine and especially a router as you need a nice fast and smooth pulse train.
If you want a stable machine and one that performs well then don't use the parallel port (Unless you use LinuxCNC, but that's for another time!) because again it will frustrate the hell out of you and is relatively slow.
I would avoid a USB-based controller because while better than PP they can still be a little flaky depending on which controller you use and the PC it's used on.
An Ethernet-based controller is the only real sensible choice if you want a rock-solid machine that performs great.

In my experience, any trying to cut corners is always regretted and costs more money. I've seen it time and time again and what's crazy is that it doesn't actually cost much more to do it right the first time. It also makes the build go much smoother, far less frustrating, and enjoyable.

Good luck and if you are unsure of what to buy just ask. The golden rule is never Buy before asking or checking if it's right for your application.

[01012f]

Hi Dean

That's a great post, thanks.

I've just has a look at your site, wow, you build some gorgeous machines

After feedback from here, the upgrades include lead screws, nema 23 steppers, decent drivers and a new controller so they'll come as funds allow.

Thanks everyone for the help and suggestions