My aluminium CNC Mill/Lathe build.

[pminmo]

Wow, impressive! Keep the pictures coming...

[digits]

Wow, impressive! Keep the pictures coming...

Thanks - but I'm not counting my chickens before I've seen it cut!

I had a quick go at getting it moving this morning, but unfortunately one of my Z-axis Gecko's is stuck in fault-mode I had a quick check of the wiring but didn't find anything, so I'll try swapping it with my A or B axis one for now - I really hope it isn't toasted, it's a long round trip to send it for repair!

I have a nasty feeling I will be taking it apart again soon - I think the Z-axis columns need to go further back as the head is near the front edge of the table when the table is only 100mm or so from its forward limit. I do need to be able to go off the front of the table to reach the ATC stations though, hmm...

Squaring it all up is also going to be interesting - I should be able to indicate off the table to get the X-axis properly horizontal, but I don't have any accurate reference to allow me to get both Y-axis screw-assemblies perfectly lined up. Any suggestions would be most welcome!

[digits]

Well, I swapped over one of my Geckos - and got my Z-axis working. Looks like one of them is dead I then had the excitement of a run away axis as one of the encoder wires had come loose during assembly - removing and remounting the motor to get to the encoder was a right PITA.

I did however finally get it all working, but then the left half of the Y-axis stopped moving I have botched boring out a couple of my motor couplings - the one on the left Y-axis is slightly too large, and despite shimming it to the motor shaft, it has spun free under light work load. I also have an off-centre bored coupler that is making one half of the Z-axis vibrate as it moves - I hope to replace both of those next week.

Anyway, I have finally run some G-code on it - it was quite scary watching more weight than an entire X-2 running up and down the Z-axis at 9000mm/min (350IPM) - and that's not even half the max theoretical speed! Peck drilling is going to be a lot less tedious now that I have a machine that will actually peck rather than slide lazily in and out of the holes!

For anyone who's interested, here are some assembly photos from last night:

One half of the Y-axis bolted down:


Both columns and both Y's:


Nothing on the right-hand side is actually bolted down - the column is balanced on the Y-axis, and the Y-axis is resting on the base frame - this is to allow me to adjust everything once the X-axis goes on between them.

Here's the X-axis resting on the Y-saddles - I had to hoist it as it's about 40-50kg and I simply can't lift it safely



This next one was rather hairy, and probably rather stupid - I needed a way to lift the X-axis upto the Z-axis saddle plates - so I jacked up one end and shored up the other:



After a lot of swearing, I got both ends bolted on to the Z-saddles. I had thought gravity would self-level the X-axis, but these screws are solid, even without power!

[digits]

Well, not much progress to report as work has been crazy and I was away last weekend, but I have got my spindle running. Unfortunately there is a lot of vibration as the spindle pulley is wonky. I don't really fancy firing up my X1 again to make a new one, so I think I'll just try and see just how badly it affects the surface finish.

I also have a few problems - the throat is so large (>250mm) that the spindle is too far forward, and so with the table fully back it is hovering over air, and with it fully forward it is only halfway across the table Why I didn't spot it in the CAD, I don't know

I am going to have to disassemble to replace some motor couplers anyway, so I will see just how much adjustability I have with the extrusions.

My other worry is how to keep chips and coolant off my Y-axis rails, screws and limit switches. The ends of the table will be right over the Y-bearing blocks, which is great mechanically, but useless for keeping things chip free. I am considering using sheet metal under the table to direct most of the coolant to the front-centre of the machine, but I am worried about accuracy issues if I have not very dimensionally-controlled sheet between the table and the fairly precise motion components. Any suggestions would be really, really appreciated!

Cheers.

[handlewanker]

Hi digs, when I'm faced with a dilemma, I just intone, "in 6 months time all this will be just like a bad dream" It actually works.
The time came when I had to adjust the end bearing cap on my Colchester bantam lathe of 1920/30 vintage, and as it was the solid casting type with just a split in one side with a bolt through the top to squeeze the inner bronze shell to the spindle, it suddenly cracked across the top where there were two 1/4" grub screws to hold the split bronze apart once it was clamped.
It was disaster big time.
Previously I had drilled the centre oil hole out from 1/8" diam to 1/4"BSP for one of those drip feed oilers, and this had weakened the casting along a line on the top.
So I sawed the broken half of the bearing off, opposite the split line, and filed the lower bearing sides to give a half bearing, and then made a new steel bearing cap for the top.
Long story short, this was then fitted and retainer bolt holes drilled on either side and the bearing bronze refitted and scraped in.
Been working for 15 years now, so it's OK.
At the time it was a gut wrenching blow, but six months later it was just a dim and distant occurence.
In January 2008 you can look back and your present problem will be just a bad dream, trust me.
Ian.

[digits]

Thanks for the moral support Ian - I'm not too dispondent about this machine yet, but I just wish something would go to plan!

I now have my new motor couplers, so hopefully I can get the Y-axis working, and the Z-axis working smoothly. I then have to get the table top bolted to the table support and then I can start squaring everything up!

I am really, really keen to use this machine in anger, but I am going to do my best not to spend any more on it in haste! I still have piles of metal, bolts and unused electronic bits all over the place, so I need to get it working, design out the weak points of the machine and do my best to do it with what I've already paid for.

I am hoping I will get it working well enough to cut a new motor pulley without any extra spending, but if all fails, I could always go back to my X-1, but I'd really rather not do!

I am also going to resist the temptation to buy a lathe - I need to get the mill bit working well enough to construct my lathe attachment from the pile of lathe bits I have already bought.

The only major splurge that may yet happen is an auto-squaring system for the twin-screw axes. I will be interested to see how well the Y-stays square once the other screw is working - I suspect the Z-axis will improve once I can bolt the tops of the Z-axis columns togther. At the moment there is enough motor power to actually deflect the Z-columns enough to allow some significant racking (~2-3mm across 1100mm). I also suspect this means that I need to strengthen the Y-axis beams that the Z is standing on, as they seem to be twisting under high load. My original CAD did have a lot more strengthening than I have fitted so far - I may yet have to bring out my X-1 to actually make these parts

Anway it's the weekend at last - if the weather doesn't stay nice, I may actually get some serious work done

[apache405]

I was looking at Haas machines and they use a stamped metal cover that has sections that telescope in and out as the machine moves. they seem pretty reliable, although I have never seen how they are built.

The vibration in your spindle sounds bad... Your idea of letting it go might work, but it seems like such a waste of the effort you put into the machine to let some stay vibration wear the machine out prematurely.
A couple of thoughts on the vibration: Maybe the pulley is not aligned/balanced and should be replaced. The other thing that might fix it is to replace the bearing assemblies in it.

Oh yeah, and try taking a break and coming back to the problems with a fresh mind.

handlewanker is right in 6 months or less these problems will have solved themselves and you will have one hell of a story to tell.

[bigbunny5]

My 'workshop' is now a proper deathtrap - I have mill assemblies everywhere, and no room to swing a cat:

I'm so sorry I don't want to ask but the Devil made me do it, now I got to know. Would that be a Live or a dead cat, you're swinging? but it's not a proper Death trap till you get the 440AC hooked up with no cover on the breaker box

[digits]

I'm so sorry I don't want to ask but the Devil made me do it, now I got to know. Would that be a Live or a dead cat, you're swinging? but it's not a proper Death trap till you get the 440AC hooked up with no cover on the breaker box

Having spent the last two evenings sitting on/inside my new machine, I'm really trying not to imagine all the exciting accidents I could have

[digits]

I was looking at Haas machines and they use a stamped metal cover that has sections that telescope in and out as the machine moves. they seem pretty reliable, although I have never seen how they are built.

The vibration in your spindle sounds bad... Your idea of letting it go might work, but it seems like such a waste of the effort you put into the machine to let some stay vibration wear the machine out prematurely.
A couple of thoughts on the vibration: Maybe the pulley is not aligned/balanced and should be replaced. The other thing that might fix it is to replace the bearing assemblies in it.

Oh yeah, and try taking a break and coming back to the problems with a fresh mind.

handlewanker is right in 6 months or less these problems will have solved themselves and you will have one hell of a story to tell.

Thanks for the words of support. The machine has 2 main problems right now - the X-axis is not square to the table. I need to measure how far out it is, but I can see that the Z is not the same on both sides. I will have to hook up one of the Z-motors as another axis so that I can tune them separately in Mach - I am planning to fit switches to allow me to do it from the control-box eventually.

Problem 2 is the spindle pulley - it was made with lots of love and care on an untrammed X-1 - measure twice, cut once eh? Anyway, I had an idea in the bath today - I have a nicely made X-2 belt drive kit, but the motor pulley won't fit on my X-3 motor and IMHO isn't large enough to be bored out. I think I am going to re-assemble my X-1, drill a 12mm hole in the centre of some round stock, then put a bolt through the hole and chuck the stock into my X-1's spindle. I am then going to put a lathe tool in a vice and then I'll try making a V-groove pulley to match the X-2 kit spindle pulley. Is this a daft idea?


There's a pic in a lathe build thread in this forum showing the back of some sliding metal way-covers - there is some sort of scissor mechanism supporting the individual 'leaves'. I am going to have a proper think about this, but I am leaning to metal way covers.

[bigbunny5]

Having spent the last two evenings sitting on/inside my new machine, I'm really trying not to imagine all the exciting accidents I could have

Well I can tell this! The Main reason I have short hair is, In High School (what you in the UK call "Day Care") I had Lang hair almost to the middle of my back, then one of our shop teachers showed us a safety film about a HIPPIE with a Pony Tale operating a Lathe. I'll let you imagine how that worked out for him and that bit of film haunted me untill I was almost 30 and one night my thumb hit the blade of a table saw, that was when I knew even with a Hair Cut and a Job All kinds of exciting accidents are waiting just on the other side of not thinking Esp. when Lucas Prince of Insuficent light is involved

[digits]

Well, I decided to do some measurement today - if I ignore my home and limit switches, I have 445mm/17.5" of X, 325mm/28.8" of Y and 300mm/11.8" of Z travel. There is about 150mm/6" extra Z in the machine, but not in the enclosure! The Y is probably good enough, but I have a plan to increase it a bit - and if I drop the direct-drive on the X, and switch to a timing belt, I should be able to increase the X and get it more centrally located. At the moment, the length of the motor and coupler mean that the X-axis is 250mm/10" closer to the left than right of the machine.

I also measured how skewed my X-axis was - from X-- to X++, I had about 0.8mm difference in Z. I hit the E-stop to turn the mill off, and the left side of the Z slid a few mm, followed half a second or so later by the right! It would seem that I am going to have to find a good way to re-level the twin screw axes (Y & Z) every time I power off or perform an E-stop. It would be nice if it could be done with home-switches, but I'm not sure they'd give me the accuracy I'd want - about 0.05mm/0.002" across the whole X-travel.

Ah well, I think I have a bit of research and redesigning to do - looks like my X-1 will be making a comeback as well!

[handlewanker]

Hi man, as the Amway secret password is intoned to the faithfull, "You can do it", have faith, in six months............Yeahhhhh, only thing you have to worry about is what is your next project going to be, CNC T&C grinder? for all those cutters you're going to be wearing out.
Ian.

[bigbunny5]

Hi man, as the Amway secret password is intoned to the faithfull, "You can do it", have faith, in six months............

Can I be an IBO? how can I be my own boss as an independent business owner?

[digits]

Hi man, as the Amway secret password is intoned to the faithfull, "You can do it", have faith, in six months............Yeahhhhh, only thing you have to worry about is what is your next project going to be, CNC T&C grinder? for all those cutters you're going to be wearing out.
Ian.

Cheers Ian - my next project is to actually make the parts that I had hoped I could make on my X-1 in the first place!
I think I may have spent too much time on You-tube (which to my British-English brain always translates as you-idiot) though, as ideas of 5-axes and trunion tables keep popping into my head! Walk before you can run...

I think I can solve the twin-screw alignment problem by adding fairly accurate home switches and then adding index pulse generators to the screws. Unfortunately my servos don't produce index pulses, and it'd cost me about $45 per servo to swap over the encoders, but I think I can DIY something good enough for a fraction of the cost. I will need to do some more milling on my X-1, and it'll have to be dry as I can't have coolant all over my unprotected new machine, but at least I'd be milling again!

I am pretty confident this twin screw design will work eventually - it seems to work brilliantly in several commercial designs, but none of them seem to move the whole X axis up and down the Z like I do.

[digits]

Well, after having spent 15 mins painfully tramming up the X-axis by running the two Z-screws independently, I did get it perfect I even did some gentle Z-moves and it stayed perfectly level I then did some Z-rapids and to my horror, the frame moved enough to allow cause the Z-axis to rack

I suppose I could tune the servos down so that they don't have enough power to bend the frame, but that's really not a good solution IMO. I suppose I could throw time money and metal at it to strengthen it, but it might just be a band-aid on a compound fracture. Instead, I'm going back to the drawing board to see what I can do by rearranging what I've got, using the experience I've gained by building this design. Wish me luck!

[handlewanker]

Hi Digits, sounds like trying to drive a car with independent steering wheels for each front wheel.
In the 50's I was employed by a firm to build a plate rolling machine.
The job was for the dished end plates for wine storage tanks, about 3 meters diam and 6mm thick 316 stainless steel.
The job was previously done by rotating in a press and manually pressing the dished shape while rotating the plate held on a crane hook, a three man job.
The new plan was for a rotating platform with a gantry across it to carry a roller to progressively dish the plates, like a metal spinner on a grand scale.
The rotating platform was from an ex anti aircraft gun rotating base, with heavier motor and gearbox.
The two columns were made from Very big "I" beams and carried a cross beam that was raised and lowered by Acme screws in each of the vertical columns.
The two Acme screws were bevel geared to a shaft across the top of the columns and driven by a single motor to raise and lower the cross beam holding the roller.
The crossbeam was levelled initially by slipping the bevel gears out of mesh, and maintained a level status against the pressure of the roller as it traversed the slowly rotating plate.
I don't think we would have considered having two motors driving the Acme screws, even if well synchronised or electrically coupled.
Ian.

[digits]

Hi Ian - I guess if I'd designed in a mechanical link between the screws from the start, it would have been fairly easy to implement. I'd be a bit worried about backlash though.

I have seen designs which use twin motors on the Z, and I'm still convinced that my main problem is a lack of rigidity in the frame - I really need to tie the columns together. I am having a good think at the moment, but I have a nasty feeling that my credit card will have to come out again to buy some more 80/20 style extrusion. Over here in the UK, it's £45/$90 a metre!

[handlewanker]

EEEK, 45 pounds, thats a bundle, but in the end if it solves the problem that's it, game set & match, you win.

That still leaves the uneven motion of the twin motors.
You could still have some synchronisation with the twin set-up by having two toothed belts coupled together at the middle with an idler pulley.
The two motors now act as one and nothing goes out of synch even if one motor fails to drive or misses a pulse, which could happen, and if the other motor was strong enough the whole beam would go skee wiff.

Personally I would drive from the centre and have a toothed belt to each Z axis screw, but this would add more weight to the crossbeam.
If the two belts prove to be too long, it would be a simple matter to have 4 shorter toothed belts with two idlers per side.

On the job I mentioned earlier we had a cross shaft with bevel gears driven by one motor and gearbox on top in the middle, but we only had to cater for parallel up and down movement without backlash being important.
Ian.

[digits]

Hi Ian - everything in the UK costs an arm and a leg - you have to just grin and bear it! I have just finished paying for this machine, so I'm not too keen to spend too much more on it. That said, it is currently a very,very expensive paperweight, and I'm not ready to retitle this thread "Digit's Folly" yet

I have seen some pics of a Societie Genovese twin column, moving bridge mill which is amazingly similar to my own design. There are however two important differences:

1) The moving bridge bridges the narrowest axis - i.e. Y, and the table moves along the longest axis i.e. X. This gives them much greater rigidity, and less potential for skew.

2) The bridge seems to be driven by two screws but only one motor


I can't fix 1) because I don't have enough floor space for a 750mm long table with 500mm travel. I could cut down the table to 600mm or so, but still, it'd be a bit bigger than I'd like.

I can fix 2) with timing belts - I'd still use two motors l think.

I am however unhappy with the amount of moving mass I have in the the Z-axis ~60-70kg. I had tried to aim to keep X the lightest becase I thougt I'd be doing a lot of X-moves in the 3D stuff I want to make, but I'd forgotten about important things like peck-drilling.

So, I think a big redesign is on the cards at the moment. I have a few ideas, but a lot of CAD to do to see if they'll give me the travels I want in the floorspace I have.

Cheers.