30"+ work space low cost diy CNC router parts kits?

[Goemon]

If built right, they won't.
But everything flexes to some degree. as long as you can keep the flex to an acceptable amount, then you'll be fine.

Have you ever sat in a car on an overpass, and feel the bridge beneath you bounce up and down? Everything flexes, no matter how strong and stiff it is.

Those bridges are designed to move. If they had no flex then they would break. I know what you meant though.

One of the reasons I like working with the carbon fiber is that it is so rigid. Nothing else comes close without a huge increase in weight. It's also easier to make it flat than it would be with steel with the tools I have here. Plus, if it is not rigid enough, I can keep adding additional layers until it is.

[louieatienza]

Given my general lack of engineering knowledge, it sounds like I am better off ordering the extra ball screw and motor. Basically, 4 motor drivers like the Gecko G540 are 3 axis devices, not 4 as advertised, unless I switch to a fixed gantry design.

I still don't understand how one side of bearing blocks gets ahead of the other on the THK rails if they are connected with horizontal steel reinforced carbon fiber plates but maybe I don't need to fully understand every last bit if I want the machine to get built while I'm still alive.

For the size of your machine, a fixed gantry may be the better choice, for the obvious reason that it's fixed. If you're going for cost-effectiveness/strength, you'll blow your budget with CF. While steel is heavier, generally weight will increase the resonant frequency of the machine, which should lead to less vibration. If this were not the case, VMC builders would make their machines with CF mostly, because they could use smaller servos and linear bearings.

While the Steel reinforced CF plate my be stiff, it still would be subjected to torsion/twist when you are cutting at the extremities of the cutting area. You would mitigate that by increasing the distance between your bearing blocks on the gantry legs, which would increase the width of your transverse plate. But then you'd also need to increase the length of that axis as well to have the same cutting area. Also, all THK profile rail/block are not equal. They have an interchangeable type, which has almost no preload (no designation), light preload (C1), or medium preload (C0). Obviously, the C0 would give you less gantry leg "translation", but that will cost you in terms of the extra torque needed to move that axis (in addition of the torque needed to overcome the friction of the anti-backlash leadnut, weight of the axis, and even detent torque of the stepper). Of course the slower you cut, the more torque you'll have available, but if your plan is to mill aluminum you'll need some speed to achieve a good finish.

My lead screws are half inch 2 start for all three axis. I'm not sure if I should have gone for 5 - start but either way, it's done now and I don't care if the machine is slower. It's not a particularly large machine anyway. My main concern is achieving a quality finish, even if it's super slow.

I'm going with those 381ozlow inductance motors. In addition to asking here, I asked multiple CNC stores which ones they recommended for my needs and they all recommended the ones they sold.

The only thing nagging the back of my mind on those motors is that the CNC mill motors kits all seem to have 8x the amount of holding torque. I understand that those things are meant for cutting harder metals like steel but it indicates that more torque is better for harder materials. I would like to try making aluminum molds at some point so I am hoping that having 381oz motors doesn't become the thing that stops me.

Anyway....time to stop stressing about stepper motors and time to start stressing about my choice of spindle. On to the next sea of indistinguishable choices I don't fully understand!

Those larger torque numbers is more an indicator that the axes on these mills are a LOT heavier, and you need to also overcome the stiction that comes with dovetail ways. And yes, milling harder materials with lower-speed high-torque spindles requires more holding power. Aluminum on the other hand cuts relatively easily. I've had success cutting aluminum with an even wimpier setup - 425in-oz steppers running 1/2"-8, 8 start leadscrews and G540 at 48V, and still cut aluminum at 60ipm with good finish. With a wood frame CNC. Almost 8 years ago! Knowing how to work your machine to the limit, and then backing off for the finish cut, helps. Also, I discovered the use of high-speed (constant-engagement) toolpaths actually benefit our more flex-prone DIY routers. I bought in at the time to a mid-level CAM package early on, but nowadays, Fusion360 makes this available to everyone.

So I wouldn't worry about the steppers, as long as you have a spindle with decent power. The limits to your surface finish are going to be with your ACME screws, because of the way they're made. Unless you have a precision ground set, they'll have very slight fluctuations in lead and stiction that will translate to your work. The acetal AB nuts help smooth that some, but they're no match to even cheap ballscrews.

One of the reasons I like working with the carbon fiber is that it is so rigid. Nothing else comes close without a huge increase in weight. It's also easier to make it flat than it would be with steel with the tools I have here. Plus, if it is not rigid enough, I can keep adding additional layers until it is.

Again, weight used judiciously can be your friend. Here, because your axes are relatively small, you don't have to worry about making it heavy; I doubt that it will be too heavy for your motors to move. Again, I used 425in-oz steppers with a G540 at 48V. Those steppers are capable of higher speed with more voltage, and I don't have the mechanical advantage of a finer leadscrew pitch (I basically had 1tpi or 1" lead leadscrews.) Still, I moved an 85lb+ gantry with two screws and two steppers, at 750ipm, and cut at 200ipm. And I was able still to resolve fine detail like 1/64" serifs on raised lettering. (I did scale it back to about 400ipm rapids however.) Good enough for me. CarveOne made a large 4 x 10 table moving a 250lb gantry with two NEMA23 steppers, at high speed. Gives you a little respect for the kind of forces these small motors generate...

[wizard]

Those bridges are designed to move. If they had no flex then they would break. I know what you meant though.

I think the point he was trying to make is that no matter what you build you machine out of it will flex.

One of the reasons I like working with the carbon fiber is that it is so rigid.

Carbon fiber fishing pole??

The point here is thé carbon fiber by itself isn't extremely rigid, it has to be woven into materials that are eventually saturated with some sort of glue. Those materials are then formed into structures that can be very stiff. Those structures still have to be designed though.

Nothing else comes close without a huge increase in weight.

I see a lot of people with concerned about weight in the various forums I visit. It can certainly be a problem if your goal is portable machinery but for the most part that isn't what most people are looking for. Very few machines are being built in this forum where weight is an overwhelming negative, in general most machines are to light. Often they are light simply due to economics, that is the builder can't afford to build out of the heavier materials that would benefit the machine.

It's also easier to make it flat than it would be with steel with the tools I have here. Plus, if it is not rigid enough, I can keep adding additional layers until it is.

Obviously you have some experience with carbon fiber here, most of us don't have such experience. The general opinion though is that carbon fiber is very expensive relative to the competing materials. Maybe that opinion is wrong, obviously if I haven't worked with the materials I can't say. What I do know is that even after molding up carbon fiber or even fiberglass you don't have a net shape part so you still need to machine the material some how.

So if you can pull off a carbon fiber based router design I'd really like to see it! I especially would like to see the methods used in building it and realistic test of its performance. I don't have the $$ to throw at an experiment but would be very interested in the results of such a build.

[Goemon]

For the size of your machine, a fixed gantry may be the better choice, for the obvious reason that it's fixed. If you're going for cost-effectiveness/strength, you'll blow your budget with CF. While steel is heavier, generally weight will increase the resonant frequency of the machine, which should lead to less vibration. If this were not the case, VMC builders would make their machines with CF mostly, because they could use smaller servos and linear bearings.

While the Steel reinforced CF plate my be stiff, it still would be subjected to torsion/twist when you are cutting at the extremities of the cutting area. You would mitigate that by increasing the distance between your bearing blocks on the gantry legs, which would increase the width of your transverse plate. But then you'd also need to increase the length of that axis as well to have the same cutting area. Also, all THK profile rail/block are not equal. They have an interchangeable type, which has almost no preload (no designation), light preload (C1), or medium preload (C0). Obviously, the C0 would give you less gantry leg "translation", but that will cost you in terms of the extra torque needed to move that axis (in addition of the torque needed to overcome the friction of the anti-backlash leadnut, weight of the axis, and even detent torque of the stepper). Of course the slower you cut, the more torque you'll have available, but if your plan is to mill aluminum you'll need some speed to achieve a good finish.



Those larger torque numbers is more an indicator that the axes on these mills are a LOT heavier, and you need to also overcome the stiction that comes with dovetail ways. And yes, milling harder materials with lower-speed high-torque spindles requires more holding power. Aluminum on the other hand cuts relatively easily. I've had success cutting aluminum with an even wimpier setup - 425in-oz steppers running 1/2"-8, 8 start leadscrews and G540 at 48V, and still cut aluminum at 60ipm with good finish. With a wood frame CNC. Almost 8 years ago! Knowing how to work your machine to the limit, and then backing off for the finish cut, helps. Also, I discovered the use of high-speed (constant-engagement) toolpaths actually benefit our more flex-prone DIY routers. I bought in at the time to a mid-level CAM package early on, but nowadays, Fusion360 makes this available to everyone.

So I wouldn't worry about the steppers, as long as you have a spindle with decent power. The limits to your surface finish are going to be with your ACME screws, because of the way they're made. Unless you have a precision ground set, they'll have very slight fluctuations in lead and stiction that will translate to your work. The acetal AB nuts help smooth that some, but they're no match to even cheap ballscrews.



Again, weight used judiciously can be your friend. Here, because your axes are relatively small, you don't have to worry about making it heavy; I doubt that it will be too heavy for your motors to move. Again, I used 425in-oz steppers with a G540 at 48V. Those steppers are capable of higher speed with more voltage, and I don't have the mechanical advantage of a finer leadscrew pitch (I basically had 1tpi or 1" lead leadscrews.) Still, I moved an 85lb+ gantry with two screws and two steppers, at 750ipm, and cut at 200ipm. And I was able still to resolve fine detail like 1/64" serifs on raised lettering. (I did scale it back to about 400ipm rapids however.) Good enough for me. CarveOne made a large 4 x 10 table moving a 250lb gantry with two NEMA23 steppers, at high speed. Gives you a little respect for the kind of forces these small motors generate...

I guess that some of the process process of learning the limits of my machine so I can set it accordingly will come from experience based on trial and error.

Please can you explain the potential disadvantage you described with using carbon fiber? I don't understand why you are saying it will effect the width of my machine or cutting area, or why it would be subject to additional twist or torsion (that wouldn't be present if I used plain steel or aluminum). If made right, carbon fiber is stiffer and stronger than steel, aluminum or titanium. It has less heat expansion / distortion too.

The reason for me using some steel sandwiched in my carbon fiber gantry plates is only to allow me to use screws to attach them to my bearing blocks. Carbon fiber doesn't hold threads or screws very well. I use the same method (give or take) when I make carbon fiber rifle stocks with aluminum pillars.

I am am not worried about the cost of the carbon fiber. I am using materials I had here anyway (making carbon fiber parts is my business). I make all the plates and tubes myself and I am set up to work with it here. I am not really set up to work with steel. I have no tools capable of cutting or shaping it well and no welding skills etc. it's more a case of working with what I am comfortable with. I assume that a lot of people use wood for the same reason (comfort in working with it).

Btw, these guys sell what looks to be a rather large fixed gantry CNC machine with a carbon fiber gantry bridge. They claim it is better for all the usual reasons but.... then again... they would say that given that they probably want to sell a few...

[Goemon]

I think the point he was trying to make is that no matter what you build you machine out of it will flex.

Carbon fiber fishing pole??

The point here is thé carbon fiber by itself isn't extremely rigid, it has to be woven into materials that are eventually saturated with some sort of glue. Those materials are then formed into structures that can be very stiff. Those structures still have to be designed though.

I see a lot of people with concerned about weight in the various forums I visit. It can certainly be a problem if your goal is portable machinery but for the most part that isn't what most people are looking for. Very few machines are being built in this forum where weight is an overwhelming negative, in general most machines are to light. Often they are light simply due to economics, that is the builder can't afford to build out of the heavier materials that would benefit the machine.


Obviously you have some experience with carbon fiber here, most of us don't have such experience. The general opinion though is that carbon fiber is very expensive relative to the competing materials. Maybe that opinion is wrong, obviously if I haven't worked with the materials I can't say. What I do know is that even after molding up carbon fiber or even fiberglass you don't have a net shape part so you still need to machine the material some how.

So if you can pull off a carbon fiber based router design I'd really like to see it! I especially would like to see the methods used in building it and realistic test of its performance. I don't have the $$ to throw at an experiment but would be very interested in the results of such a build.

Compared to some of the carbon fiber parts I make, gantry plates are relatively easy. The typical method of making anything with woven carbon fiber is to wet it out with epoxy resin and lay it in a mold in multiple layers. What you pull out is pretty close to the final part (aside from some trimming and polishing. It isn't like wood or metals where you have the opportunity for further milling and shaping. If it doesn't come out of the mold in the right shape, you can't change it's shape much afterwards by sanding or grinding.

Most of the difficulty in making cf parts is in making a good mold. After that, it's all down hill. The reason why gantry plates are relatively easy is because they are flat. I don't even need to make a mold for the sides of the gantry if I wanted to replicate the designs I have seen others use with aluminum.

For example, for the the gantry bridge, which has to be super flat to hold the rails, I am using two optically flat mirrors as my mold. The cf fabric is layered up in multiple directions (for strength) and sandwiched between the two mirror surfaces under 300lb or pressure while it is cured in my diy curing oven at a high temperature (same as I would make any flat cf plate).

The plates I pull from the mirrors will be flatter than most people could ever achieve with wood or steel.

I plan to use a slightly different design for my gantry side plates though. Cf parts are more rigid if they are not flat. I am working on my plugs now so I can make a mold for them.

Carbon fiber parts are expensive if you get someone else to make them for you but the raw materials to make your own are not too bad. You can buy the fabric off 50" rolls for as little as $10 a yard these days.

Compared to the prices I have been given for decent steel or aluminum plates, the costs won't end up being too different.

I have no interest in reducing the weight to make it portable. It's more that everything I have read indicates that too much weight will slow it down with a moving gantry design. Using cf seems like an obvious way of making it stronger without worrying about stalling my little motors or reducing performance.

Anyway, when I'm done, (if it works out) I will happily share my methods if there is any interest. It's the least I can do given all the help people have given me. I may not know anything about building CNC machines yet but carbon fiber, that I do know.

[Goemon]

Is there any reason why using two 7.5a 48v PSU's instead of one 12.5a 48v psu would be a problem for me?

The reason I ask is because it's a lot cheaper to buy two 7.5a 48v PSU's than it is to buy one 12.5a one. I am assuming it's ok given that each motors only draws 3.5a or less but I wanted to check just in case....

[louieatienza]

Please can you explain the potential disadvantage you described with using carbon fiber? I don't understand why you are saying it will effect the width of my machine or cutting area, or why it would be subject to additional twist or torsion (that wouldn't be present if I used plain steel or aluminum). If made right, carbon fiber is stiffer and stronger than steel, aluminum or titanium. It has less heat expansion / distortion too.

The reason for me using some steel sandwiched in my carbon fiber gantry plates is only to allow me to use screws to attach them to my bearing blocks. Carbon fiber doesn't hold threads or screws very well. I use the same method (give or take) when I make carbon fiber rifle stocks with aluminum pillars.

I am am not worried about the cost of the carbon fiber. I am using materials I had here anyway (making carbon fiber parts is my business). I make all the plates and tubes myself and I am set up to work with it here. I am not really set up to work with steel. I have no tools capable of cutting or shaping it well and no welding skills etc. it's more a case of working with what I am comfortable with. I assume that a lot of people use wood for the same reason (comfort in working with it).

Btw, these guys sell what looks to be a rather large fixed gantry CNC machine with a carbon fiber gantry bridge. They claim it is better for all the usual reasons but.... then again... they would say that given that they probably want to sell a few...

On the first point, I never said that the CF would effect the width of the cutting area. Having a moving gantry using a center-mounted screw can cause the gantry to rack when cutting at the extremities. Doesn't matter what material you use. The way that is mitigated is by increasing the width between the gantry upright bearing blocks, which in effect will necessitate making the axis larger for a given cut area.

With regards to fasteners, I've built a few machines using paper, cotton, and fiberglass based phenolic sheet, successfully using inserts for smaller screws and even threading directly for larger screws, with no failure.

I hear you on using what you have on hand. But your thread title is "30"+ work space low cost diy CNC router parts kits?" So even if you have it in stock, it was purchased at some point. Maybe 8-10 years ago wood frame, plywood, MDF routers were pretty abundant because it is cheap, readily available, and relatively easy to work. Even then, the machine was more of a stepping stone toward aluminum extrusion and steel; and the skate bearings, angles, unsupported black pipe, and ACME screws replaced with supported round rail or profile rail, and ballscrews. It's a misnomer that a steel build requires welding and heavy equipment. My first machine was made of hardwood. I then made a few out of phenolic, and a few of aluminum extrusion. I am a cabinetmaker and woodworker by trade, but there was no fear in working with unfamiliar materials to me like aluminum. Sometimes you have to get out of the comfort zone to achieve the results you desire.

I believe DMG makes or made a VMC that has CF components. I personally don't envision how this works in a much smaller scale. If your goal is to eventually make aluminum molds, check out the commercial machines out there, and see how they're constructed and why they're made the way they are. I also build guitars. The plate assemblies with bracing have to be stiff, but light. This is because I'm trying to achieve a low resonant frequency to effect a fuller bass sound; I want the plate to vibrate. It's the opposite with a CNC; I want as little vibration as possible, so I use the densest materials I can. It's why mills are made with cast iron. Not saying that a CF machine won't be fine, just that I don't believe that the extra labor and time and cost (time is money) will give you a machine that is just as good, if not better, for a lot cheaper, and faster.

There was an older gentleman that used to post here. Loved to work with aluminum plate. So much so, that he spent over $3000 on the stuff, and built beautiful frame components. Bought huge NEMA34 steppers. Which caused him to cheap out on other stuff, like using 1/2"-10 single start ACME and a wimpy 24V PSU. Had no money for even a basic CAM package, and was using CorelDraw and Mach3's DXF converter. Then he was wondering why he couldn't move faster than 40ipm... got plain acetal leadnuts, and when he tried to rapid, the acetal would swell from the heat from friction, causing the motors to stall, because the PSU was limiting the amount of torque he had at higher RPMs. If I had the money at the time, I would have bought him a copy of Cut2D I felt so bad.

[wizard]

Compared to some of the carbon fiber parts I make, gantry plates are relatively easy. The typical method of making anything with woven carbon fiber is to wet it out with epoxy resin and lay it in a mold in multiple layers. What you pull out is pretty close to the final part (aside from some trimming and polishing. It isn't like wood or metals where you have the opportunity for further milling and shaping. If it doesn't come out of the mold in the right shape, you can't change it's shape much afterwards by sanding or grinding.

Sounds a lot like boat building out of fiberglass.

The problem with a machine tool component is that it needs to be pretty accurate for flatness and squareness of various surfaces. I can't see a molded component holding tolerance and not require some triming or machining. Then again the tis what I know of boat building.

Also you seem to be thinking in terms of plates, not beams. Simple plates just aren't that stiff no matter what they are made out of.

Most of the difficulty in making cf parts is in making a good mold. After that, it's all down hill. The reason why gantry plates are relatively easy is because they are flat. I don't even need to make a mold for the sides of the gantry if I wanted to replicate the designs I have seen others use with aluminum.

I suppose you could laminate up hundreds of layers of fabric to get plates that are stiff and suitable for a machine tool. I'm still thinking in terms of beams and box sections for economy and strength.

For example, for the the gantry bridge, which has to be super flat to hold the rails, I am using two optically flat mirrors as my mold. The cf fabric is layered up in multiple directions (for strength) and sandwiched between the two mirror surfaces under 300lb or pressure while it is cured in my diy curing oven at a high temperature (same as I would make any flat cf plate).

The problem here is are the mirrors stiff enough not to deflect or is the frame they sit on stiff enough?

As for the curing oven any info you have on the resins and oven design would likely be of interest to many people in these forums.

The plates I pull from the mirrors will be flatter than most people could ever achieve with wood or steel.

I'd like to see that.

I plan to use a slightly different design for my gantry side plates though. Cf parts are more rigid if they are not flat. I am working on my plugs now so I can make a mold for them.

Almost anything is more rigid if it isn't flat, that is why the use of tubing is highly suggested for machine builds out of metal and the use of torsion boxes (beams) are suggested for wood.

Carbon fiber parts are expensive if you get someone else to make them for you but the raw materials to make your own are not too bad. You can buy the fabric off 50" rolls for as little as $10 a yard these days.

That is actually surprising, carbon fiber use to be horribly expensive. Still you will need hundreds of yards to build you machine. At least that is what I imagine as you wouldn't want cross sectionals any thinner that 1/2" anywhere in the machine.

Compared to the prices I have been given for decent steel or aluminum plates, the costs won't end up being too different.

Aluminum is pretty expensive purchased new. Steel can be considerably cheaper. The real value in steels though is to buy drops and used which can dramatically lower costs.

I have no interest in reducing the weight to make it portable. It's more that everything I have read indicates that too much weight will slow it down with a moving gantry design. Using cf seems like an obvious way of making it stronger without worrying about stalling my little motors or reducing performance.

You may be getting overly concerned about speed. First when machining the speed or feed rates are determined by the materials, cutter and spindle capacity. In other words there is an ideal feed rate for a specific operation and those feed rates seldom reach the speed of your rapids.

As for rapids there comes a point where rapids become a safety problem. I wouldn't be too concerned about achieving amazing rapid sunless you want to totally enclose the machine. I've seen and have even worked on devices that can achieve really fast movements and it can be very unnerving. Also depending upon your type of work, the quality of your CAD/CAM and other factors rapids might no make a lot of difference in the over all cycle times. Often you rapid move velocities look like pyramids as max speed is never achieved.

In any event I'm not saying that weight isn't a factor but with most low end machines it is hardly a problem.

Anyway, when I'm done, (if it works out) I will happily share my methods if there is any interest.

I'm very interested to see what you come up with. Your methods, materials, suppliers and so forth are all of interest.

It's the least I can do given all the help people have given me. I may not know anything about building CNC machines yet but carbon fiber, that I do know.

You certainly leave the impression you have lots of experience with carbon fiber! For us out here with zero experience we would love to see this project go forward even if we believe the use of carbon fiber isn't required. Anything that leads to a low cost way to achieve good performance from a DIY machine is of interest.

From the machine design standpoint all you really need to realize is that stiffness is more important than anything else really. Weight is actually a good thing in many cases, a machine that is too light will walk all over the place as it reacts to accelerations from the axises. Vibration is also an issue when it comes to quality finish passes. If you can design out the tendency of the structures to vibrate then you have the workings of a decent machine.

[wizard]

Is there any reason why using two 7.5a 48v PSU's instead of one 12.5a 48v psu would be a problem for me?

The reason I ask is because it's a lot cheaper to buy two 7.5a 48v PSU's than it is to buy one 12.5a one. I am assuming it's ok given that each motors only draws 3.5a or less but I wanted to check just in case....

For motor drivers, it shouldn't make a difference. Run one power supply to two drivers. A lot of commercial drivers take 120VAC in and in effect have a power supply per driver. As for driver current draw do check with the manufactures data sheets as micro stepping impacts current draw and many drivers have controls to fold back current when idle. Generally I like to see current capacity on the plus side of what we add up as simple motor current ratings.

[louieatienza]

Is there any reason why using two 7.5a 48v PSU's instead of one 12.5a 48v psu would be a problem for me?

The reason I ask is because it's a lot cheaper to buy two 7.5a 48v PSU's than it is to buy one 12.5a one. I am assuming it's ok given that each motors only draws 3.5a or less but I wanted to check just in case....

I don't believe it matters. That said you could be fine with just one 10A 48V PSU. This is because most stepper drives are chopper drives; to keep the current constant the drive will cut or "chop" voltage to the coils when it reaches the current set on the drive, and reapply voltage when the current drops in the coil.

[Goemon]

Sounds a lot like boat building out of fiberglass.

The problem with a machine tool component is that it needs to be pretty accurate for flatness and squareness of various surfaces. I can't see a molded component holding tolerance and not require some triming or machining. Then again the tis what I know of boat building.

Also you seem to be thinking in terms of plates, not beams. Simple plates just aren't that stiff no matter what they are made out of.

I suppose you could laminate up hundreds of layers of fabric to get plates that are stiff and suitable for a machine tool. I'm still thinking in terms of beams and box sections for economy and strength.

The problem here is are the mirrors stiff enough not to deflect or is the frame they sit on stiff enough?

As for the curing oven any info you have on the resins and oven design would likely be of interest to many people in these forums.

I'd like to see that.

Almost anything is more rigid if it isn't flat, that is why the use of tubing is highly suggested for machine builds out of metal and the use of torsion boxes (beams) are suggested for wood.

That is actually surprising, carbon fiber use to be horribly expensive. Still you will need hundreds of yards to build you machine. At least that is what I imagine as you wouldn't want cross sectionals any thinner that 1/2" anywhere in the machine.

Aluminum is pretty expensive purchased new. Steel can be considerably cheaper. The real value in steels though is to buy drops and used which can dramatically lower costs.

You may be getting overly concerned about speed. First when machining the speed or feed rates are determined by the materials, cutter and spindle capacity. In other words there is an ideal feed rate for a specific operation and those feed rates seldom reach the speed of your rapids.

As for rapids there comes a point where rapids become a safety problem. I wouldn't be too concerned about achieving amazing rapid sunless you want to totally enclose the machine. I've seen and have even worked on devices that can achieve really fast movements and it can be very unnerving. Also depending upon your type of work, the quality of your CAD/CAM and other factors rapids might no make a lot of difference in the over all cycle times. Often you rapid move velocities look like pyramids as max speed is never achieved.

In any event I'm not saying that weight isn't a factor but with most low end machines it is hardly a problem.

I'm very interested to see what you come up with. Your methods, materials, suppliers and so forth are all of interest.


You certainly leave the impression you have lots of experience with carbon fiber! For us out here with zero experience we would love to see this project go forward even if we believe the use of carbon fiber isn't required. Anything that leads to a low cost way to achieve good performance from a DIY machine is of interest.

From the machine design standpoint all you really need to realize is that stiffness is more important than anything else really. Weight is actually a good thing in many cases, a machine that is too light will walk all over the place as it reacts to accelerations from the axises. Vibration is also an issue when it comes to quality finish passes. If you can design out the tendency of the structures to vibrate then you have the workings of a decent machine.

The process of making carbon fiber parts is not massively different from working with fiberglass but the properties of the material and the level of difficulty involved is very different. Carbon fiber parts are lot stiffer than fiberglass and it won't take anywhere close to hundreds of layers to make it rigid enough for the size of plates I am making.

With 10 layers of properly made carbon fiber it is usually so stiff that it can't be visibly bent by hand. You can drive a 2 ton vehicle over some carbon fiber rifle stocks without causing any damage. I seriously doubt I'll need any more than 5 yards of 50" fabric.

I am going to make rectangular tubes for the gantry side plate and probable the same for the bridge. In addition to adding strength with less material, it also offer the opportunity to add a filling to reduce vibrations. I sometimes use urethane foam for the same purpose when making stocks.

I keep hearing how steel is a cheap option but I can't find anywhere that sells decent size pieces of quality steel for a low price. Where should I look to find cheap steel for sale? And what are people calling cheap? How much would you expect to pay for 1/2" thick 8"x24" or 12"x24" slab to use as a gantry bridge?

[wednesday]

I am going to make rectangular tubes for the gantry side plate and probable the same for the bridge. In addition to adding strength with less material, it also offer the opportunity to add a filling to reduce vibrations. I sometimes use urethane foam for the same purpose when making stocks.

You could always use ready made fiberglass structural U channel as your core build up material. Make an ID carbon tube and glass the U's together around it then wrap the outside with as much CF as you felt necessary. Be sure to imbed some phenolics for screw retention as CF doesn't tap well IMO. Garolite xx and ce take mechanical fasteners well. G-10 not as well. Hollow tube framing is very useful for running wires cleanly but be sure to properly isolate any holes that the wiring will pass through.

Also be very diligent to account for proper grounding of your finished frame. CF is highly conductive and that could be problematic around the electronics you will be using to finish out the whole. If you plan on adding any kind of vacuum hold down setup and/or dust collection the issue is tripled. We acquired a copper mesh material from a defunct blimp builder a few years ago. They used it as a layer inside the carbon laminates to protect the parts from lightning strikes. It could be adapted for a kick ass grounding set up.

[ger21]

How much would you expect to pay for 1/2" thick 8"x24" or 12"x24" slab to use as a gantry bridge?

You don't use steel plates, you use tubes. A 1/2" steel plate wouldn't be rigid enough.

[Goemon]

You don't use steel plates, you use tubes. A 1/2" steel plate wouldn't be rigid enough.

I have some steel steel and aluminum tubes here that I could use but I would still need a flat surface to mount the rails on wouldn't I?

everywhere I have looked for large steel plates has been very expensive. We are talking $250 -$350 per piece before shipping.

I don't know if people have outdated views or if I am looking in the wrong places but as far as I can tell, steel is the most expensive option. It's a lot more expensive than even a top quality T6 aluminum and far more than any composite solution.

[ger21]

You have to make sure you're not comparing apples and oranges.
As I said, nobody is using steel plates.

If you compare rectangular or square tubing, it looks like the cost between aluminum and steel is similar, but steel is much stronger. When you compare equal strength, steel comes out cheaper, as you need a much larger aluminum tube to get equal strength.

Most people make the comparison between steel tubing and aluminum extrusions, where steel is far cheaper.
You also need to find a local steel supplier, and not get pricing from online dealers, where the cost can be 3-5x higher.

As far as getting the tubes flat. The most common method is to build dams, and pour layer of epoxy for a bed for the rails. The epoxy self levels, providing perfectly flat mounting surfaces.
Another method I've seen is to use a thickened epoxy mixture on the tubes, and place them, epoxy down, on a surface plate, or other flat surface with a release film. This will also give you a flat surface, and may be a bit easier.


I don't think that there's anything wrong with your plans to build a carbon fiber machine. But expect to build at least 2 or 3, as you'll learn things with each part you make.

[wizard]

I have some steel steel and aluminum tubes here that I could use but I would still need a flat surface to mount the rails on wouldn't I?

Depends upon your design! If the steel tubing is think enough and you want to spend the money you can have the tube machined flat

everywhere I have looked for large steel plates has been very expensive. We are talking $250 -$350 per piece before shipping.

Sound like you are looking online, this is the wrong place to be looking. Fine a local steel supplier, preferably one that sell drops at a discount. Last time I looked scrap steel was shockingly cheap, like 35 cents a pound if you tried to sell it to a scrap dealer. You will pay more for drops but it is normally less than a dollar a pound. The price is variable in time and variable Witt the amount of steel you purchase. At least it is with my local steel supplier.

I will try to dig up receipts from recent purchases (at work at the moment), but you should be able to purchase a box section beam suitable for a 30" span gantry for under $100. Depending upon the beam you could come in under $75.

I don't know if people have outdated views or if I am looking in the wrong places but as far as I can tell, steel is the most expensive option. It's a lot more expensive than even a top quality T6 aluminum and far more than any composite solution.

When talking about aluminum I'm talking about T-sloted extrusions, all the extras that you buy to get everything to go together and most likely a plate to face the extrusion with. Al that stuff adds up pretty fast. Also while T-Slotted extrusions are available in many shapes, most of them leave a lot to be desired structurally.

[louieatienza]

I have some steel steel and aluminum tubes here that I could use but I would still need a flat surface to mount the rails on wouldn't I?

everywhere I have looked for large steel plates has been very expensive. We are talking $250 -$350 per piece before shipping.

I don't know if people have outdated views or if I am looking in the wrong places but as far as I can tell, steel is the most expensive option. It's a lot more expensive than even a top quality T6 aluminum and far more than any composite solution.

The way it's usually done with steel tubing would be to weld strips of steel flat stock, not much wider than the linear rails, onto the steel tube. Those surfaces would later be machined or ground to accept the linear rails. With some setup and imagination, you can do this with epoxy or even CF. You'd use steel plates judiciously, like for the gantry carriage Y-Z plate maybe for example; though a lot of folks here use Mic6 cast aluminum plate, because drop-offs are pretty inexpensive on eBay, and it cuts and drills easily, even with woodworking tools (saws, drill press, etc.) and it's Blanchard ground flat on both sides, so you have an accurate datum plane on which to install LM rails.

Steel is cheaper than aluminum by weight, but it's about 3-4 times heavier than aluminum, hence the cost.

Probably the most famous large scale steel router using epoxy leveling would be the MadVac router. The website is still up, and you can see how he used epoxy to level his rail surfaces as well as bearing surfaces between bolted joints. A hybrid design of sorts with CF over steel, aluminum, or even wood would be another option...

[Goemon]

I have a lot of epoxy bar top here so I can make anything level. Whatever I use, it's going to have to be fairly flat to start with or the epoxy bar top will become a point of weakness (if too much has to be used).

The tube I have here are round. with solid bar clamps, they could probably be made into a fairly strong and rigid frame that is quick to construct. It would be ugly though.

I pulled my first carbon fiber plate out the mold (the optical mirrors) and I am satisfied that they will be rigid enough for the gantry side plate. I can stand on them with the edges held up by two chairs without any visible bending. They will be a lot strong than many of those over-priced aluminum gantry plates people sell on eBay for $200.

I don't want to offend anyone who makes and sells those things (ok maybe I do a little.... I can't help it) but when I look at what you get for $200, I can't help thinking "is that it?" and "are you serious?" and "is everyone else just seeing two small thin aluminum plates with some red paint?".

Take a $10 aluminum plate, change the name to "CNC aluminum plate" and now it's worth $200. Good scam. No.... wait a sec... you have to cut off one of the corners first, then you can increase the price 20x...

[Goemon]

The way it's usually done with steel tubing would be to weld strips of steel flat stock, not much wider than the linear rails, onto the steel tube. Those surfaces would later be machined or ground to accept the linear rails. With some setup and imagination, you can do this with epoxy or even CF. You'd use steel plates judiciously, like for the gantry carriage Y-Z plate maybe for example; though a lot of folks here use Mic6 cast aluminum plate, because drop-offs are pretty inexpensive on eBay, and it cuts and drills easily, even with woodworking tools (saws, drill press, etc.) and it's Blanchard ground flat on both sides, so you have an accurate datum plane on which to install LM rails.

Steel is cheaper than aluminum by weight, but it's about 3-4 times heavier than aluminum, hence the cost.

Probably the most famous large scale steel router using epoxy leveling would be the MadVac router. The website is still up, and you can see how he used epoxy to level his rail surfaces as well as bearing surfaces between bolted joints. A hybrid design of sorts with CF over steel, aluminum, or even wood would be another option...

I had had a look at that MadVac site. That guy has a large budget for a diy build. You could probably buy what he made ready to play for $16,000 these days but I have to say, what he built looks amazing. He has every right to be proud of what he made and show off a little.

[ger21]

You have to remember that he built that about 20 years ago. Things were a LOT different back then.

And unless it's made in China, you won't find a welded steel 4x8 router for under $20K-$25K.

I pulled my first carbon fiber plate out the mold (the optical mirrors) and I am satisfied that they will be rigid enough for the gantry side plate. I can stand on them with the edges held up by two chairs without any visible bending

What size, and how thick are they?

Care to share some tips on how to get rigid parts? Materials, Orientations?

I have a 40"x90" vacuum frame press, with a table that's been cnc machined flat. It's flat enough to make router parts.