Frustrated/failed experiments

I’m having a bad cutting day :frowning:

I had been trying to do detail work in hardwood with 1mm cutters. I have run tests and have determined that parameters for the 1mm cutter should be close to that of the built-in 1/16th end mill. This makes sense, since 1/16th is pretty close to 1.6mm.

In test cuts, it works fine. But when doing something larger or more involved, it seems that forty minutes into the work, something will happen and the cutter will get snapped off. I haven’t seen it happen, but I hear it from across the room.

So for now, I’ve switched to only using the provided cutters and MeshCAM wizard settings. The next problem I’ve run into is that the work won’t stay put. I’m cutting old hard oak, in this case. At that same critical forty minutes into the process (in this case the roughing phase), and the work starts breaking free, causing gouges.

I think I need to find a better system than the double-sided tape. The problem is I can’t use a vise or clamp, as the outline is cut out on these guys, and there would be nothing left to hold.

So I switched gears again, and started carving into a softer wood (still a hardwood). This design is much simpler. I’m just cutting an outline. I set up MeshCAM to do the entire process with the waterline and pencil cleanup. I did not machine the top of the stock, or do a roughing pass.

This time, I witnessed the problem. As the cutter was coming around the edge, it suddenly started scribbling like a kindergartner who didn’t want to stay within the lines.

So I’m not sure what I should have done here. I’m guessing the problem has to do with feed rate or plunge rate being too high for the wood and the cutter is getting snagged or deflected? Perhaps if I had done the roughing cut, it would have left less resistance to the waterline cut?

I’d welcome any words of advice from people who are being more successful.


Same issue here, lost 2 carvings today after about one hour or so, one was definetely because the tape came loose, the other one im not sure since i was not watching it. . Image below. It also had some suspecious looking straight holes drilled in areas where there shouldnt be any.

One answer might be better tape. The double sided tape that came with my machine was fine for some jobs but failed in a way similar to what you describe early in your posts with thicker parts especially when the cutter reached the tape layer. So, now I use ‘duck’ brand double sided duck tape for heavier jobs. Found it on amazon. Duck model 240200 1.4 inch x 12 yards. It has much higher grip. I also make sure the sacrificial board and my stock are dust free before attaching the 2 parts. It is sticky enough that it is hard to get some parts off.


yeah i now am using the crimping solution i posted Self made solution to affix wood without tape and it seems to prevent the wood from moving, although its not perfect still since i had to tighten the screws that hold the stock several times since the nuts wiggled themselves loose it seems.

I think once the vise is available, i will go for that since i get too frustrated at the moment. I will get the tape from Amazon though to see if that helps, thanks for the hint!


I put a note about fixturing over in the “self-made solution” thread, just cross-posting here for reference.

OK. So today I tried making my own jig. Unlike oliverblum’s approach, I cut some strips of soft aluminum, and drilled them at 7 inches to match the mounting holes for the wasteboard. I then used washers and bolts to secure everything (for reference, they’re 6mm 1.0x50mm bolts and M6 1mm hex nuts).

I still used the two-sided tape under the stock.

This is the same design where I had the “drawing outside the lines” problem mentioned above. I loaded the very latest MeshCAM beta, and recreated the toolpath making sure I was specifying hard wood and including a roughing pass this time.

While it took about 3 times as long as MeshCAM estimated, it completed successfully this time!

So, some things I need to do next time: cut a touch deeper. The stock appears to have been 0.1 or 0.2 mm thicker than I specified. I need to remember to tell MeshCAM that the stock is larger than the input model when cutting shapes from a larger piece of stock. I learned this before, but failed to put that learning into action. The result is that MeshCAM does not complete the cut all the way around as you can see at the bottom of the picture.

It looks like my wood is not strong enough for the fine details of this model. Some of the smaller spur points broke off. This could have been in the removal from the wasteboard, because the tape is strong enough to make it difficult. I’ll have to come up with some way of handling that, or finding denser wood.

There is also one curiosity - the “branching” shapes are not uniform. If you look at the larger branch that’s a little to the right of twelve o’clock in the picture, you’ll see that it’s thicker than the similar branches. In the original model, they’re all identical. I’m not sure what’s going on there. The STL file looks like my original model, but maybe in the conversion to STL I’m losing some detail. Or maybe it’s lost in MeshCAM.

A few things:

  1. Good job on the fixture! You may want to use a shorter bolt for the back left corner to avoid accidental collisions with the a tool going to or coming from the tool height sensor. You might also want to consider using L, C or U channel for your next iteration, as that won’t flex like flat stock does, so it’ll clamp the stock with even pressure across the surface. The flat bar will flex, so it ends up pinching down on the corners and bowing over the middle.

    As long as you can keep from cutting into it, that should work like
    a champ. In fact I may have to make myself a pair of hold-down bars myself! :+1:

  2. Does your y-axis have any play at all in it when the machine is on? When it’s off the motors won’t resist you, but when it’s on you shouldn’t be able to wiggle anything without using a fair bit of force. It’ll only move once you’re exerting more force than the holding torque rating of the motor. If you have any wiggle, refer to this thread about how to fix it, as that’ll resolve any backlash that could be causing the unevenness in branch size a poor cut quality.

  3. The grain in a given wood makes a huge difference how things will come out—wood with larger grain is much more likely to splinter & not hold small features well, especially if your cutters aren’t really fresh and sharp. It looks (at least from what I can tell in the pics) like you’re using red oak—you may want to try something with a finer grain to see if you get a better result in it.

Looking forward to the next cut!

Excellent suggestion on the U-Channel. That would be a lot better.

I don’t seem to have any backlash. I think the irregularity is in the toolpath. Here’s an image of the STL file from Meshlab, and a simulation cut from OpenSCAM (I’d rather use Rob’s Cutviewer, but can’t afford it at the moment). On the simulation, you can see the irregularity as well.

I’ve been working through a small bag of hardwood samples. I’ve been meaning to get to the local artisinal plywood / hardwood place. I just need to make time to get there, and see what they have available.

It looks like the irregularities are caused by the tool being too big to cut the small thorns in those locations. The CAM is doing the “right” thing and protecting the model by cutting what it can.

1 Like

I agree with Jonnie—looks like it’s time to do multi-tool machining plan!

Probably want to use the bit you have (1/8" flat end-mill) for the roughing + primary finishing paths and then use a considerably smaller bit for pencil clean-up. Just make sure you have the right feeds/speeds for the smaller tool, and keep in mind that they’re considerably more fragile than larger cutters so you have to be more gentle with the chip-loading.

For what smaller bits to use, I just found bits & bits from someone else’s suggestion here in the forum and ordered a few things on Wednesday, so until Carbide expands their own tooling range to include much smaller stuff, that might be a good option for smaller cutters.


i like the bars and 2nd unionNine’s comments about watching out for the height of the bolts. Meshcam has a good “do not machine” feature that I would use in combination with lower bolts to avoid accidents. Also thanks union for the link to bits and bits. nice find.