I tossed up this idea but ended up going the weldment route. Made it out of 100mm x 200mm x5mm tubing with full intention to fill it with epoxy granite. Having it together now I don't think I'll need the eg but the option is there. I'm lucky to have access to large milling machines to machine the milling base but I think you could do it easily without them. I love the idea of the spacer and keeping the dove tails.
Nice work Sam. Commercial machines do something similar by pushing the rail up against a machined edge and then using lots of tiny finger clamps on the other side of the rail to hold it against that reference edge. As you say it will help hold position under heavy loads or minor crashes, but also worth mentioning that it gives the rail more stiffness than just relying on the clamping force of the rail bolts. Should work well so looking forward to the next one.
Great ideas here Sam! I believe I will 'borrow' some of them, as I am currently in the process of rebuilding my ZX45 machine .. Keep up the great work! Gary
Sam, thank you for making these videos, Its truly impressive! Waiting in anticipation for the next episode! Btw, will you be using epoxy aggrigate to fill the castings to make the even stiffer? I bought two of these machines a while back, in a rather dilapidated state, with the idea of rebuilding them, and converting them to cnc, your timing is perfect, as one of them dont have a saddle, and your plan with an inch and a quater plate or 30mm in my part of the world, is the perfect solution! Once again, thank you for taking the time to video your work!
Ian, Your Welcome!, I may do some Epoxy I have not decided yet. It just did not seem right to Hack up the Original castings. There would be voids etc. Steel Plate was the Best solution to me.
Do you think there's any lost of rigidity in the space that lies between X/Y? I had this exact same plan in mind, minus the datum reference points. My only thought was to model it up in a manner that 'sinks' the blocks into the X axis plate to minimize that large gap - which in turn will get you some Z clearance back. The same thing in Z, that void will lose some of your Y axis travel towards the column - right? Or, will it remain with the spindle centerline in the stock position in relation to the Z column surfaces?
I dont think the space will hurt rigidity i do need room for the Ball screw / Ball nut On Z ac=xis i use a head spacer anyway since I extended my Y axis to 10" so that will not be a Issue
After experiencing linear rails on a 96 Bridgeport VMC I have come to the conclusion that beyond rigidity you need axis precision and a high quality spindle with plenty of RPM. From there decent feeds and speeds can make a 200lb mill deliver parts with higher finish quality than a modern day VMC run at a high chip load. I have also found certain rails can be purchased for 1/10th the price of name brand products. Sadly I haven't found any roller type clones so I would suggest going 20mm even on a G0704 as the rails themselves add quite a bit of rigidity as a bonus, so I don't mind giving up a little travel on the Z. My biggest issue at moment is how to finish my X axis and Z axis with rails without a second machine. I think I've come up with a plan to build a temporary new saddle and bed but it's a tricky route you commit to hoping a lot of pieces come together.
Hi Dean, You bring up some good points and I understand these risks. Stick-tion of the Dovetails help add rigidity to these mills with out it the lack of resistance will put all the force on the Ballscrews. I know the motor I have chosen will handle the load it is a 1000 to 1 constant torque inverter duty motor. But I will not be reaching rpm with this spindle of over 5500 rpm. My plan is to keep the machine intact so as to not destroy the castings. I will machine a new Saddle and preserve the original.
I really like your design. I’m just wondering why you need precision ground bars (datums) if you are needing to machine them to a smaller width? The thickness wouldn’t really matter but the with would be more important right? The design is great. I think I’m definitely going to used these ideas with my conversion. Great job
I only needed to machine the Y axis to width the others were stock width. Its nice to start with straight flat stock. I suppose you could use Bar stock but you may need to do some work on it.
how will you access the y axis bearing block screws that are on the inside against the y axis dove tail? Will the inner screws attach from the top? or would you only use two screws per y axis block? I do like the idea on not milling back the dove tails, as you can always go back to stock if it doesn't work out. But if you remove the dove tails, you can fit a significantly larger linear rail. I guess more risk, more reward, for removing the dovetail. Do you think 15mm will be enough? I think the haas mini mill uses 30mm rails, as a frame of reference. I was sizing up my machine last night, and am considering 25mm on all axises. I looked at steal plate, and man is it expensive. But I think at the end of the day it will probably be required over the aluminum. I guess you just need a few people to split a length of it so you can make several saddles out of it.
Lol You saw the mounting dilemma! I will have to build up the x axis first and get them mounted to the saddle. Then assemble the Y axis and just snug the inside bolts so I can move them if needed. Then slide the assembly on the base y rails. Once it is all aligned. I will tighten the outside bolts. Slide the assembly back off and tighten the inside bolts. At least that's the plan will see how it plans out. I scored a couple pieces of steel stock off Ebay so I'm good there and I plan to use 7075 for the Z axis to save weight.
Couldn't you make the saddle wider in the X axis direction. That way the X Axis blocks are outboard of the Y axis blocks, and the outer edge of the y axis dove tails? this way you would have plenty of clearance to tightened all the bolts from the bottom of the saddle. I would think the only down side would be loss of X axis travel. But that shouldn't be too big of an issue, since most bench top machines have more X travel than most ever use. I would think even with the reduced X travel you could still use two 6" vises or three 4" vises.
Yes I did consider that option and as you mentioned there will be a loss of X travel. At the moment I have 18" the Stock mill has 15" and most likely since I will not be able to over extend the X travel as I do now I will be back to around 15". Making the saddle wider might put me around 12" so not something I would like reduce. of course if I have to then that's another story.
12" would be a little small, but not a deal breaker. Do you have any plans to upgrade your spindle? I think once you have the linear rails and servos your cutting feeds cold easily be 150 ipm +. and your rapids would be over 250 ipm. As long as you have the spindle rpm to maintain your chip per tooth. With those speeds I could easily see you getting 4 cubic inch per min removal rate with a 1/4 or 3/8 end mill in aluminum. It is sort of an endless cycle, improve one area, then move on to the next. After that I think the limiting factor is tool pull out with the TTS.
It is a vicious Cycle that is for sure. I will be upgrading the Spindle motor to a 1000 to 1 constant torque motor. Any spindle upgrade will be later down the road. with the Dmm servos rapids will be well over the 300IPM I'm comfortable with. Feed rates will most likely increase to a point where the pucker factor is the limiting issue. But never say Never who knows what the future brings.
Hi, I am also designing a desktop CNC milling machine with working area around 200 X 200 X 40 mm. Here i am facing a problem. However i have completed most of the design. When it comes to force analysis i have no clue how force will act on machine parts during milling process. I belive the forces depends on the material (workpiece), spindle power and speed . But i dont know any formulas to calculate the forces. Could please suggest me any book or material which can lead me to solve this problem.
As long as you do not try and machine more than the Tool and Spindle motor will allow the force applied is minimal. These machines have a tendency to flex when to big of cut is taken. staying within the limits of the machine is the best practice
Well I already had the mill model drawn from my original build I just had to design the new saddles. 2 weeks of making changes checking rethinking deleting redoing. Etc. Lol
Really clever solution with that parallel stock. Was that stuff very pricey? I'd think it'd be worth a good bit to not have to do the machining on those parts!
Why not fill in the Dovetail area with steel bar stock machines to fit in the bevelled area and bolt them in place in a few places then machine those to fit the Linear rails to them.
@@CNC4XR7 all you need is a machine large enough to machine the smaller bar stock to fit in the dovetails to make it all the same height. you might be able to use a grind stone in a hand router to surface the top flat with the other surfaces. a machinist finds way to repair and machine out side the box that how it always been done back before all these NC machine, i worked for Linamar machines in the 1980's and did things to get the parts right that wasn't in the normal machining handbook and i worked on military missile and guns parts with machines that where at that time old manual milling and lathes. we used them to make the prototype PLC machines to build the missile parts where each operation was started by the last operating ending point.
Looking good so far, really like the concept!! .... You mentioned that all you had to do was drill and tap some holes??? I wouldn't rely on any fasteners for positional accuracy on anything; their concentricity is nowhere near good enough. Id drill undersize, then single point precision bore for dowels.. preferably tampered ones for the datums.... you also need access to a machine big enough to machine this in a single setup. Cant you use the holes in the rails as datums or are they not positionally accurate? If so, scrap the whole datum idea and maybe use 3 parallel dowels along each rail... theres enough counterbored holes to sacrifice a few for pins :/
I may not have stated correctly , The datums will be wedged against the stock dovetails then drilled and bolted into position. The rails will be pushed against the datum and bolted into position. Also a bracket will be used to Clamp the rail into position. None of this requires a separate machine to mount which was the whole Idea. I have seen the pins used but the holes will need to be precise once pinned there is no way to make any adjustments.
Maybe put the saddle back on the mill with a mag base and sweep out the horizontal and vertical surfaces you are hoping to use for alignment/ mount the datum... usually only the top bearing surface and the undercut of the dovetails are accurately machined. All other surfaces are only roughed and id be amazed if the horizontal surfaces either side of Y dovetail were even the same height . But I see the issue if you don't have access to a larger machine.. you need to just take the plunge, Good luck!
How do you want to tighten the 8 screws in the middle of the Y axis when assembling? It would be better to turn the cars. The bottom to top and the top to bottom and the Konstrucktion adapt. If you have the cars from the bottom up, you can easily screw in all the screws. You can then first tighten the outer 8 screws firmly and then slide the table to the left .... and firmly tighten the first inner 4 screws then move the table to the other side ...... and tighten the other screws. Good luck
Hi Stephan, Yes it is going to be a Challenge to secure the Blocks I will have to tighten the X Blocks and secure then to the saddle and get them secured then I will be able to tighten the 4 outside Y axis block mounting screws the inner screws will be a challenge I think I have a solution.
Very interesting Sam - Re a steel plate, have you looked into moldset modular plates. Even in NZ there is a big range of ground steel plates so there will be a much bigger range in the USA. eg: www.egwhiter.co.nz/Mild-steel-plate Cliff
Thanks Cliff, I was able to score a good deal on some 1.25" steel plate off Ebay, I also picked up some 7075 for the Z axis to save a little weight. I would have like to use Dura Cast but its just to expensive.
im looking to make a manual milling table using this kind of linear rail. Super helpful
love the intro.
Thank you
Oh yea this is gonna be a awesome build here!!
Yes should be a fun Project!
So cool and love it!
Thank You!
So cool! This is your most impressive project yet! Love it.
Best matt.
Thanks Matt!
Such a cool project. Nice work.
Thanks Russ ,
I tossed up this idea but ended up going the weldment route. Made it out of 100mm x 200mm x5mm tubing with full intention to fill it with epoxy granite. Having it together now I don't think I'll need the eg but the option is there. I'm lucky to have access to large milling machines to machine the milling base but I think you could do it easily without them. I love the idea of the spacer and keeping the dove tails.
Thanks! Manny
I like to keep as much in fact as possible. And design builds with those who may only have the one mill to work with.
Pretty sweet concept.
Thank you! No on to machining.
Same project, i to... Thanks for vidio..
You are welcome
Nice work Sam. Commercial machines do something similar by pushing the rail up against a machined edge and then using lots of tiny finger clamps on the other side of the rail to hold it against that reference edge. As you say it will help hold position under heavy loads or minor crashes, but also worth mentioning that it gives the rail more stiffness than just relying on the clamping force of the rail bolts. Should work well so looking forward to the next one.
Thanks Yes i think that is the Idea exactly!
Great ideas here Sam!
I believe I will 'borrow' some of them, as I am currently in the process of rebuilding my ZX45 machine .. Keep up the great work!
Gary
Thanks! Gary
Go right ahead and you don't even need to return them. Lol
Sam, thank you for making these videos, Its truly impressive!
Waiting in anticipation for the next episode!
Btw, will you be using epoxy aggrigate to fill the castings to make the even stiffer?
I bought two of these machines a while back, in a rather dilapidated state, with the idea of rebuilding them, and converting them to cnc, your timing is perfect, as one of them dont have a saddle, and your plan with an inch and a quater plate or 30mm in my part of the world, is the perfect solution!
Once again, thank you for taking the time to video your work!
Ian,
Your Welcome!, I may do some Epoxy I have not decided yet. It just did not seem right to Hack up the Original castings. There would be voids etc. Steel Plate was the Best solution to me.
Do you think there's any lost of rigidity in the space that lies between X/Y? I had this exact same plan in mind, minus the datum reference points. My only thought was to model it up in a manner that 'sinks' the blocks into the X axis plate to minimize that large gap - which in turn will get you some Z clearance back. The same thing in Z, that void will lose some of your Y axis travel towards the column - right? Or, will it remain with the spindle centerline in the stock position in relation to the Z column surfaces?
I dont think the space will hurt rigidity i do need room for the Ball screw / Ball nut
On Z ac=xis i use a head spacer anyway since I extended my Y axis to 10" so that will not be a Issue
@@CNC4XR7 awesome. I'll be watching for the update!
After experiencing linear rails on a 96 Bridgeport VMC I have come to the conclusion that beyond rigidity you need axis precision and a high quality spindle with plenty of RPM. From there decent feeds and speeds can make a 200lb mill deliver parts with higher finish quality than a modern day VMC run at a high chip load. I have also found certain rails can be purchased for 1/10th the price of name brand products. Sadly I haven't found any roller type clones so I would suggest going 20mm even on a G0704 as the rails themselves add quite a bit of rigidity as a bonus, so I don't mind giving up a little travel on the Z.
My biggest issue at moment is how to finish my X axis and Z axis with rails without a second machine. I think I've come up with a plan to build a temporary new saddle and bed but it's a tricky route you commit to hoping a lot of pieces come together.
Hi Dean,
You bring up some good points and I understand these risks. Stick-tion of the Dovetails help add rigidity to these mills with out it the lack of resistance will put all the force on the Ballscrews. I know the motor I have chosen will handle the load it is a 1000 to 1 constant torque inverter duty motor. But I will not be reaching rpm with this spindle of over 5500 rpm.
My plan is to keep the machine intact so as to not destroy the castings. I will machine a new Saddle and preserve the original.
Will you do the furthermore parallelism checking on the pair of rails against with each other after they were mounted on machine.
Yes lots more checks before I'm done.
I’m thinking about doing this with a Bridgeport
Interesting
I really like your design. I’m just wondering why you need precision ground bars (datums) if you are needing to machine them to a smaller width? The thickness wouldn’t really matter but the with would be more important right? The design is great. I think I’m definitely going to used these ideas with my conversion. Great job
I only needed to machine the Y axis to width the others were stock width. Its nice to start with straight flat stock. I suppose you could use Bar stock but you may need to do some work on it.
how will you access the y axis bearing block screws that are on the inside against the y axis dove tail? Will the inner screws attach from the top? or would you only use two screws per y axis block?
I do like the idea on not milling back the dove tails, as you can always go back to stock if it doesn't work out. But if you remove the dove tails, you can fit a significantly larger linear rail. I guess more risk, more reward, for removing the dovetail. Do you think 15mm will be enough? I think the haas mini mill uses 30mm rails, as a frame of reference. I was sizing up my machine last night, and am considering 25mm on all axises.
I looked at steal plate, and man is it expensive. But I think at the end of the day it will probably be required over the aluminum. I guess you just need a few people to split a length of it so you can make several saddles out of it.
Lol You saw the mounting dilemma! I will have to build up the x axis first and get them mounted to the saddle. Then assemble the Y axis and just snug the inside bolts so I can move them if needed. Then slide the assembly on the base y rails. Once it is all aligned. I will tighten the outside bolts. Slide the assembly back off and tighten the inside bolts. At least that's the plan will see how it plans out. I scored a couple pieces of steel stock off Ebay so I'm good there and I plan to use 7075 for the Z axis to save weight.
Couldn't you make the saddle wider in the X axis direction. That way the X Axis blocks are outboard of the Y axis blocks, and the outer edge of the y axis dove tails? this way you would have plenty of clearance to tightened all the bolts from the bottom of the saddle. I would think the only down side would be loss of X axis travel. But that shouldn't be too big of an issue, since most bench top machines have more X travel than most ever use. I would think even with the reduced X travel you could still use two 6" vises or three 4" vises.
Yes I did consider that option and as you mentioned there will be a loss of X travel. At the moment I have 18" the Stock mill has 15" and most likely since I will not be able to over extend the X travel as I do now I will be back to around 15". Making the saddle wider might put me around 12" so not something I would like reduce. of course if I have to then that's another story.
12" would be a little small, but not a deal breaker. Do you have any plans to upgrade your spindle? I think once you have the linear rails and servos your cutting feeds cold easily be 150 ipm +. and your rapids would be over 250 ipm. As long as you have the spindle rpm to maintain your chip per tooth. With those speeds I could easily see you getting 4 cubic inch per min removal rate with a 1/4 or 3/8 end mill in aluminum.
It is sort of an endless cycle, improve one area, then move on to the next. After that I think the limiting factor is tool pull out with the TTS.
It is a vicious Cycle that is for sure. I will be upgrading the Spindle motor to a 1000 to 1 constant torque motor. Any spindle upgrade will be later down the road. with the Dmm servos rapids will be well over the 300IPM I'm comfortable with.
Feed rates will most likely increase to a point where the pucker factor is the limiting issue.
But never say Never who knows what the future brings.
Wow Sam, that's going to be nice! What kind of rapids are you going for?
Thanks
Well as far as Rapids I imagine it's going to be faster than my nerves can take.lol most likely I will limit it to around 300ipm but will see.
Hi, I am also designing a desktop CNC milling machine with working area around 200 X 200 X 40 mm.
Here i am facing a problem. However i have completed most of the design. When it comes to force analysis i have no clue how force will act on machine parts during milling process.
I belive the forces depends on the material (workpiece), spindle power and speed . But i dont know any formulas to calculate the forces. Could please suggest me any book or material which can lead me to solve this problem.
As long as you do not try and machine more than the Tool and Spindle motor will allow the force applied is minimal. These machines have a tendency to flex when to big of cut is taken. staying within the limits of the machine is the best practice
As always great video Sam, would you consider doing this on a lathe conversion?
Yes!
How long did it take to draw it up?
Well I already had the mill model drawn from my original build I just had to design the new saddles. 2 weeks of making changes checking rethinking deleting redoing. Etc. Lol
Really clever solution with that parallel stock. Was that stuff very pricey? I'd think it'd be worth a good bit to not have to do the machining on those parts!
3ft X and Z bars are $22 each Y 18" bars were $17 each
MacMaster Carr
X Z 9516K46
Y 9516K48
Why not fill in the Dovetail area with steel bar stock machines to fit in the bevelled area and bolt them in place in a few places then machine those to fit the Linear rails to them.
That would work well however you need a machine large enough to do the machining.
@@CNC4XR7 all you need is a machine large enough to machine the smaller bar stock to fit in the dovetails to make it all the same height. you might be able to use a grind stone in a hand router to surface the top flat with the other surfaces. a machinist finds way to repair and machine out side the box that how it always been done back before all these NC machine, i worked for Linamar machines in the 1980's and did things to get the parts right that wasn't in the normal machining handbook and i worked on military missile and guns parts with machines that where at that time old manual milling and lathes. we used them to make the prototype PLC machines to build the missile parts where each operation was started by the last operating ending point.
Looking good so far, really like the concept!! .... You mentioned that all you had to do was drill and tap some holes??? I wouldn't rely on any fasteners for positional accuracy on anything; their concentricity is nowhere near good enough. Id drill undersize, then single point precision bore for dowels.. preferably tampered ones for the datums.... you also need access to a machine big enough to machine this in a single setup. Cant you use the holes in the rails as datums or are they not positionally accurate? If so, scrap the whole datum idea and maybe use 3 parallel dowels along each rail... theres enough counterbored holes to sacrifice a few for pins :/
I may not have stated correctly , The datums will be wedged against the stock dovetails then drilled and bolted into position. The rails will be pushed against the datum and bolted into position. Also a bracket will be used to Clamp the rail into position. None of this requires a separate machine to mount which was the whole Idea. I have seen the pins used but the holes will need to be precise once pinned there is no way to make any adjustments.
Maybe put the saddle back on the mill with a mag base and sweep out the horizontal and vertical surfaces you are hoping to use for alignment/ mount the datum... usually only the top bearing surface and the undercut of the dovetails are accurately machined. All other surfaces are only roughed and id be amazed if the horizontal surfaces either side of Y dovetail were even the same height . But I see the issue if you don't have access to a larger machine.. you need to just take the plunge, Good luck!
How do you want to tighten the 8 screws in the middle of the Y axis when assembling? It would be better to turn the cars. The bottom to top and the top to bottom and the Konstrucktion adapt. If you have the cars from the bottom up, you can easily screw in all the screws. You can then first tighten the outer 8 screws firmly and then slide the table to the left .... and firmly tighten the first inner 4 screws then move the table to the other side ...... and tighten the other screws.
Good luck
Hi Stephan,
Yes it is going to be a Challenge to secure the Blocks I will have to tighten the X Blocks and secure then to the saddle and get them secured then I will be able to tighten the 4 outside Y axis block mounting screws the inner screws will be a challenge I think I have a solution.
Very interesting Sam - Re a steel plate, have you looked into moldset modular plates. Even in NZ there is a big range of ground steel plates so there will be a much bigger range in the USA. eg: www.egwhiter.co.nz/Mild-steel-plate
Cliff
Thanks Cliff,
I was able to score a good deal on some 1.25" steel plate off Ebay, I also picked up some 7075 for the Z axis to save a little weight. I would have like to use Dura Cast but its just to expensive.
Nice video, but talking is a little bit slow, speeding up to 1.5x is the right pace. Maybe it's just me.
Yeah thanks something happens in the editing I think