Instrument maker

[Grambot]

Hi Boyan

Thanks for getting back to me. I've had a chance to look at your build. Looks amazing!

I'm experimenting with small keyed xylophones.

I'm wondering if I can make the main casework out of timber, machined on CNC. It would be like a large wooden pencil case, or dough bowl at the bottom, housing most of the mechanics. This would be hopefully milled out of one block:


And there would be an aluminium shape like this, secured above it, around the keyboard. Hopefully this could also be CNC'd?:


Ideally I'd like an accuracy of +-0.1mm. Is this realistic?

And a machine with a working area of at least 500mm x 200mm x 120mm (Z)

Thanks

[Grambot]

Well, definitely looks possible! :-)

[Boyan Silyavski]

[QUOTE=Grambot;84109]Well, definitely looks possible! :-)

That on the video is an example of lazy cnc-ing or test cnc-ing. He uses a bit i will use only for sth smal like 50x50mm. i can make that bowl on my machine faster than he, using 1/2 or bigger cutter. Plus he does not use roughing at all.




My instruments were 0.01mm overall even on my old machine which was to say "flimsy". "Overall" means in some places could be and is worse, but on wood that's not a problem. In reality now with my machine i have always to slow down things and spend time adjusting perfectly toolpaths so to avoid chipping. So at the end whatever they say, but CNC is an art


So you don't need super fast machine , real life 10 000mm/min will be ok. 5000mm/min will be slow to my liking, but it will not be bad if your machine can reach 20000 mm min so you use HSm techniques there like digging full depth / see my video/ . What happens at the video is at 16000mm/min. The problem though is that now i don't use that toolpath, cause it makes a lot of noise and around me people pay to live in quiet near the sea
You will need servo drives and motors for that speeds. And looking at that bowls i greatly suggest that, otherwise they will take forever to make


in fact one of the reasons to build my machine was to make my instruments more 3d, like the bowls. And soon i will start experimenting with the sounds of it, ast shape must be matched to sound in my case

[Clive S]

What happens at the video is at 16000mm/min. The problem though is that now i don't use that toolpath, cause it makes a lot of noise and around me people pay to live in quiet near the sea
You will need servo drives and motors for that speeds

Boyan why do you think you can't run at 16000 mm/min with steppers?

[Boyan Silyavski]

Boyan why do you think you can't run at 16000 mm/min with steppers?

Because i have never seen it achieve serious speeds except on RP builds

As far as i know to reach let's say 16 000 mm min on a xx10 ball screw you will need to spin that motor at 1600rpm. At 200 steps per rev then resolution will be mere 0.5mm . So microstepping then and loss of the incremental torque. The acceleration also counts when fast machine is desired, and many changes of direction have to be made when routing large 3d file.

At the other side for the machine to be able to work at that feedrate the gantry and Z should be very strong / ok, just strong-not talking OTT here/ which means weight

The spindle to achieve serious feed rates and deep cuts or fast shallow cuts with big cutters needs to be powerful. In fact the way i see it my 3kw spindle is limiting me now, not the speed that the machine could achieve. Though more or less i am finding now that my machine is well balanced. Anyway, i wanted to say that the 3kw spindle needs serious Z axis. Otherwise the bit will be prematurely dull because of vibrations.


So i don't see 100kg gantry on small machine moven by steppers at that speed. Not to speak of 200kg gantry on 8x6 machine. From what i have read in forums and builds i have looked and studied, people using steppers normally achieve 5000-mm min on a normal build and 10 000mm / min on the better thought builds. I doubt that this 10 000mm /min is really usable with good acceleration on actual cutting, not rapids.


I am not sure about current prices but 4x30 +4x80 =440 euro for 4 Leadshine drives plus motors. When at the same time i said it many times here for a bit more/ ~200/ you could get the cheapest servo motors plus drives. Which i believe is to be more reasonable considering a good build.

And now after testing the cheap chinese controller for a couple of weeks- Let's cut this money from the expensive controllers.

[Grambot]

Thanks for all this info everyone. This is so valuable for me.

My instruments were 0.01mm overall even on my old machine which was to say "flimsy". "Overall" means in some places could be and is worse, but on wood that's not a problem. In reality now with my machine i have always to slow down things and spend time adjusting perfectly toolpaths so to avoid chipping. So at the end whatever they say, but CNC is an art

Does this mean, I'll be spending a lot of time getting the right toolpath and speeds etc, and eventually I'll have a program which I'll be able to keep using for one particular design?

in fact one of the reasons to build my machine was to make my instruments more 3d, like the bowls. And soon i will start experimenting with the sounds of it

I'm also interested in trying out many designs, timbers to see what effect it has on the sound.

The problem though is that now i don't use that toolpath, cause it makes a lot of noise and around me people pay to live in quiet near the sea

Yes, it does seem loud. I'm wondering if I can use my machine at home during the day without too much extra soundproofing. I know how to soundproof, so its not the end of the world, but it would be great to be able to occasionally leave a window open :-)

If not, I have to hire a small place somewhere, as I don't have a garage. Where do people keep their machines? Or I suppose it must depend on how big/noisy they are?

[Boyan Silyavski]

Thanks for all this info everyone. This is so valuable for me.



Does this mean, I'll be spending a lot of time getting the right toolpath and speeds etc, and eventually I'll have a program which I'll be able to keep using for one particular design?

It means you will start cutting fast, then lower speeds when you encounter problems. It also means sometimes instead of profiling, you will have to pocket some extra pockets, so as the rests from wood not to obstruct the vacuum shoe, or chip when they fly away. Meaning last pass should be carefull and well thought. / Lastpass is when you cut off the detail/ . As i said experimenting should be done with each material for optimum results.

I'm also interested in trying out many designs, timbers to see what effect it has on the sound.

Thats the spirit

Yes, it does seem loud. I'm wondering if I can use my machine at home during the day without too much extra soundproofing.

Forget about that

I know how to soundproof, so its not the end of the world, but it would be great to be able to occasionally leave a window open :-)

Its not the machine that makes the noise. Its virtually silent with the water cooled spindle. you can not hear it if you are doing sth small. Its the bit in the wood that makes the noise. A big noise. No windows open. For home scenario, you need a to make an enclosure. And if you have neighbours touching from one side, forget about that too. Even the frequent vacuum cleaning will make them crazy.

[JAZZCNC]

So i don't see 100kg gantry on small machine moven by steppers at that speed. Not to speak of 200kg gantry on 8x6 machine. From what i have read in forums and builds i have looked and studied, people using steppers normally achieve 5000-mm min on a normal build and 10 000mm / min on the better thought builds. I doubt that this 10 000mm /min is really usable with good acceleration on actual cutting, not rapids.

Boyan you are correct that machine with 100kg gantry won't reach 16,000mm/min with steppers but at 100kg it shouldn't be fitted with steppers so should be no surprise.!!

That doesn't mean correctly built and spec'd machine using steppers won't reach 16,000mm/min. most of the routers I build range between 600x400 and 1250 x 700 region and just about every one will reach those speeds. While testing/setup I push them to the extreme limit to find the saturation point where they stall and 20-24mtr/min is not uncommon.
Yes they don't get run at those speeds and I'll whind them back to between 10-15mtr/min rapids. All will happily cut at 10mtr/min if the tooling and material allow for it.

Servos are ok but as you know they also come there fair share of issues and requirements. Servo's while being stronger are not always better and in DIY enviroment mostly not required or recommended. KISS is the key at DIY level and steppers tick all the boxs. Servo's can quickly crush the boxs and frustrate the hell out of people. Also the cost doesn't just stop at the Servo's because again as you know suitable controller and all the correct cabling are required. (before you reply with "use chinese controller" not everyone want's to take this route)


Grambot:
Noise is big concern for most DIY users and can't be avoided because it mostly comes from the cutter not the spindle but it's simply case of enclosing the machine. Just make sure it's well ventilated.

Boyan the movement of your machine was nice to see but the trochoidal toolpath was IMO the wrong choice for this type of Job/material. It would have been quicker roughing in conventional manner then finish pass and much less stressful on the machine.

[Boyan Silyavski]

the way i see it :

long 3d toolpaths with serious material removal rate requirement → more spindle power needed hence 2.2 and even better 3kw /for home use/ → heavier and stronger Z with long travel /remember deep bowls/ → stronger gantry → heavier build → at least 70kg gantry with all mounted




Dean, i agree, The trochoidal path was only to test vacuum fixture, machine and board capabilities. I wanted to see what happens when cutter is routing full depth. Now i use again HSM technique, instead i changed depths with width and dont use trochoidal path, except for profiling the piece and also its only 3mm deep as deeper leads to chipping of edge. If that makes sense.

[JAZZCNC]

Ideally I'd like an accuracy of +-0.1mm. Is this realistic?

And a machine with a working area of at least 500mm x 200mm x 120mm (Z)

Thanks

0.1mm is more than achievable and your cutting requirements are also very easily achieved. Thou you may find little more than 200 will be more usable and give more scope.