[vivid]

Hi guys,

I wonder if anybody can help me with the problem I have.

I've made a router for a friend of mine for milling and drilling aluminum profiles for LED lighting. Milling pockets works well, but I have a problem with drilling.

The profile is "U" shaped aluminum extrusion 200x200x600mm , around 7mm thick. It's well supported, leveled and held in place via pneumatic cylinders.Basically all the driilling (40-50 holes array) is done with 2.6mm drill bit, holes should be around 5mm deep. I do it in one pass, no peck drilling. Spindle motor is 1.5 kW air cooled. Speed is set to 20.000 RPM.

The problem is, the holes are getting deeper with every profile (and probably with every hole, too), so after aproximately two profiles, the drill goes all the way throught the 7mm plate, and the height has to be corrected to avoid drilling through.

If I test the program adjusting Z so that the drill only barely touches the surface, everything works correctly from start to finish. I've checked if the drill bit is maybe for some reason pulled out from the collet, but that's not the case.

As the macghine doesn't seem to have any problems milling 3mm deep pockets, and the motors are cool I don't think the strenth is a problem either.

Any ideas? I'm attaching a render, just the spindle motor is bigger and the profile fixture not shown...

Thanks in advance for any suggestion.

Miso

Attachment 317124

[A_Camera]

Is the depth increase linear? I mean, it starts right and as the gantry is traveling to the other end the depth is increasing and when you start a new profile it starts with the last drilled depth and the depth increases as much as it increased with the first profile and so on?

Otherwise I'd suspect the unsupported rods. When drilling, especially in one go, the material is pulled up towards the spindle by the spindle. As the spindle travels towards the center of the unsupported rods it gets easier to pull it upwards and the hole deepens. Of course, the depth should decrease again as it moves towards the other end. This is an issue I had until I upgraded my CNC from unsupported rods to fully supported ones. When milling it is a bit different, the pulling forces are not as high so you may get better results, even though, especially for milling aluminum I'd definitely not use anything less than supported rods.

Of course, the problem can be in the code, taking some kind of offset and increasing the depth automatically. What software are you using to generate the code and to handle the axis movements?

[DonKes]

Does your Zaxis show the depth changing?

If not, you may be losing steps.

Don

[awerby]

It's a typical lost-steps pattern. That straight drilling plunge with no pecks is too much for the stepper, so it loses steps going down, leading it to think that it's higher than it really is. Then when it goes up, it's lower than it should be, and goes deeper on the next down-stroke. Milling a pocket is different; you just go down a little bit and the rest is horizontal.

If I'm right about this it will show in your DROs; when you've run your program trying to drill those holes and you tell it to return to Z zero afterwards, the tool will be considerably lower than originally. To remedy the situation, slow the plunge rate way down, and preferably use peck-drilling instead of trying to do it all in one go. Use plenty of lubrication as well; aluminum tends to stick to drill bits, which keeps them from working effectively.

[LeeWay]

Most drill bits should not be going anywhere near that fast. If you are just using standard drill bits, then there is the problem. I would instead use a high helix end mill and ramp in with it.
A two flute straight carbide bit may do it too.
The fastest I drill with in aluminum is about 3600 RPM and that is a small drill bit. I do drill in plastic on my router and the slowest it will go is 8000 RPM. 3/16" drill.
It works okay, but I peck it. It is drilling 1" deep.

[pippin88]

It's a typical lost-steps pattern. That straight drilling plunge with no pecks is too much for the stepper, so it loses steps going down, leading it to think that it's higher than it really is. Then when it goes up, it's lower than it should be, and goes deeper on the next down-stroke. Milling a pocket is different; you just go down a little bit and the rest is horizontal.

If I'm right about this it will show in your DROs; when you've run your program trying to drill those holes and you tell it to return to Z zero afterwards, the tool will be considerably lower than originally. To remedy the situation, slow the plunge rate way down, and preferably use peck-drilling instead of trying to do it all in one go. Use plenty of lubrication as well; aluminum tends to stick to drill bits, which keeps them from working effectively.

I think it would be the opposite.

Stall (lost steps) while plunging. Controller thinks it is still moving down, but the tool is not. Now the tool is higher than the controller thinks.

To get deeper with each hole it would need to be losing steps on the retract (moving up).

Agree that the Z zero needs to be checked at the end.

[A_Camera]

I think it would be the opposite.

Stall (lost steps) while plunging. Controller thinks it is still moving down, but the tool is not. Now the tool is higher than the controller thinks.

To get deeper with each hole it would need to be losing steps on the retract (moving up).

Agree that the Z zero needs to be checked at the end.

Right. It can't be lost steps due to the forces and the deep drilling in one move, otherwise the holes would reduce in depth, not increase. Must be extra steps, or offset added by the software or some other issues, but definitely not lost steps.

[LeeWay]

Oh it could be lost steps due to a hot drill bit sticking to the aluminum and losing the steps on the retract. I would slow the velocity and acceleration down along with the spindle speed and try it again.

[coherent]

Lost steps on retract. slow down the z retract speed. Peck drill and plunge slower also. Double check your coupler setscrews etc also.

[Louis_Cannell]

Agreed losing steps on up stroke whether due to hot drill bit or some sort of binding on Z axis motor, just seen another case with same problem on 6040 motor coupling moved rubbing causing lost steps on upstroke so check coupling and bearings and if present thrust bearing.

[awerby]

Yes, Pippin, you're right - the steps would be lost on the upstroke, not the plunge. I've had that happen when the spindle was too heavy, but I got it backwards in my post - good catch.

This problem might be due to that, or to something coming loose in the Z-axis mechanicals, like a set-screw. I've had that happen as well.

[medu2015]

Yes, Pippin, you're right - the steps would be lost on the upstroke, not the plunge. I've had that happen when the spindle was too heavy, but I got it backwards in my post - good catch.

This problem might be due to that, or to something coming loose in the Z-axis mechanicals, like a set-screw. I've had that happen as well.

I experienced the same problem on my PC board router. The problem tuned out to be lost steps on rapid retract due to the mass of the Z axis. The problem was eliminated by slowing the rapid
Good Luck

[vivid]

Hi Guys,

Thanks for all your answers. I'm sorry I was away, and didn't set up the emailing properly, so I just thought I didn't get any answers.

I also think that steps are lost while retracting, otherwise the depth would decrease. A friend suggested compensating for the spindle and Z support weight by using something like elastic band or spring, I might try that when I get back home first, it can't hurt...

I'm using Mach3. I've written the Gcode manually, and there's nothing progressive in it, all the depths are copied from one hole to another, and the Z axis DRO shows the right height at the end. I've also tested with the Z depth set to just touch the surface, and it worked flawlessly from start to end, so i think it definitely must be something else, as it only occurs when actually drilling.

Maybe I should try the bigger Z axis motor, I would definitely prefer not having to peck drill if I can, as with these profiles time is critical. I will also try to change the stepper drivers. At first I had the borrowed Leashine 8A drivers, clearly an overkill for the motors installed. Now i have some cheap TB6660 4.2 A drivers, and while the current should still be more than enough, maybe they are causing the problem.

I don't think that unsupported rods are the problem, as they worked fine previously with Leadshine drivers.

Milling will be eliminated completely with the new profile design, so the machine will be used exclusively for drilling. The drill bits were recommended by the local store which specializes in the CNC machines and equipment. I know the price doesn't guarantee the quality, but these are the only ones to last so long (50+ profiles). They cost €17,00 each, quite a sum for a 2.6mm bit... Although there's no cooling, the aluminum doesn't stick to the drill bit, at least there's none visible, and spraying with WD40 also doesn't change much...

Thanks again for your help, will try everything you suggested and report the result as soon as I get back.

Miso

[A_Camera]

Hi Guys,

Thanks for all your answers. I'm sorry I was away, and didn't set up the emailing properly, so I just thought I didn't get any answers.

I also think that steps are lost while retracting, otherwise the depth would decrease. A friend suggested compensating for the spindle and Z support weight by using something like elastic band or spring, I might try that when I get back home first, it can't hurt...

I'm using Mach3. I've written the Gcode manually, and there's nothing progressive in it, all the depths are copied from one hole to another, and the Z axis DRO shows the right height at the end. I've also tested with the Z depth set to just touch the surface, and it worked flawlessly from start to end, so i think it definitely must be something else, as it only occurs when actually drilling.

Maybe I should try the bigger Z axis motor, I would definitely prefer not having to peck drill if I can, as with these profiles time is critical. I will also try to change the stepper drivers. At first I had the borrowed Leashine 8A drivers, clearly an overkill for the motors installed. Now i have some cheap TB6660 4.2 A drivers, and while the current should still be more than enough, maybe they are causing the problem.

I don't think that unsupported rods are the problem, as they worked fine previously with Leadshine drivers.

Milling will be eliminated completely with the new profile design, so the machine will be used exclusively for drilling. The drill bits were recommended by the local store which specializes in the CNC machines and equipment. I know the price doesn't guarantee the quality, but these are the only ones to last so long (50+ profiles). They cost €17,00 each, quite a sum for a 2.6mm bit... Although there's no cooling, the aluminum doesn't stick to the drill bit, at least there's none visible, and spraying with WD40 also doesn't change much...

Thanks again for your help, will try everything you suggested and report the result as soon as I get back.

Miso

OK, thanks for the feedback. The only thing I'd disagree with is using of elastic band. It is better to use counterweight because elastic bands are non-linear, the more they are stretched out the more they are pulling back, so the pulling force will not balance the motor well enough. I'd use counterweight because with that, regardless of the depth of your spindle, the force is constant, all the way down to the deepest hole you manage to drill.

Though... I'd suggest to start with reducing speed and acceleration, even if you don't find that as a good solution. If it works it says that the stepper, the driver or both are not up to the task.

[vivid]

OK, thanks for the feedback. The only thing I'd disagree with is using of elastic band. It is better to use counterweight because elastic bands are non-linear, the more they are stretched out the more they are pulling back, so the pulling force will not balance the motor well enough. I'd use counterweight because with that, regardless of the depth of your spindle, the force is constant, all the way down to the deepest hole you manage to drill.

Though... I'd suggest to start with reducing speed and acceleration, even if you don't find that as a good solution. If it works it says that the stepper, the driver or both are not up to the task.

I agree the band is not the best solution, I meant it just as a quick and dirty way to see if the counter force helps at all. The vertical travel is relatively small, 10mm or so, so I don't think the band's non-linearity would change much, at least for the testing purposes.If that helps, I'd look for the more permanent solution, counterweight, and/or the bigger motor.

Thanks for your input. I will probably try everything on Monday, and I'll post the results here.

Miso