[D.L]

Folks, what do you think of the new Centroid 5 axis?
The patent was only lodged last year on the head, I reckon it is awesome!
5 axis CNC Cylinder Head Porting Machine. Digitize and CNC port Intake, Exhaust ports, and Combustion chambers.
Could we build one? LinuxCNC maybe?

[RICHARD ZASTROW]

Just like the oldest profession, asking the question "HOW MUCH?" LOL

There is value in the fixturing, tooling and resident CNC part programs for some common engines. Otherwise, you will have to compare these Centroid machines to other 5-axis machining centers, building the tooling and CNC programs.

If the knowledge base, machining capability and capacity, time and money are there it could be done. My cohorts in crime and I do this for a living. (actually they do, I just help with the mech. eng. part). End product 4-8 axis gear hobbers.

One other question, Who is WE???

Dick Z

[D.L]

Just like the oldest profession, asking the question "HOW MUCH?"

Well i was thinking a chinese ballscrew with a nema23 stepper, plus a pivot point added to the design of the carriageway, but this is not a machine build (yet).

Who is WE???

Well, you've cornered me there, we as in anybody who cares to dabble with 5 axis.
It seems so much energy has been spent on draw bars and tool carousels added to Syil mills, really, if you're making parts in the shed what's the point?

Would'nt a turbocharger impeller, porting a motorcycle cylinder head or a dry sump bed plate for a hill climb motor be the kind stuff we'd like our machines to build?

Thats my 2 cents, each to their own.

[miljnor]

Would'nt a turbocharger impeller, porting a motorcycle cylinder head or a dry sump bed plate for a hill climb motor be the kind stuff we'd like our machines to build?

Pluck ya! (substitute your Non-Pc word as needed) :nono:

lets do it! :cheers:

[RICHARD ZASTROW]

My question was to determine if WE meant you and some associate(s) or the fellow members of this forum. That is all.LOL

Dick Z

[D.L]

Well I've traced out photos of the mechanism and it would seem that this could be done using screw mapping - a pitch error ranging from 0 to +21mm and back to 0.

If mach3 can handle this range of error, then there will be no need for LinuxCNC.

Does Mach3 'interpolate' the errors into a curve, or does it just make a linear jump from one point to the next?

The 'we' was just a collective term, everybody shares ideas here and information.

But with respect to the company that holds this patent, i intend to NOT gain financially from mimicking this idea.

Purely a tinkerers appreciation of fine machinery!

[mactec54]

D.L

This takes the shine off things,Centroid thinking this is there Idea,is wishful thinking

Directmotion have been doing this for may years, for Centroid to copy there machine design, & call it there own is shameful,& applying for a patent for a design that is not there's

Direct Motion - 5 Axis CNC Machine Tool Control Technology

http://www.rottlermfg.com/cylinder_h....php?model=P69

[D.L]

Nice find mactec,
That rottler is sweet!

The brochure mentions

Rotary Vector Gear Reducers (or Cycloidal gear drive) give less than 1 arc/min on the 4th and 5th axis, reducing wear, extending life and boasting near zero backlash

That's a whole other bag, the ballscrew method seems superior in terms of not having fulcrum induced backlash (lever arm effect).

[mactec54]

D.L

They used to do it with a ballscrew drive the same way as Centriod is, & must of gone to the gear type drives, I don't know what the problem was with the ballscrew setup to make them change, There original machines were not like the Rottler, they used Knee mills with there own head on it, & used the same ballscrew setup to move the head

I think with the ballscrew you are limited to how much movement the head has

So the Centriod is not different in anyway just that machine that I posted is different in the way it is driven, Directmotion were the first to do this & many years ago

[D.L]

I will dig around for more variations, mazak has one as well.
Below is a quick and dirty summary of the basic movements and also an early concept draft of converting a bench top mill.

[mactec54]

D.L

What you have done in your drawing looks good, but you can see the limits of travel( 120deg ) with this system are not ideal for normal use, for things like porting, that is plenty of movement, 270deg is the ideal movement, for normal 5 axes use with this type of moving head design

[holbieone]

looks like the old Boston Digital

[holbieone]

the more i look at it that is the old Boston Digital

B axis is driven the same as a rotary and with a counter balance

[mactec54]

holbieone

Yes Boston Digital had the 505 5axes that has the moving head, there may have been other models as well, but they did not use a ballscrew to move the the head

[ame=http://www.youtube.com/watch?v=va9TPg_UhmQ&feature=player_detailpage]B1MC 5-axis CNC porting - YouTube[/ame]

[D.L]

There is another good 5 axis design, the heads pivot is roughly located in the region of the tool tip. This maximises the workspace, on a small mill every mm is valuable 'real estate'.

Nice find!

[D.L]

Well, the Mach3 screw mapping is a dead end, it requests input from an encoder and range is only 2 units of movement.
It does have a feature called 'formulas', if the axis movement can be described mathematically, then PROBABLY no problems. (never done this before)

[holbieone]

holbieone

Yes Boston Digital had the 505 5axes that has the moving head, there may have been other models as well, but they did not use a ballscrew to move the the head

B1MC 5-axis CNC porting - YouTube

we have a 605

no the head was driven like a rotary (worm gear)

[D.L]

Back in 3 months!

[D.L]

O.k, I think I'll do this, it's not that hard if you use the right mix of donor parts (car parts).
I still need to convert to 3 axis to kick the project along, my Z axis will be a quick and dirty stepper to the quill feed.
(not using original column or head for final build)
I won't build log the 3 axis, but I will start a log once 3 axis is running and I start welding and machining the column.
The early cad sketch I've shown you was a flawed design, this is all clean sheet stuff, November to December expect photos.

[mactec54]

D.I.

Good to see you are back, & moving on with your build