Squares are square, circles are not round

[brtech]

Newbie here trying to get my router doing something useful.

I am trying to cut some countersinked bolt holes in plywood (eventually MDF, but I'm just playing around with a scrap of ply. These are for 5/16" bolts, and I've got a 1/8" upcut spiral bit.

If I cut a profile of a square, it comes out fine. When I try the countersink with a spiral pocket, the circle is misshapen. I'm guessing the tool is bending for some reason.

I've got it set low: 60 ipm, .0675 doc. It should be able to do this at 600ipm and .25 doc, which is how deep I want the countersink, i.e. one pass. The router is set to around 20000 rpm.

I'm probably doing something stupid.

This is an FLA300, R&P drive on both X and Y, with an M12V router, Mach3.

[ger21]

Picture? It's probably backlash or flex somewhere.

How are you creating the spiral code?

[brtech]

The left one is from sheetcam, the right one was generated from Cambam.

The path in Mach looks round.

I notice that it's repeatable - these have 4 passes, and each pass is pretty much the same as the other passes.

[judleroy]

Try cutting it with one pass at 5-10ipm and see how that works. If you were cutting something bigger the higher speeds would be fine but on something that small the motors may just need more time to accelerate and then decelerate.

[brtech]

Interesting

This is 10 ipm, .25 DOC, around 10000 rpm. I got burning with these settings, but the circle is much better. It's not perfect.

Does this mean I have acceleration too high on the motors?

[ger21]

I still think it's flex or backlash. Can you get the pinion tighter to the rack?

[brtech]

I currently have the spring about 1.25x collapsed length, so I can get it tighter.

What would you think a reasonable speed for this cut should be?

[ger21]

Depends how rigid the entire machine is.

My Z axis has a bit of flex, so, the smaller the diameter, the slower I go.

Try lowering the accel quite a bit. If the pinions are climbing out of the rack, then that may help.

[brtech]

Tightening the springs on the R&P and lowering the acceleration (it's now at 10 ipm) didn't help as much as lowering the speed rate to 5 or 10.

The entire frame is pretty rigid. I can't get it to move any perceptible amount yanking on the router motor.

This is at 10 ipm cut rate

[judleroy]

The problem is most likely caused from the motor needing to reverse almost instantly when cutting small circles at high feed rates. This is a problem at high federates because your motor has come to the end of a move after accelerating and then needs to reverse directions with no time to decelerate the move first. This can be tested easily by cutting a 6 inch circle at your original 60ipm or more and seeing how the cut works out. It should look fairly good even at those speeds. I see this most often with rack and pinion machines. If you were using a carbide tool it could not flex enough to cause this problem because it would have broken the tool before it flexed that much. It is possible that the spindle has that much deflection at the end but that's a bigger issue.
Judleroy

[judleroy]

To try to prevent the burning drill the inner circle first then plunge into that hole and spiral outwards.
Judleroy

[ahren]

I'm not an expert in Mach, but I have seen problems, especially with small moves and high feedrates, with the constant velocity mode in Mach. Here's an article I found on the topic. You might try toggling this on/off to see if it makes a difference.

I'd suspect the tightness of the bearings on the carriages, but you have indicated that the machine feels rigid when you pull on it, so probably best to leave this alone for now.

Ahren
CNCRouterParts