[handlewanker]

Hi......unless someone sets up to build BLDC motors, for a hobby business, similar to those that build precision spindles etc, we will not see anything that fits the hobby budget.

You can source all the parts on AliBaba.....if you can generate the quantities large enough, from stator stampings, magnets and Hall effect devices.....all those parts that are exotic and impossible to find locally......the rest such as end plates, bearings and spindles can be sourced or made anywhere local.

As we only need one or more motors, the prospect of hand winding a bunch of stators will not appeal to anyone due to the complexity of the build.

I had great fun some time back playing with a small burned out motor, just messing about, but the real reason was to see if the cooling blades on the end could be turned into a turbine rotor.

Well.....don't try this at home comes to mind........I bent the blades at right angles to be in line axially with the spindle.....a pseudo De Laval turbine design....... and fixed an air nozzle to the side of the casing aimed through the slots where the air was drawn in......it was spectacular.

Once the air stream was properly directed and clamped the rotor started spinning and emitted a low howling sound like a siren, getting louder and higher in pitch.........at around a hundred thousand rpm..... or so it seemed...... the rotor disintegrated.....actually the bearing on one end, smoked, seized, collapsed and bits went every where.....but it did work as a turbine.

BTW, the windings that were left on the rotor also attempted to part company and ended up wrapped round the rotor between the stator and the rotor gap.

I suppose, if the rotor had been stripped down......or even removed from the shaft all together, as the modified fan is all that was needed to make a turbine, and the stator too, just leaving the outer casing and end bearing housings.....you would have a better turbine almost ready made and add some form of governor to keep the bits together......such fun.

Now if we could drive that from a small steam boiler there would be no need for a compressor banging away as it did use a bit of air.

I'd much prefer to have a hydraulic motor powered from a car's power steering motor as it's more compact and quieter and the pump comes with a pressure regulator to prevent blowing up the motor.

Going to that design and I think you could use a die grinder as the hydraulic motor with thin hydro oil as the medium instead of air, and power the pump from a 3 phase motor off the machine........die grinders have very small body diams and are designed with moving vanes, so low speed and lots of torque are the result without having a burn out from stalling......just a thought.

I've got an old die grinder that uses gallons of air, so I rarely use it at all......might be food for experiment.......if my son, who dabbles with Gemini car building, has a spare power steering pump around I might have a go at experimenting....time allowing..
Ian.

[louieatienza]

As I understand it trochoidal tool paths have no benefit with facing but have a huge impact on pocketing and profiling in the x/y plane

Maybe not trochoidal, but there are high speed toolpaths used for facing as well. It's also about maintaining a constant load on the spindle, and making even non-cutting movements smooth (i.e. no sharp moves or corners) that can cause jarring moves that slow down feedrates and cause stress and wear on the machine. For higher end CAM the radial DoC would be dynamically controlled depending on the situation (i.e. "nibbling into a corner"), and long, arc-shaped sweeps are used as much as possible (not interpolated x-y moves that generate tons of code.)

The argument against using a HS toolpath for surfacing is that it may or may not leave a pleasing "finish" depending on your preference. I kind of like the look of pocket floors.

[daniellyall]

the trochoidal tool paths and the other one`s don't know what there name is, the way it goes in and out of the cut is a sweep (it looks like a french man`s mow with curled ends) from zero to full engagement to starting the cut it has full engagement when it`s moving side to side in a sweep then goes from full engagement to 0 engagement, it sweeps out of the cut.

the HS tool path`s it depend`s on the tool used as well I agree Lou they do look nice in a artsy fartsey way

[handlewanker]

the trochoidal tool paths and the other one`s don't know what there name is, the way it goes in and out of the cut is a sweep (it looks like a french man`s mow with curled ends) from zero to full engagement to starting the cut it has full engagement when it`s moving side to side in a sweep then goes from full engagement to 0 engagement, it sweeps out of the cut.

the HS tool path`s it depend`s on the tool used as well I agree Lou they do look nice in a artsy fartsey way

Hi Dan......would a circular/spiral path be better as it's a constant cutter engagement path?...........that is, starting at the centre of a face and going down a mill then just cut with a spiral path until you break out of the sides......the G code would be pretty simple to manually generate even by my limited knowledge, or so it would seem.

For simplicity, I think I'd do the same with a slot, that is, create a circular path, width of the slot, and just move forward with it.....simple G code manoeuvring and repeat......is this a problem?
Ian.

[daniellyall]

that`s what the trochoidal and helical tool paths do when it gets to the edge it swoops in and swoops out moves back to it`s start the repeats if it`s set to conventional or climb. if it`s set to both ways it keep`s a constant contact where it can, I set it to Helix into the cut what ever way it`s doing it, its just in a trochoidal or helical shape

Attachment 286450
the first pick is both ways with constant engagement if you look you can see it is trying to remove the same amount all the time

Attachment 286452

this pick it`s doing a slot with conventional cutting the green lines is the swoop out yellow is moving across to the next cut the green line it run`s into is the swoop in, the blue is the cut, red is the helix in.

the first one is more of a helical cut (the other fancy cut) than a trochoidal but they both have constant engagement that`s what makes them better and be cost saving they produce a constant forces on the machine and cutter when they go around a corner they slow down to keep it constant. I think that right it was ages ago that I read a paper on it.
it has a major pucker factor when using them for how deep you can run the cutter with it not snapping a tool or browning out.

[handlewanker]

Oh dear....it's starting to get complicated......in a piece of work like the first pic, I would just make a 70mm diam circular cut with a 10mm diam cutter...... whatever......moving forward a mill at a time and work back and forth across the job in straight paths, same as using a fly cutter.

In a clock wise circular cutting path, once the cutter is fully into the cut and cutting a full 70mm circle it advances at the 6 o'clock position and starts cutting at the 9 o'clock position..... finish cutting at the 3 o'clock position.......even though it loses time for 180 deg at the lower half of the circle, the G code is relatively simple to manually generate as it repeats continually.....I think.
Ian.

[daniellyall]

it is but you lose constant forces the first pick is 300 X 500 4mm depth off cut. it`s a cut what I have done at 25mm deep , the second pick is a 20 x 100 x 10 deep cut. ( go hard or go home)

if it term`s to a clock its in at 9,6,3,out at 9,6,3 in a arch, loop what ever

[RCaffin]

Hi Ian

not my dreams.......a poster Khoac3 post #125 a couple of posts back on this thread clearly stated it is what he achieves or does........that is, it's his preference to go with a 10mm diam carbide cutter in steel at 7mm deep and with a small 1.2mm step over at 5,000 rpm, as opposed to my 10mm end mill with a shallow full diam cut of 2mm.......not quite my cup of tea.

Yeah, I know.
I should not have commented on TWO different postings in the same reply.

Cheers
Roger

[handlewanker]

Hi, at the end of the day the machine will decide on what it can do, so until a suitable motor manifests.....anything can happen.......I'm resigned to the fact that small is beautiful and the metal removal rate will be proportional to the grunt......you wouldn't expect a boy to do a man's job.
Ian.

[daniellyall]

well if you can do 25% in steel,and in aluminum 50% would be about right as those chips need to come out you could get away with big cut`s with a 1 Hp what ever motor using the fancy cut`s or cutters

a 10mm diam carbide cutter in steel at 7mm deep (70%) and with a small 1.2mm step over (12%) at 5,000 rpm, as opposed to my 10mm end mill with a shallow full diam cut of 2mm. (20%). adding coolant changes the cut amount, adding fancy tool path`s lesion power required same with the fancy cutter coatings you can get now.

so what to get has to be a call on what you will cut the most if it`s steel sometimes get something in the middle and use fancy cutter`s and tool path`s for steel then you can get away with having less HP, yes the cutter`s cost a lot if they are used correctly they last a long time. and fusion 360 is cheap for a program that has the fancy tool path`s yes it`s still buggy but it works.

[louieatienza]

that`s what the trochoidal and helical tool paths do when it gets to the edge it swoops in and swoops out moves back to it`s start the repeats if it`s set to conventional or climb. if it`s set to both ways it keep`s a constant contact where it can, I set it to Helix into the cut what ever way it`s doing it, its just in a trochoidal or helical shape

Attachment 286450
the first pick is both ways with constant engagement if you look you can see it is trying to remove the same amount all the time

Attachment 286452

this pick it`s doing a slot with conventional cutting the green lines is the swoop out yellow is moving across to the next cut the green line it run`s into is the swoop in, the blue is the cut, red is the helix in.

the first one is more of a helical cut (the other fancy cut) than a trochoidal but they both have constant engagement that`s what makes them better and be cost saving they produce a constant forces on the machine and cutter when they go around a corner they slow down to keep it constant. I think that right it was ages ago that I read a paper on it.
it has a major pucker factor when using them for how deep you can run the cutter with it not snapping a tool or browning out.

The first example is not a high speed toolpath to me, looks like a standard offset toolpath. A high speed toolpath does NOT come unto a corner like that for the fact that the tool engagement angle spikes, and the tool has to slow down.

Here is a sample, in iso as well as top view. The tool is helixed in, so that there is no full tool flute engagement. If you plough into a corner you basically double to tool engagement, the machine slows down, and surface finish suffers. The constant accel/decal from stop is not good for the mechanicals either, as well as the spike in tool engagement (nearly double) in the corners.

Attachment 286478
Attachment 286480

You can see in this example how a constant spiral is used for the majority of the cut, switching to the "trochoidal" toolpath into the corners. The radial DoC (tool engagement) decreases slightly as the arcs get tighter. I should note it's only cost-saving if you optimize your axial DoC (cut depth), feedrate, and radial DoC (stepover). Also you should strive for the deepest cut your machine can handle, and adjust the stepover respectively. Also it should always be done climb cut, to prevent the tool's flute from "digging in" to the material at high speed. The deeper the cut, the more rigid the tool.

[daniellyall]

the tool path images`s in fusion are a bit of it may of been climb milling they don't quite match what is happening, the second image is trochoidal at max depth of what was required, I know the first one is not trochoidal I said that it more a helical cut. they both helix in and did the cut at full depth with full width of cut the first image`s cut is 10mins faster than Vcarve pro, they both have full width of cut also the first one is both direction if I had it climb milling it looks very different like your second image also that`s a pick of the second op so it has no edges to run into. and i did say this, I think that right it was ages ago that I read a paper on it.

so if I am wrong that`s life I will get over it but your using a different program

[Feist92]

I just recieved a chunk of 32mm steel plate that will be used as the table. I didnt really think about this but it weighs aroud 23kg by its self.
Does anyone have any suggestions for which size steppers to go for and where to get them? Im leaning towards nema 34 size but thats just a guess based on the fact that there will be alot of weight on the x and y axis to throw around.

With regards to the spindle motor I think Ill be going with one of the Adlee 2,2kw motors shown here
Adlee AM2200-H BLDC Motor & Drive
Or one of these servo motors.
130ST cnc servo kit ac servo motor electric motor 220v 10N.M 2.6KW servo motor cnc 130ST M10025 B025D-in AC Motor from Industry & Business on Aliexpress.com | Alibaba Group

The servo motor ends up around $400 cheaper but im not sure how suitable it would be

[daniellyall]

you will have to way up the risk`s get from here where you have come back our law`s or a may be may be not

that`s a lot of weight to move and stop stick it on the rail`s and work out how much force is needed to move it there are a lot of online calculators you can use to work out how many Nm you will need post results here as well what you think may be ok could be wrong

[Dubbie99]

Don't forget to add a vise and stock weight. And no, you don't need to assemble it to work out the torque you need. You can work out your torque requirements using high school math.

[daniellyall]

stick it on the rail`s

[RCaffin]

Hi Feist

I just recieved a chunk of 32mm steel plate that will be used as the table

Remember that the steel plate will warp if/when you mill T-slots into it. You may need to mill the slots, heat the plate a bit for a little while, then mill both top and bottom surfaces flat. That's why most tables are made from cast iron: it does not warp as much.

Cheers
Roger

[Feist92]

Don't forget to add a vise and stock weight. And no, you don't need to assemble it to work out the torque you need. You can work out your torque requirements using high school math.

This is true but Im finding it dificult to work out what numbers to plug into what equations. I have no problem working it out long hand once i have all the data. I plugged some numbers into the Linengineering stepper calculator but it said I could use a 100oz/in stepper in 1/2 microstep which sounds wrong so Im guessing Ive got some numbers wrong. Im going to do a bit more research this weekend

stick it on the rail`s

Haha that was the first thing I did. The plate will be skimmed down to about 30mm with a large 6mm deep pocket on the under side and covered in holes so at a guess it might end up closer to 16kg

Hi Feist
Remember that the steel plate will warp if/when you mill T-slots into it. You may need to mill the slots, heat the plate a bit for a little while, then mill both top and bottom surfaces flat. That's why most tables are made from cast iron: it does not warp as much.
r

Im planning on more of a tooling plate design so wont need to worry about that so much.

[daniellyall]

use a forces calculator what out puts Nm like this Force, Mass, Acceleration Calculator -- EndMemo

read this as well http://www.geckodrive.com/support.html

you just need to know how much forces is needed to move the axis if you pull on it with a set of scales to it moves that is how much forces is require to move it. working out just the weight of what ever the axis weight is wont be quite correct on my big router the Z axis carage weight is 60 kg it does not take 60 kg of forces to move it you can do it with one finger

60 kg works out to need 7.5 N at 1250 m/sec what works out to a motor of 1062 oz-in, it has a motor of 750 oz-in on it and it works fine as the forces to move it is not 60 kg

friction and gravity have an effect as well as the temp at the time of measurement plus a few a things

one other thing I was told with a ballscrew 10 mm pitch is best for steppers and 5 mm for servo motor`s

there are MIT calculator that you can use what you put in weight/mass, Acceleration, gravity,friction and what you are moving the mass with how many oz-in it gives a negative or a positive number on what is move the object.

or just email a suppler what you need

[RCaffin]

I think you can ignore friction in the calculations [assuming linear bearings]. All you need to worry about is acceleration, for the load, the ballscrew and the motor..

FWIIW, I haven't seen even quite small motors having any real problems once you have a 2:1 or 3:1 reduction from the motor to the ball screw, plus a reasonable ball screw pitch. Plus, any of the CNC programs allow you to limit the acceleration to be applied by the SW. [Small enough steppers at low voltage may present a problem.]


I tried a tooling plate design approach, and went back to a few good T-slots. But a different work regime or set of jobs might make a different demand.
Fwiiw, my mill table sags 30 microns outside the edge T-slots. Most annoying. And that is cast iron.

Cheers
Roger