Yet another fixed gantry machine...

[jfong]

Have redesigned the base a bit. My philosophy is that its good to have the ball screw level with the linear s. May be very wrong on that though.
Need to know what linears and ballscrew to use and the height of it so im heading to ebay now. Any tips on what to search for? Hiwin? Size?

My search terms for linear rails are
THK linear
NSK linear
IKO linear

I much prefer THK and since you are mounting them flat horizontal with radial loading direction you can use SR type rails or the equivalent NSK LS. If you can find HSR type rails, even better.

I would think 15mm with rails would be good enough. Check datasheet for load rating. Of course 20mm width if you can find them.

I just bought two new 46" NSK LS20 rails with 4 carriage blocks for $80 free shipping, two days ago on eBay. Dirt cheap!!! Longer than I need so I will cut down the rails to fit my application. For the price I paid, I don't care if I had to cut them. There are eBay deals if you look long enough.

[handlewanker]

Ha Ha....some of us are night owls.

The reason that U section steel is unstable is because it lack the ability to resist torsional twisting forces compared to a tube, and when you have the Z axis hanging down near the table that is where the leverage of the tool is greatest against the bridge beam.

Having 2 square tubes across the X axis, one on top of the other, spaced apart to clear the ball nut, will be easier to fit and weld in place while also having inherent torsional twisting resistant properties due to their spacing.

Once again, this is only my opinion, so take it with a pinch of salt....you must analyse the suggestion to see if it has credence with how you visualise the build.

BTW, a flat metal plate is the most springiest material you can work with.......it has no strength in the flat plane and requires the edge turned at 90 deg to make it stiff.

Those moving gantry routers on EBAY have thick aluminium side plates which if they were in steel would weigh a ton......and as such if they were in steel would only need to be relatively thin, as steel has 10 time the tensile strength of aluminium, but they would flex and bend if you just looked at them.

A steel tube has two pairs of sides.....two long vertical sides and two long horizontal sides.....top and bottom, front and back.. etc.

As a beam they will have compression and tension applied alternatively on opposite sides to resist the loading.

Take one side out of the tube to give you a U shape and you lose massively the ability to resist torsional twisting.

In your latest design, I would take the two end plates out of it and replace one at the bottom end with another cross tube,......the ball screw drive end bearing and stepper motor are attached to a cradle frame that bolts to the top end cross tube as you now have it, so no need to have a plate across the end as well.

While you are in the designing and drafting stage, consider having the Y axis linear rails down on the cross tubes, inside the side tubes, instead of on top of them, and on flat steel bars welded to the cross tubes.

This will enable you to shorten the side columns as the table will be lower in the frame.

Oh dear.........here I go again.......do it my way etc........please forgive as it seems the logical way to do it.....you be the judge.
Ian.

[wizard]

You have right Wizard. Better to move on and understand that I will never get the perfect machine on the first try with my complete lack of structure knowledge and force calculations and so on.
I will make some changes and hopefully set on a design to start with on monday. Then I will need some assistance to order all the hardware.
You guys are really kind and helpful, priceless.

One thing that might help with understanding the strength of the materials being used is to try to manually bend or twist them. Channel iron for example could be held in a vise at one end and twisted at the other. Given a robust vise and a couple of hundred pounds of torque you should be able to see some twist over 4 feet. Size matters of course just like 1/2" pipe is far more flexible than 4" pipe.

Edit:

When I say "see" above I mean just that, being able to visually see the twist without the aid of a dial indicator or other tool. Channel iron and angle iron can be very twisty if you will. A gantry has to resist twisting forces if you expect to maintain accuracy at the tool.

So what I'm trying to say here is that what you want to do with the channel or U shaped profiles isn't impossible, you just need to be aware of the limitation and extra work involved. That extra work might involve a lot of welding and machining not normally required. Lots of welding introduces its own issues. As I understand it you are using this material due to having access to it which is fine. Given that you might want to consider welding up a box beam for the gantry if you can address the post welding work required.

Also when it comes to your design realize nothing is perfect. You could pull a free plan off the net, that has been tested by dozens and it still won't be perfect. In the end the machine needs to meet your needs.
Edit:

Not knowing your design and build capability I'm going to suggest buying tubing for the gantry beam might be a good idea. You might get good results from channel iron if well equipped to do so but I really believe it will be a lot of effort on your part.

[wizard]

Just thought about the time zones. Ian and Daniell is probably at sleep now and wizard having a burger for lunch. Dinner for myself here in Sweden. Internet is fantastic.

Wonderful system here. To put a twist into things I'm working nights this weekend. Beyond making guesses about lunch impossible it also means I will be a grouch by Monday.

[handlewanker]

Hi, If you have channel steel going cheap you could put two together and weld two long seams to make a square steel tube.......in that case use stich welding, that is weld in short spaced intervals...... 25mm with 25mm spacing etc, not totally along the length.

100mm X 50mm channel iron gives you a 100mm X 100mm square tube, probably with a thick wall too......8mm would be fantastic.

BTW, if you do not have welding skills, now would be a good time to add another string to your bow........welding is the easiest construction method once you get the hang of it, and steel is so forgiving if you need to do a bit of make over.
Ian.

You'd still need to have flat steel bars for the linear rails to bolt onto.

[Hartzell]

Hi
Searching ebay for linear stuff.
I think I go for Hiwin HGW15 for all axis.
All Japanese rails I could find was used ones, or to expensive.

What do you think about the x-axis (bed) ball screw, can RM1605 be a good choice? The screw will be ca 350mm long. Lower speed with 1610, but Iḿ afraid that I lose some precision/resolution with coarser pitch, am I wrong?

Also, an design question. Is having the ball screw level with the linear bearings a good thing? Or is it better to aim at getting the top of the bearing level with top of ball nut housing? I cant find that website that explain those things in a pedagogical way.

[handlewanker]

Hi, when I laid out the design for the 3025 bridge mill/router I'm building I projected the end view to ascertain the clearance between the bottom of the table and the various heights of the linear rail bearing blocks and the top of the ball nut housing and as well to determine the ideal thickness of the linear rail bearing block base mountings.

It doesn't really matter if the ball screw is significantly above or below the linear rails.......it has a definite position when attached to the table and base relative to the linear rails......your biggest problem is to get the linear rails and the ball screw in line with each other in two planes, so when you come to machine the linear rail mountings you also machine the ball screw end bearing seats at the same time.

You need to take into account the heights of the ball nut housing and the linear rail bearing blocks.....the nut housing is bolted to the bottom of the table as well as the linear rail bearing blocks, and the ball screw end bearing mountings are bolted to the base of the machine, or as you have drawn it to an angle bracket.

I decided to use 1605 ball screws as this, with a 200 step motor will, for my needs, give adequate rapids and also adequate resolution........time for rapids being not as important as accuracy in the resolution of the moves etc........the short travel lengths means rapids won't give you much gain in the time it takes to get from A to B, unless you need split second production times which for a hobby mill is not going to happen.

BTW, have you decided on what to make the table from, as it's the most difficult part of the build apart from the spindle which in most cases is a buy in item.

Any accuracy you build into the machine will be lost if the table flexes when you bolt down a piece of work etc.
Ian.

[handlewanker]

Hi, I wouldn't worry too much if the suggestions clash with your design ideas........you have to use them purely as a guideline, not as a must do .....sometimes the build becomes too problematic when you start to apply all the "ideal" solutions to a design.......no one can go the whole distance in your boots, but if you get led part of the way it'll get you there eventually.

I worked with a complete full size model build, constructed in full scale in all aspects to see if it was feasible, working from sketches and a stand back and look approach as I get a better feel for somethoing when I can see it in the flesh or nearly so.

This is the same or nearly so as when you have finished construction of the real thing and realise with hindsight that it could have been done differently.......I hate having to re-do a project for lack of hindsight.

The model got cut up a number of times to explore different aspects of the design so it was easy peasy at that stage as the model only cost the glue to hold the scrap material together.
Ian.

[louieatienza]

Looks good... So you don't have to go crazy they do make a block with side bolt holes, which would make the Z carriage design a bit easier.

[Hartzell]

Hi!
louieatienza I apologize, but dont really understand what you saying. Who does which blocks? Im terrible bad at english so please have understanding for that.

Anyway. I have cut the 80x80 tubing with a band saw, nice thing. First time.

Do i need any thicker than 8mm for base to the linear rails and ball screw bearings?



Yes Handlew, regarding the table I have thougt about what you said. I need to have that part, and de spindle/z, thought out before building the bridge up rights.
Just begins with the base now.

[Hartzell]

[handlewanker]

Hi.....you "could" do it this way by piercing the cross members for the screws but getting the holes in that thick steel will be difficult.......also the faces of the cross members may not be 100%b square in both planes, which is vital for mounting the ball screw support bearings.

I think the design of the base is getting there.....just my opinion......but it would work better if you mounted the side tubes on top of the cross members.....much simper welding.....this will enable you to mount the ball screw bearing housings on the tops of the cross members.

They can be machined with the same setting and at the same time as the linear rail mounting seats........you would probably also need to have a cross member at each end and another one about the middle due to the short travel of the table and a short screw.......the screw support bearings at the stepper motor end (two angular contacts for thrust resistance) are bolted on the top cross member, but the other end of the screw ends only half way down the base, about the middle and has an axially floating radial bearing to allow expansion etc......this is bolted to the middle cross member.

BTW.....if you look at the design you posted in the latest drawing you will see that if the linear rails were down inside and on the tops of the cross members instead of on the tops of the side tubes you'd get a lower profile and shorter more rigid columns......table to clear the sides of the side tubes by 10mm and be slightly above the tube tops.

Having said that I'm leaning towards the design I'm currently working on, but you'll have to decide for yourself.
Ian.

[Hartzell]

Hi
Have to get a grip over how big the z-axis would be. Looking at motors and there seems to be a few "standard" china models on ebay. I already have a ABB VFD. 3x230V, 9A, 0-300Hz.
Do you think 0.8kw is enough or should I aim at 2.2kw? Also a 2.2kw would be a lot heavier, which maybe isnt so good.

[handlewanker]

Hi, the 2.2 KW is a water cooled high speed model .......as far as I know........the lowest speed is 8,000 rpm.....as far as I know.....top speed is 24,000rpm......that is stated on the EBAY sellers blurb.

At 8,000rpm you won't get much torque but as you're mainly interested in milling aluminium etc that would not be a problem using the higher speed and carbide cutters.

My aim is specifically to mill steel and have a BLDC motor and a belt drive spindle, top speed probably about 8,000 rpm......still have to research that one out.

You're going to have some weight on the X axis rails and the Z axis screw, just can't get away from that one.

Judging by the comments of 6040 router users, the 800 watt spindles are quite adequate for aluminium etc.

The fact that you're designing this machine in a steel build endows it with a lot more rigidity than you probably need to mill wood, plastics and ally, so the weak link in the chain has to be the spindle.
Ian.

[Hartzell]

Oh my...
So you now saying that my effort to get at rigid machine is overkill for my needs?
Just kidding Ian, I much appreciate your knowledge. As I understand you are retired and have more years behind metal constructions than I have stepped on this earth. I,m 35. You have my respect sir!

800 watt, the Chinese ones, are smaller in dia (65mm) witch helps me a lot in my wish to design a machine that can handle a A3 with the long side under the bridge than the bigger motors that is 80mm.

[handlewanker]

Hi......no, over engineering is better than working down to a flimsy structure even though on paper it appears to be strong enough.

A moving table router or bridge mill can be made rock solid in the support structure as there are very few moving parts that need to be lightened.......this is totally different to a moving gantry type where any weight is detrimental etc.

The only problem is in the making and the machinery to make it with, but that is the name of the game......the design is only as good as the means to make it.

There'll come a time when you want to have it as rock solid as they come, and after it's made it's probably too late to re-work it again.

BTW.......I'm as new to this CNC game as the next nooby, the only advantage of the years is that I've got metal working skills and machinery knowledge from previous occupation, so that's an advantage when it comes to working with metals.

As far as design goes, everyone has his own ideas on how it should be made, so pick and choose accordingly.

One thing I've learned over the years......simplicity is top of the ladder.....anyone can make something complicated until you come to make it........... usually there is an ideal formula for simplicity that you find once you make the dang thing and see a better way.....it's called the exact science of hind sight......one reason I made a full size model before I even finalised the design layout.....it got cut up a few times but is a quick and easy way to visualise the outcome.

The only real advice I'd give anyone is to think about how you're going to make it......if that doesn't clarify the issue, making it will not happen.

The burning question will always be.....how big is your budget........a small purse does limit your potential build program, but the local scrap metal yard is a gold mine.
Ian.

[wizard]

Hi
Have to get a grip over how big the z-axis would be. Looking at motors and there seems to be a few "standard" china models on ebay. I already have a ABB VFD. 3x230V, 9A, 0-300Hz.
Do you think 0.8kw is enough or should I aim at 2.2kw? Also a 2.2kw would be a lot heavier, which maybe isnt so good.

There are several things to consider here. First is how will you plug the machine in? By this I mean does your power source have the ability to supply 2.2kW to the spindle and the rest of the machines controls. The last thing you want is a machine that trips a breaker in the middle of a run when it is trying to get full power out of the spindle. I only mention this due to popping a breaker more than once in a home shop, even on manual processes it sucks.

Second; you Also you can't dismiss the possibility that a larger spindle will have a larger chuck which sometimes is reason enough to go for the larger spindle. There is a lot of value in a spindle with a larger chuck. For one it should be stiffer but you can also accommodate a larger range of collet sizes.

Going back to power available at the outlet you plug into don't forget that the rest of the machine has to be considered when it comes to total load at the plug. Each axis will have its own burden from the servo/stepper drive, a DC power supply is likely needed and then you have the computer controls. It may be advisable to dedicate a feeder to the machine depending upon how electrical systems are done locally. I only bring this up to hopefully help you avoid disappointment if you haven't already considered the power that may be needed by the machine.

Now that I've said all of the above do understand that many users seldom fully load their spindles. If you are doing 3D carving with very small step overs you won't likely fully load the spindle. Still it is a good idea to design the machine to use its full capacity.

[Hartzell]

Hi
Yes, my plan is to dedicate a 230V/16A feeder for this machine, so a 2.2kW spindle would do regards to that.
Also, what you said about collet sizes etc made me to set for a 2.2kw.

Have not ordered any parts yet. A lot of other things to do now, around the house and so. That never ending to-do list you know.

I have sketched the z-axis now. Do you think a Nema 17 stepper with a 1204 ball screw will do for z-travel?

[handlewanker]

Hi......Nema 17's are a bit on the small size for the Z axis work on the machine you envisage.....all the weight of the spindle and the forces required for drilling too will be on the Z so a bit of muscle is all to the good......I'll be fitting Nema 23's at 425 oz/in all round on the build I'm involved in.......... but that's just my opinion.....I think the weight of a 2.2KW spindle is around 9 KG and that's a lot to move up and down.

I can't comment on the 1204 screw as I've never used one or know of it's capability to handle the loads you anticipate.......I'll be using 1605's all round as that's how the design worked out with the dimensions of the ball nuts and linear bearing blocks etc.

To get everything into perspective I built the model full scale to see how all the parts interacted as it's easier than going to a cad drawing......... which I do not have......Turbocad is still a mystery to me and a long learning curve..........the model was built in a single day and paid dividends in seeing where everything fitted etc.
Ian.

[Hartzell]

Hi
Had some vacation, now back to work.
I hope the hardware from China arrive this week.

Have some considerations that you guys might help me with.

1. The bridge up rights will be bolted to the base frame. M10 I think suffice. But how many bolts do I need? (See picture)
2. I have not decided what type of bed to use yet. Thought about two 5mm aluminium sheets glued together and mounted on top of a 20mm thick torsion box frame. Any (better) ideas?
3. What free height between bridge and bed are sufficient? Is 50mm to narrow? I now it depends on what I want to do, but the thickest i can think of right now is 22mm wood board. May be forgetting something here.