-
5 Attachment(s)
Definately a scrapyard challenge
Hi guys. This is my first attempt at building anything like this myself. I have a history in motorcycles for 45 years and used to be a mech eng designer. Its a bit diferant without a full blown workshop doing your bidding. This project has a limitation on funds but I will not skimp if possible.
As it is the first I have decided to make a torsion box base. If this all works then I can upgade to a metal frame later. I also thought that linear rail bearings would be better than bearings on angles.
I realise that regidity is essential so the y gantry has been constructed from 10mm Ally. Although it came cheap the angle was not exactly 90deg :rolleyes:. Bit of patience and it appears to be fairly accurate. When the rails arrive for it the will be measured off the x rails to locate. If it ends up being too heavy then I can cut sections out to lighten the load. It rolls well but inertia maybe a problem.
I have ordered ballscrew, 2 for the x and 1 for the y. My next decision is stepper size (3nm maybe) and whether to build or buy the z.
Comments are always welcome. Already had great advice on spindle choice.
Bruce
-
Re: Definately a scrapyard challenge
I too started with my X-axis rails mounted to a torsion box made from MDF/pine with a gap left either side to hide the 2 leadscrews. I'm still using the torsion box as the machine bed. Making the steel frame for mine made a huge difference. What size are your X-axis rails, they look like only SBR16 or 20mm?
Plate is weak for the gantry since when the machine is cutting parallel to X the cutting force will try and bend the gantry, and that's in the direction that's worst for a rectangular cross section. It's also the worst cross section to choose for twisting. Adding extrusion, or more plate perpendicular to it to the back would help, but you could try it and see and make sure you leave space to add it at a later date if required. What is the width of the plate, so I can calculate the deflection (roughly)?
I'd be inclined to add some triangular pieces to the aluminium angle to stop that flexing.
3nm is a good choice for the stepper motors. Lots of people, including me, use them. If you can afford 70 volt drivers (like MDS752) then you will get a higher feedrate than 50V (like MDS542), but it's more difficult/expensive to find/make a suitable power supply. The mass/inertia of the gantry is not a problem with these motors. Chip's and mine (and many others) using the same motors weigh a lot more.
Have you bought the ballscrews yet? If not then get 10mm pitch (RM1610) for X/Y to get a higher feedrate and use HTD timing pulleys (although 1:1 ratio is about right for those motors and 10mm pitch screws adding the pulleys still helps as it reduces resonance). 5mm pitch screws (RM1605) will work but it wont go anywhere near as fast, or more importantly the acceleration will be less and possibly vibrate more as the screw is spinning twice as fast.
-
Re: Definately a scrapyard challenge
Thanks for the heads up guys. All the material is 10 mm thick and 150mm wide. I had wondered about the the plate for the gantry but thought I could add box/angle to the back if required. the rails are 16mm
Unfortunately I have already ordered the lead screws and they are rm1605. I really should ask first DOH!. I am going to look into ordering the motors controllers this week sometime.
The angles on the rails were left hanging for, just as you say, the support for the nut on each side. I also thought that this allowed for the Y screw supports to be further apart and thus allowing me to use more of the bed width.
I really appreciate all your help guys
Bruce
-
Re: Definately a scrapyard challenge
x and y screws plus the y rails arrived today. Had a thought about the gantry design and have to agre especially with the angle on the rails. Looks like a bit of a rebuild required. I will also have to think about the z gantry.
-
Re: Definately a scrapyard challenge
Jonathan
Quote:
Have you bought the ballscrews yet? If not then get 10mm pitch (RM1610) for X/Y to get a higher feedrate and use HTD timing pulleys (although 1:1 ratio is about right for those motors and 10mm pitch screws adding the pulleys still helps as it reduces resonance). 5mm pitch screws (RM1605) will work but it wont go anywhere near as fast, or more importantly the acceleration will be less and possibly vibrate more as the screw is spinning twice as fast.
He is using an MDF box and i think the screws he has ordered are much more suitable, you need to stop pressing everyone with the need for speed? shake rattle and role will not bode well here.
Bruce your screws will be just fine for the application you have there.
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
2e0poz
He is using an MDF box
So was I, and I am now though it's no longer there for strength. To start with that will be a limiting factor, but there's always the option to add a steel frame ... or even a simple rectangle on the bed made from steel box would help a lot.
Quote:
Originally Posted by
2e0poz
you need to stop pressing everyone with the need for speed? shake rattle and role will not bode well here.
That's why I mentioned you'll get better *acceleration* with 10mm pitch screws, and less vibration from the screw as it's not spinning so fast. I agree high rapid speeds are pretty much useless, but good acceleration is important. When I cut the mayan calender using lower aceleration values (as I was limited by the computer) than Jazz did added over an hour to the time it took. That's quite an extreme example admittedly, but the general trend is true.
Yes they will work fine, but not as well as could have been.
-
Re: Definately a scrapyard challenge
The problem is every time you mention 1610 screws it makes it sound like everybody has to rush out and buy them?????? just remember to when you first started mister and how difficult it was. I use 1605 screws and belts on mine and it is more than enough for it's job, i could have my X axis twice as fast as yours with the belt but it would be pointless.
Thats it my rant is over and i'm stepping back from the computer for a cup of tea.
-
Re: Definately a scrapyard challenge
I can appreciate both sides of the argument. My mistake was getting confused on some of the other sites and deciding to go it alone.....and then finding you guys. This is not a quick project and will take some time. I have many commitments between my work as a landscaper and my wood turning as well as other Things. Need to decide on the z so I can order parts for it. Probably just use a plate with 2 rails I still have some of the 10mm plate left. I will have to order some rails and a short ball screw.
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
2e0poz
The problem is every time you mention 1610 screws it makes it sound like everybody has to rush out and buy them??????
In my opinion 1605 has very few advantages over 1610 for a router, which is why I recommend it so often.
Quote:
Originally Posted by
2e0poz
just remember to when you first started mister and how difficult it was.
When I was designing my router, the rotating ballnut design was far more difficult that picking the ballscrew, but that's just my opinion.
Quote:
Originally Posted by
2e0poz
I use 1605 screws and belts on mine and it is more than enough for it's job, i could have my X axis twice as fast as yours with the belt but it would be pointless.
Even if you manage to get 60m/min, it will only be for a short time before failure. Realistically it's impossible. In terms of speed and acceleration 1610 screws outperform 1605 by a large margin. Even hypothetically running a 1605 at 2:1 is still inferior in performance to a 1610 at 1:1 due to higher K.E.
I started with a 2mm pitch screw on Y and 2.5mm on X since that was the biggest I could afford at the time (not ballscrews), and I made a rotating nut for X to compensate and that funny wonky bearing drive, that I wish there was a simple name for, on Y to lower the friction.
-
Re: Definately a scrapyard challenge
Bruce i apologies for tramping your thread but things like this puts doubt in new comers minds and they then abandon their projects because it gets too expensive. Advice on your Z though is definitely not go for 1610 screw enless you are using high powered steppers. The greater the thread angle the easier it is for the stepper to not be able take the weight. Please do not ignore Jonathan's advice as he has a lot of experience learned first hand. Just stick to your budget and if it is not enough for what you really need then save rather than waste it on buying the first thing that comes along.
A great effort on your build so far
-
Re: Definately a scrapyard challenge
Hey I enjoy seeing people put their theorise to the test. I have a budget but I also like do the best I can. As I already have the screws then thats fine. The upgrade to a water cooled spindle was good advice and still stays , just, in budget. If I cannot afford it then as you say I will save. Biggest problem is deciding in what order to buy stuff.
I will do a sketchup of my gantry redesign and get some coments. probably at the weekend.
Thanks again
Bruce
-
Re: Definately a scrapyard challenge
Right...cough... remember when you were a young engineer you always had to know how things worked. I remember taking a clock apart, and my electric train, my mums iron. Then I got into motorcycles and that was great fun. been doing it for 45 years with varying degrees of success. Always have to find out how they work and maybe put them back together. Well.. oh dear I am not sure how to say this....I was very interested in exactly how the ball screw worked. You guessed it. I unscrewed it and the balls fell out. After spending the last hour trying to rebuild it I realised that the plastic bungs are where they put the balls in and then press hot plastic in to close it off. Is this now an ex ballscrew system or is it fixable? Thankfully it was the shorter one if it has to be replaced.
On that high note of failure goodnight
Bruce (yes i am going to spend the night standing in the corner)
-
Re: Definately a scrapyard challenge
Just to be clear, all my previous comments about 10mm over 5mm pitch were referring only to X and Y. Also a 5mm pitch screw can still backdrive on Z (mine does) if it's aligned well.
I took the ballnut off once just to see what all the fuss was about. I spend a while trying various methods of poking the balls back it which didn't really work. The easiest way (other than buying just a new ballnut) is to get a very strong magnet and stick it to the side of the ballnut. then you can carefully place the balls back in one after the other as they'll stick to the nut inside. Then you machine/find a bar that's just under the pitch diameter of the ballscrew minus the diameter of one ball so it just slides in (doesn't have to be accurate as long as the balls can't fall out). Introduce the ballscrew into the nut allowing it to push out the bar as you turn it on. Only took about 10 mins once I'd found the magnet (2"x1"x0.5" neodymium).
You can also try putting grease on the balls so they stick to the nut inside, but that's not as effective as the magnet. I think you should demagnetise the nut afterwards.
-
Re: Definately a scrapyard challenge
You university boys so darn clever. This is definatly something to tackle at the weekend.
Good night all
bruce
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
Jonathan
In terms of speed and acceleration 1610 screws outperform 1605 by a large margin. Even hypothetically running a 1605 at 2:1 is still inferior in performance to a 1610 at 1:1 due to higher K.E.
Ok can't resist got to jump in.!!!:twisted: . . . Don't need to worry about 5mm pitch being inferior because of K:E bull shit etc because it don't mean jack shit for a DIY machine, it will still perform more than good enough.! . . ( If built good enough.!!)
Yes jonathan's correct when he points out that the calculated physics way of looking at it that 10mm pitch wins for a router!. . . BUT. . . I can tell you because my machine use's 5mm pitch geared 2:1 it's absolutly no problem and in no way limits the performance. Actually IMO it's a bennifit because it's just a simple belt/pulley change to gain resolution.
That said I agree that 10mm pitch (except Z axis then 5mm) would be better choice for a router which mainly cuts wood or plastics. Anything harder IE Ali etc and then extra speed 10mm pitch gains is wasted as the feed rates dictate. 99.9% of the time you'll be lucky to cut any where near quarter the speed 10mm pitch allows.!
An often over looked and misunderstood area which Jonathan touched upon earlier is acceleration.!! . . . IMO and experience for most jobs (Esp 3D) that don't need high feeds (2mtr/min or less) then your much better off tuning your motors to give higher accel than velocity. . . Often less is more.. . Like in the Aztec case where there are lots of short moves it can and does often make job's faster. 3D and v-carving really see the most increase's from high accelleration.
I use Mach3 and it's such a simple job to create profiles that you can easily tailor for specific job types or conditions. IE I have 3D profile tuned for Accelleration. 5mm pitch profile (pulley/belt change)for when resolution needed. I also have profiles for Inch units for each so when I get given code that use's imperial measurements I don't have to fanny about changing settings or code.
Doing it this way makes it very simple and quick to change from one to the other.
So Like 2eOpoz I 100% agree 5mm pitch will be fine in your case.!. . . . BUT . . . 100% Agree with jonathan 10mm pitch rules for a wood/plastic router.
Anyway the screw pitch or speed will be the least of your troubles if you don't beef that gantry up.!!
-
Re: Definately a scrapyard challenge
Ok Ok I promise to do a redesign on the gantry:rolleyes:. I have learnt more in a few days here, mainly how little I know but I am so grateful for all your help.
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
JAZZCNC
Don't need to worry about being inferior because of K:E bull shit
I think you'll find just about everyone apart from you agrees kinetic energy isn't 'bull shit' and is very important, I'm not even going to argue this point because it's so widely accepted. No matter if a machine is made in a home workshop or professionally, it can't escape the laws of physics. A 3nm Nema 34 motor wont go as fast as a 3nm Nema 23, it's due to exactly the same calculation!
Quote:
Originally Posted by
JAZZCNC
my machine use's 5mm pitch geared 2:1 it's absolutly no problem and in no way limits the performance. Actually IMO it's a bennifit because it's just a simple belt/pulley change to gain resolution.
As far as I'm aware you've not tried 10mm pitch rails on your router, so I can't see how you could possibly know that. I'm well aware that you don't respect calculation, and that might well be true for you, but for just about everyone else, calculation plays a hugely important role, and if it didn't work like you say, the theory simply wouldn't be there.
Quote:
Originally Posted by
JAZZCNC
it's just a simple belt/pulley change to gain resolution
That clearly also applies to 1610 so it's not a benefit. Since you only recognise empirical evidence, your version of the Mayan calender I cut on 1:1 looks amazing.
Quote:
Originally Posted by
JAZZCNC
99.9% of the time you'll be lucky to cut any where near quarter the speed 10mm pitch allows.!
It's not about cutting faster, it's the acceleration gain from using 10mm pitch that's most important. You told me your acceleration on the Y-axis is 900mm/s/s, with 5mm pitch. I now have mine set to 2000mm/s/s with 10mm pitch (more shakes the machine too much).
I get the impression you think I'm saying 5mm pitch won't work, seeing as you say "it will work fine". I agree with that, but when designing a machine I think it's foolish to aim for it working fine, you should aim for it to work as well as possible, not just well enough and 5mm pitch screws will simply limit the potential of your machine.
I'm not going to say anything more on the subject since apparently it provokes a particularly strong reaction, I think motoxy has all the info to make an educated decision. Good luck.
Motoxy: There's a quote for that - The more I know, I know, I know the less. (John Owen)
-
Re: Definately a scrapyard challenge
-
Re: Definately a scrapyard challenge
Jonathan . . .My point was that 5mm pitch geared 2:1 works fine because I use it.!. So all your bullshit spouting about K:E means sweet F:A whether nema 23 or 34 because for 99% of DIY use both are far more than good enough.
Also If you look again ya touchy lickle git you'll see I agreed with you on the 10mm pitch.!!
Presumption is the mother of all F:U.!! . . . SO NOW. . How the F~@K do you know what I've done or not done with my machine.? Or any other machine for that matter.:exclaim: . . . . . reading a few old post's don't make you Physic Sally.:mad:
If you want to get into machine design and foolish ness. My approach is that you build a machine to be strong and accurate as possible to do the best it can at it's intended purpose.! .. . To build with just speed in mind at the sacrifice of accurecy, repeatabilty resulting in poor quality in-accurate work.!!. . . is plane stupidity IMO.
Now wasn't having a go at you or saying what you said was wrong or wouldn't work so suck your petty lip back in and drop it.!:tup:
-
Re: Definately a scrapyard challenge
ok but no kissing please:heehee:
-
Re: Definately a scrapyard challenge
I wonder what would happen to lock in Jazz and Jonathan in a big brother house. Make them to make ONE cnc machine.
That would be most interesting BB ever. For a first time I would watch it.
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
wiatroda
I wonder what would happen to lock in Jazz and Jonathan in a big brother house. Make them to make ONE cnc machine.
That would be most interesting BB ever. For a first time I would watch it.
Easy question to answer.!!. . . My bezzy mate and I would agree to disagree then have it built in record time both hopefully learning something new in the process:tup:
-
Re: Definately a scrapyard challenge
For me and I bet for many others both of you are great source of knowledge and experience. You just have totally different approach hence you are great contributions to this forum complementing one another.
Both of you developing one project would create perfect WHOLE :tup:
-
Re: Definately a scrapyard challenge
LMFAO. . . I've got to tell you my son's response to seeing me smiling to my self while writing that.!
" No you wouldn't, You'd test the Moment of inertia of his head with the ballscrews to see if they produced enough Kenetic energy to clear the Big Wall"
Personally I don't know why he'd think that.?? . . . . I didn't know there was Big wall:naughty:
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
JAZZCNC
LMFAO. . .
" No you wouldn't, You'd test the Moment of inertia of his head with the ballscrews to see if they produced enough Kenetic energy to clear the Big Wall"
Personally I don't know why he'd think that.?? . . . . I didn't know there was Big wall:naughty:
:rofl::rofl::rofl: Did he mean Big Wall around Big Brother' house?:rofl:
So you would do empiric aspect of the job , leaving formulas and calculation for Jonathan ??:naughty:
-
Re: Definately a scrapyard challenge
Jonathan - i am having that quote as my signature. Thanks.
As for all the spat Your both absolutely right. Without calculations we would have no idea what could be achieved. But suck it and see can be so much fun. Somewhere in the middle would be where you find success.
Now if its alright with you guys I would like to have my thread back. :wave:
I have thought about my gantry today. Repairing fences can be boring. How about 80 x 40 Al profile as the cross rail with the 10mm thick ally side fixed to cut down version of the 10mm thick angle? I will draw over the weekend.
Bruce
Value frame are cheaper than marchant and dice.
-
Re: Definately a scrapyard challenge
Just been given an old computer for the project. Checked it out....chucked it out. Boy some comps are really old.
Got the balls back in the screw but i will have to make up a guide tomorrow as they push out when you try to screw it together.
I am going to check out my supplier to see what ally box he as available and will then decide what to do about the gantry.
Bruce
-
Re: Definately a scrapyard challenge
Bruce roll up some cereal packet cardboard for now and find something to pack the middle out once inside.
-
Re: Definately a scrapyard challenge
2e0poz you are yet another genius on this site....I have a functioning ball screw. Thanks everyone for your help. Nice to realise I am not the only curious one on this site that justhad to look inside:wink:
Bruce
-
Re: Definately a scrapyard challenge
The main problem with the 150x10mm plate is that that cross section has very poor torsional stiffness. The deflection neglecting twisting is not too bad, though it can easily be improved. Torsional stiffness is very important since the Z-axis acts as a lever applying a large moment to the gantry, so poor torsional stiffness leads to high tool deflection = chatter, poor finish, lower loot life and not being able to cut aluminium well.
The best cross section for torsional stiffness is one where as much material as possible if far from the centroid, that's clearly a tube, and since the stiffness constant is proportional to the radius to the power 4, a small increase in wall thickness or diameter has a big effect. Unfortunately it's rather difficult to attach rails to a curved surface! So the next best thing is to use box section. As a reference, the torsional stiffness of 80x80mm aluminium box section, with just 3mm wall thickness, which is what I use on my machine is almost 40 times greater than 150x10 plate. For forces parallel to X (i.e. orthogonal to the gantry) the stiffness is also significantly greater for that size box section.
If you add two pieces of 50x10mm aluminium to the plate, attaching 10mm surface to the top and bottom of the 150mm surface, the deflection parallel to X is massively reduced (to the point where it's about equal to that for the box section) however the torsional stiffness is still 16 times less than my box section. Add a thin plate on the back (3mm or whatever) and it's a whole lot better. But then you've effectively just make a box section! Adding aluminium profile could work to, but bear in mind the same rules as in this little example apply (look at the cross section of the aluminium profile).
Perhaps you could instead use the 150x10 for the Z-axis (as torsional stiffness there is irrelevant) and add the plates as above to the Z-axis as 10mm is probably not enough on its own.
Hope that helps / gives you some ideas.
Another important factor is the rigidity of the joints between the piece(s) between the gantry, so bolt it well. Doesn't matter how rigid the beam is, if the ends are not connected firmly it will bend and the joints. At the moment you want some more bolts nearer the edge of the aluminium angle, but I guess you've just not got round to putting those in yet.
(Note in the calculations above I didn't take into account the rails themselves. These will add to the strength, but I think relying on the strength of the rails to make the gantry sufficiently rigid is going about it the wrong way. You should get it rigid enough to start with.)
What's the width of the gantry between the supports?
-
Re: Definately a scrapyard challenge
Hi Jonathan
I am beginning to see where your coming from. I have taken on board previous comment about the over hang over the x rails. At the moment their 700mm but I could reduce this to 600. I am going in on monday to see what box is available.I thought about 80 x 40 with the rails top and bottom.
Bruce
-
2 Attachment(s)
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
motoxy
I thought about 80 x 40 with the rails top and bottom.
80x40x3 box section is less than a third as torsionally stiff as an 80x80x3 box section. Same reason as before - material is distributed further from the centroid with a square compared to a rectangle.
These photos show mine (old pictures!) with the size I mentioned. I put 20mm thick plate in the ends of the box section to make as strong joint.
Attachment 5176Attachment 5177
-
Re: Definately a scrapyard challenge
emmm. okay. I have looked on line and 3" x 3" x 10 swg (3.2mm) is easily available. My limitation is no access to a mill. All my work is bench work. But I am sure I can come up with something.
I feel I have come a long way since i started reading the site where someone was building a plywood cnc on his kitchen table. I have t admit that I am glad to be getting so much advice that does make sense.....so far
BTW I have made an offer on the spindle. See if he bites???
Jonathan may I ask what you are studying at uni?
Bruce
closing off see you all tomorrow some time.
-
Re: Definately a scrapyard challenge
See your mum was right? there was reason why she said to eat up your weetabix in the morning:whistling:
-
Re: Definately a scrapyard challenge
Quote:
Originally Posted by
motoxy
Jonathan may I ask what you are studying at uni?
Electrical engineering.
Quote:
Originally Posted by
2e0poz
See your mum was right? there was reason why she said to eat up your weetabix in the morning:whistling:
Er, who was that aimed at - I don't follow?
-
Re: Definately a scrapyard challenge
Cereal packet - Weetabix........Bruce
-
Re: Definately a scrapyard challenge
-
Re: Definately a scrapyard challenge
Just ordered my spindle and inverter. Made an offer of £140 instead of £172 and it was accepted:dance: Other than postage that makes it the same price as a Kress. Bargain.
Now off to redesign the gantry.
Bruce
-
Re: Definately a scrapyard challenge
Yup
Here is the item no. 190618666009
There are more to be had
Bruce
-
3 Attachment(s)
Re: Definately a scrapyard challenge
OK Here is a thought on redesigning the gantry. Any similarity to Jonathan's machine is purely down to plagerism. I have looked at a construction that I can do myself. I will make some enquiries tomorrow about acccess to a m/c shop.
Attachment 5180Attachment 5181Attachment 5182