This isn’t a router, but this seems to be the only thread for DIY machines so I’m putting it here…
I just got done building this 16 ft x 5.5 ft fabric cutting machine which is a tangential cutter that has both a cutting tool and pen on the same tool head for doing both operations without needing tool changes. This is my first CNC machine and was a lot of fun so I plan to build a mill next summer. It might have made more sense to start with the mill, but the fabric cutter is actually a simpler machine that requires less precision and lower cost components, so it was a better choice for a first time builder I think. The down side is that there seems to be no information anywhere on building fabric cutters, as not many people appear to be building their own. While router and mill plans are all over the place, fabric cutters are nowhere to be seen. Perhaps this is due to the fact that the CNC hobby is 99% guys, and most guys do not sew! I have to admit I catch a lot of crap for even owning a sewing machine, so I just want to state up front that my only purpose for sewing is to make hot air balloons, not dresses. I use one type of fabric, one gauge of thread and only know how to sew one type of seem used for making balloons. So THAT, my friends, is why there is a sewing machine in my man cave. And it’s not some plastic betty homemaker Jo Ann special, it’s an industrial double needle, all metal Japanese Juki built like an engine block, so yeah… just wanted to clear that up
Because of the large size of this table, it was necessary to find the absolute lowest cost option for rails and sliders, which turned out to be makerslide from inventables.com. Speed is a higher priority than precision with a fabric cutter, so the aluminum 80/20 type construction was ideal. I looked at a lot of options for how to build a machine this large without breaking the bank and makerslide was the optimal solution. Not only is this the most economical way of building a large format fabric cutter, it is also one of the most light-weight and expandable designs. The machine shown here uses 43 linear feet of makerslide.
Another criteria for a fabric cutter is the ability to expand the length of the X-axis later on if you need more length. The primary reason I built this cutter is for cutting large fabric panels used for making hot air balloons. The 16 ft X axis would handle most balloons, but if someone were going to make sails, ultralights or paragliders then a longer table might be needed. The table can be expanded in 4ft increments by building more table segments and then adding more rail. Chain drive was chosen for the X axis so that the chain can be easily lengthened by splicing onto it. Timing belts would require replacing the whole belts with longer ones. Gear racks can be expanded by adding more rack segments, but they are very expensive and would have busted the budget.
I’ve noticed a lot of chain drive machines just leave the chain hanging on the X-axis, which requires an excessive amount of tension on the chain to pull out the sag. For a shorter X-axis this might not be an issue, but the longer the table gets the more weight of chain you have to try and pull the sag out of. So what I did was use a chain gutter to support the chain for the full length of the table. The ¾” aluminum channel you can buy at hardware stores works great. With the chain gutter holding the weight of the chain, you are free to apply only the amount of tension required to remove play, thus putting less load on the motor bearings. I was worried that the chain might make a lot of noise rattling around in the metal channels, but it really doesn’t.
Another complication that arose from having such a long X axis was cable management. Originally I was going to use the catenary wire festoon method of running the cables, which is where you have a tight steel cable from which loops of wire are strung such that they can expand or compress as they slide on the wire. This turned out to be a more difficult method with several disadvantages over drag chains, such as entanglement issues, requiring longer over twice as much wire, bunching up at one end and just doesn’t look clean. It seems to be a fairly standard way of cable management for commercial fabric cutters, but I just didn’t have any luck with it. I went with a long drag chain instead, using a support shelf to prevent sagging issues as described here:
http://www.cnczone.com/forums/diy_cn...ong_spans.html
Like most fabric cutters, the entire table surface is a vacuum table. Rather than using commercial blower motors that cost over $300 each, I used two 5HP Rigid shop vacs under the table costing $80 each. This saved a good bit of money and works just fine. Not only that, when I need to clean off the table I can pull one of the shop vacs out and use it to clean the table. I will mainly be using this for cutting coated ripstop nylon, which air can not pass through so the vacuum table holds it down quite well.
Unlike the DIY vacuum tables you might have seen built for router table drag knifes, a fabric cutting vacuum table must have very small holes—otherwise the fabric will get sucked into the hole and produce a dip which the knife will fail to cut. The holes in this table are 1/16” diameter with 3” spacing, drilled through 3/16” hard board. Each panel has a 16 x 22 hole grid and is sealed so that airflow does not occur between adjacent panels. This allows you to power only the panels you need for a given job rather than the entire table. So that was 1408 holes that had to be drilled for my four panel table, and since my machine has no Z axis or spindle, these had to be drilled by hand. It wasn’t as bad as it sounds though.
Because of the light weight of the gantry and even lighter weight of the tool head, I can easily run rapids at 1300 IPM or more. Rigidity is not as much of an issue here since there is no heavy spindle with side forces acting on it the way there is with a router. The levelness of the table does not matter much either, since the tool head is pneumatically actuated and will hold the tool to the table regardless of any dips or peaks in the table surface (as long as they are within the range of the cylinder stroke). The air pressure is independently adjustable for the pen and cutter, both of which use springs for return. The cutter head was heavy enough that I had to add an additional external spring because the internal spring in the pneumatic cylinder wasn’t enough to bring it back up.
In the videos below I am using a hot-knife cutter which I built from an 80W soldering iron. This cutter has tangential cutting capability, so there is a blade rotation motor that keeps the blade tangent to the cutting line as it goes. For drag knife you just wouldn't use this motor, but for hot knife and wheel cutters you would need it. Currently all I have made is the hot knife, since that is the main thing I need, but other knife types could be made and inserted into the tool shaft in place of the hot knife. The ideas is to build each tool type into a 7/8” O.D. cylinder that slides into the holding tube. The hot knife is actually the most challenging option since you have heat dissipation issues. I could not use normal ball bearings on the tool shaft because the high temps would break down the grease and melt any plastic bearing seals. I went with graphite impregnated sleeve bearings for this, and still need to add a muffin fan that blows air through the gear box. Without the air flow, the heat transfer will slowly heat up the entire tool head and eventually overheat the blade angle motor.
I did have to write a custom post-processing program that takes g-code generated for 3-axes machines and converts it into what a 2 axis fabric cutter needs. This mostly consisted of stripping out Z axis data and inserting the knife/pen up and down macros at the proper places, and also calculating what the knife angle needs to be based on the X,Y data and setting the A axis motor to that value. When A axis angle changes are below a certain threshold, the tool head stays on the table. For larger angle changes, the tool head comes off the table so that it does not twist the fabric when turning. Because the hot knife has a power cord to it, the program also insures that it does not ever turn beyond 360 degrees in order to keep the cord from getting entangled. Another feature I built into the post processor is the ability to automatically add registration lines at regular intervals around the entire cutting perimeter. Registration lines are reference marks that are used to keep two pieces of fabric aligned when sewing them together, since a sewing machine actually pulls one piece more than the other and you must continuously struggle to keep both pieces in synch. Rather than having to include these lines in the original CAD files, my program will automatically generate the g-code for perfectly spaced lines based on the geometry of the patterns and add them into the final output file.
The remaining work left to do is build a computer console with joystick and other buttons running through an i-pac, and build a screenset for Mach which is specific to fabric cutting. Total build time for this machine was about three months and material costs were about $3K, which also includes a small air compressor to run the pneumatics. A commercial machine this size would run you about $30K, so that is quite a bit of savings!
Here’s some video of the first tests:
CNC Fabric Cutter: G-code modification test - YouTube
Automated Registration Marks & Labeling - YouTube