New Machine Build- Just built a new CNC - One issue: What bits to use?

[gepeto]

Hi,

First of all I want to thank the entire community here for all the useful information you contributed and continue to contribute.

I was lurking for a long time in these forums but unfortunately I didn’t have much to contribute. I like doing wood working as a hobby (mostly to relax). A month ago I had the guts to actually start building a CNC which I just finished a couple of days ago. It is a pretty light CNC (made of plywood) with a working area of 20”x24” (pictures & videos available on my website: www.adystuff.com in the projects section)

My main goals were:

• to be cheap (don’t forget I do this as a hobby)
• to be as simple as possible (yet as robust as possible)
• to be as accurate as possible (for cutting MDF or plywood sheets and wood milling

wood)

Surprisingly, it really works (beyond my expectations).

Now that the CNC is done I was confronted with a question: What kind of bits can be used if you want to cut a sheet of MDF or plywood? I used a 1/4 straight bit but the problem is that the bit gets clogged. To actually cut the MDF sheet I had to run three times the bit at three consecutive depths and clean-up between runs

In the past whenever I had to do a cut I used either the table saw or the jig saw or other tools. To my own surprise I never use the router for more than creating grooves or routing edges (never to cut).

So, I would appreciate if you can help me with some suggestions.

[sonex128]

Nice looking router. I am just getting started on building mine so I can't help you out with the bits.

I don't believe that anybody cuts MDF in one pass. I thought that it always took multiple passes.

Brian Bland
Claremore, Ok

[BobF]

Spiral upcut bits are good for removing the debris. They work well in plywood and mdf. Solid carbide is the way to go, especially for MDF.

[ger21]

If you don't start cutting a LOT faster than your video, any bit's you use will get dull very quickly from the heat generated. With small bits (less than 1/2"), dust will always get packed into the cut. Spirals will help somewhat, but not completely. It is possible to cut through in a single pass, but I doubt that you're spindle has nearly enough power. Compression spirals are typically used when cutting through in a single pass. They tend to be rather expensive, over $50 for a 1/4" bit.

[gepeto]

Thank you everybody for your comments

ger21 you are right. I tried with a higher speed with a depth of 1/2". The spindle cannot carry that for sure. I managed to cut but the sound from the spindle told me clearly that was not the way to do it. But what worked best it was to cut less with higher speed.

The spindle I attached to the CNC is a pretty crappy one (underpowered). I used it just because the form factor made it easy to install and I was anxious to try the CNC (plus in case I was doing something stupid I didn't mind breaking it).

Definitely I should try spiral bits. I look at home depot but they don't have any. So probably I need to get them online.

Somehow based on some experiences I run I tend to think that I should just make multiple passes (even with spiral bits). The piece I cut in the demo was in three passes (I kind of cheated because I recorded only one pass). That works for me too but I wanted to make sure that cutting in one pass is not really possible with the configuration I have (this is a home built CNC ...far from a pro one)

[ger21]

You might be able to cut in one pass with a 2HP router, but would probably need a 3HP or more, which are big and heavy. Multiple passes are fine. Everyone does it that way.

[BobF]

These guys make nice bits and are pretty reasonably priced. The V-bits are great for V-carve.
http://www.centuriontools.com/

[gepeto]

Thanks BobF & ger21

I made more tests and multiple passes worked just fine even with the crappy spindle installed now. The precision is amazing considering the components that went into the CNC. Just for testing I'll make a support for my 1.75 HP router.

The machine should be able to handle the weight of that without problems. The T frames I implemented in it pay off now because I can stand myself on top of the table (over 200lbs :-( ) and it will not bend at all. Actually all axis are re-enforced the same way.

I'm pretty happy that I decided to do it. I was sure the result would be way much worse (again, considering the components I used to build it ..especially the cheap rods driving the axis).

[sonex128]

What are you using for leadscrews? I have cheap 1/2" Acme rods from Enco.

I may steal your idea for reinforcing for my machine. It will be made with 3/4" MDF.

I also have a Hitachi router for it.

Brian

[gepeto]

I used ordinary rods 5/16" - 18 tpi + these lead screws just because they fit the "rollerblades" ball bearings. I have no tools to handle metal and I was not convince the whole shebang was going to work. So I went for cheap stuff.

Now, looking back I think it is worth spending some money in acme screws. The lead screws I have definitely will not cut it for a decent load on the machine. On the other hand I have no idea how much I will use this (don't forget this is a just hobby for me)

As for "stealing" ideas. There is not much to steal. All ideas are coming from one build or another on this forum.

I just tried to pay attention to what was important and what not that critical in the build.

Just some decisions I made (that proved good)

1. I didn't like to much the idea to use pipes because to build support for them (to avoid bending) required a complex structure (hard to build without another CNC). Using V profiles 1" x 1/8" x 3' I managed to create a very strong support (because they stay on the table itself and I used 2 layers of plywood where the V rod is supported (see here).

2. Also I wanted to make sure the parallelism is easy to control. As a result to use the table as the support worked just nicely. The support for pipes using screws to align them were, again to much for me.

3. Everybody was complaining about bending. As a result T frames all over for X Axis, Y Axis, Z Axis. I'm happy about choosing plywood instead of MDF. Although plywood is not stable in one direction using T frames made it extremely robust. MDF worked for a lot of people. So probably what I did was an overkill.


Some ideas that were not that good:

1. Rods & lead screws are not good. Although 18 tpi can provide a good precision, you cannot run this at high speeds. The best I was able to run it without loosing steps was 40 ipm with minimal load. They cannot support high speeds and also they cannot support load.

2. I threaded the coupler for the steppers to fit the rod (they were 1/4 and I threaded them to 5/16). They are not properly aligned. This is not the way to do it for sure.

My conclusion is that the overall design is OK (and also very simple) but using ACME screws is almost a must.

[BobF]

gepato,
Mcmaster has some 5/16" 8 2-start acme which you could probably swap in fairly easily. That works out to 4 tpi and should double your speed or better, and still fit your bearings.
You might have to make your own lead nuts. If you get a bit of extra rod, you can make a tap and tap a plastic lead nut. Delrin plastic works well and HPDE would probably work as well.
1/2" bearings are readily available if you want to go to 1/2" acme. There are several popular sizes used in lots of builds here. 1/2-10, 1/2-8, 1/2-8 2-start, 1/2-10 5-start have all been used in Joes 2006 builds.

[gepeto]

Originally Posted by BobF gepato, Mcmaster has some 5/16" 8 2-start acme which you could probably swap in fairly easily. That works out to 4 tpi and should double your speed or better, and still fit your bearings.

Thanks for the tip. probably I'll be using the machine as it is for a while to get accustom with all the other stuff (Mach3 tunning, G-Code, etc). ...and probably I'll start building a better version of the current one (definitely with ACME screws). Considering that I worked only in my spare time (not much) and it took less than 6 weeks to complete it I think it's worth thinking to do it from scratch.