CNC for a novice

[John11668]

How does a novice get into CNC
I have never tried but something is telling me I should have a go.
Do I try to convert a machine , Buy a small CNC lathe , or what .
How do I learn the requirements of CAD and programming the machine .
In short what is the best way to get in.

 

[shred]

What kind of parts are you wanting to make?
Most people seem to start CNC with mills rather than lathes (the concepts are a little easier and it's more suited to one-offs), but I think you'll want an idea what you want to do first, unless it's just mucking around with a CNC machine, which a number of people also seem to be into.

 

[ttrikalin]

i got a small sherline CNC mill.

Before buying it I installed the Ubuntu distro that has EMC2 and played a bit with the CNC simulation.

If you are not comfortable with electronics, get a ready system.

take care,

tom in MA.

 

[Tin Falcon]

how one gets into cnc for a hobby is usually a function of skill level in machining and electronics vs how deep are your pockets. There are web sites that have plans for the electronics and you can build from scratch. You can buy retro fitted machines and factory fitted machines but you are talking thousands of dollars rather than hundreds for a conversion. CNC is not necessarily an easier way of making a part but a different way. Both require knowledge of your machine and machining technique and creativity.
Wrote a thread a while back trying to explain the basics of CNC and what is needed for a conversion.
read here
http://www.homemodelenginemachinist.com/index.php?topic=9787.0
there are several build threads that could also be helpful in a conversion.
Bottom line is you will either spend some money and a lot of time researching and learning the skills to convert a cnc or lots of money buying one. as far as the programing you will need to understand some G-code.
Mach 3 has optional wizards that are pretty conversational and will pretty much write code for you you tell the computer the material you are using the cutter you are using and then answer questions about the part.
Hope this helps.
Tin

 

[1hand]

John,

I was in your shoes about a year ago. Since then I've done 2 conversions. First with a kit for my mill, and then a scratch build for my lathe. It is do able. The fine folks here walk me right through it, even though I change my mind in direction afew times along the way. After doing it now there is a few things I would change, but I would do it again in a heartbeat. The rewards are 10 fold.

..I knew nothing about it when I started, and am still learning, but it can be done.

We are all here to help if needed.

Good Luck,
Matt

 

[kvom]

In my case I took classes at the local votech school. For both lathe and mill, we first learned CAD/CAM for the machine, then simulated our programs in software, and then finally made parts on the actual shop machinery.

I don't think either lathe or mill CNC is more difficult than the other. For the mill we used a CAM program to generate g-code from the drawing, while for lathe we generated the g-code manually (the programs are generally a lot smaller).

For my current build (Kozo A3), a CNC mill is a lot more useful than a CNC lathe, but neither is actually a requirement. It's just that some parts are easier (not necessarily faster) via CNC. CNC makes sense for me when:

1) There are a lot of non-orthogonal outlines (e.g., angles and arcs)
2) There are multiples of the same part

 

[John11668]

Thanks for the prompt responses guys
My machine shop is really my main project at the moment.
Colchester bantam lathe to which I have fitted DRO ( from singapore :)
Deckel FP1 mill which has needed a bit of work on the more exotic features.
The shop is a sort of pre -retirement project which I hope will become more absorbing when I have more time to play

I was maybe thinking about a CNC mill or lathe as a project in itself, so get something basic and add a kit. I see reasonably cheap kits on Ebay (usually from Singapore)
and wondered about the functionality of those. Dont mind making the brackets and things. Redundant computer is never a problem these days so the actual outlay could be small.
Much potential though for mistakes :-
How do you size motors , drives , power units, etc ? What is the function of the breakout board? Do you( from the start ) give yourself the option of adding further axes?

Then having assembled what may be a working piece of kit , how do you make it do stuff.
Where In UK are there any courses.
Is it feasible to learn by trial and error
Is there easy software to start with and can you then step up to more advanced stuff.
In the meantime will have a look at some of your links - thanks again.

 

[Tin Falcon]

How do you size motors , drives , power units, etc ?

for a mill the power of the motor needed is based on the weight of the heaviest object moved like the mill table and how fast you want to move it.
the drivers need to be rated for the the motor amperage.
voltage of power supply is limited by the diver as well the amperage by the motors.
A full run down here http://www.geckodrive.com/support.aspx?n=626915

What is the function of the breakout board?

this is a board that takes the connection from the computer and connects it to individual drivers the drivers in turned are wired to the motors. it breaks down the signals to multiple axises.

Is there easy software to start with and can you then step up to more advanced stuff.

Mach 3 is designed to for an easy start. with built in wizards (conversational programing. ) It is highly customizable to run just about anything that can be computer controlled.
You can download a demo version free and play with it to you hearts content and have no machine attached.

Then having assembled what may be a working piece of kit , how do you make it do stuff.

you need to program the computer to make the moves to make the part.
1) g- Code
2)conversational programing like wizards in Mach GENERATES G- CODE. .
3)cad drawings converted by CAM program to g code.

Is it feasible to learn by trial and error

one way to learn is by cutting machinable wax. Recover the chips with a clean shop vac melt and reuse the wax . The wax is forgiving a crash will not damage tooling. also running the program thought the computer in off line mode will debug things.

Tin

 

[John11668]

As a total novice I feel I must go step by step!

My Deckel mill is devoid of any auto traverse on the Ram (Z axis)
I had considered the £300 traverse motor as used on the Bridgies but it would need special adaptors machined. and a 110V power unit so not cheap for a semi solution.

If I add a stepper and a timing belt drive then I am sure It can be driven by a computer /breakout /driver setup and I am keen to follow this as a project.

It would seem logical to also go for electronics which can handle spindle control (0-10V output would control my VFD from 10 to 80 cycles which would give me 200 to 3000rpm so that would seem a desirable option rather than just relay output)

If I can successfully control the ram, then table traverse (y axis) would follow naturally and z axis ( table rise and fall ) should be feasible too.

Just by operating these motions manually it is clear that they require vastly different effort . Ram is almost effortless, traverse requires a bit more effort, lifting the table needs a fair bit more still.
A bit of research shows that there are a variety of motors out there all of which can be geared up or down but most of the available drivers and BOBs seem to have three four or five equal outputs.

So firstly how do I measure the torque needed to drive a particular motion?
How do I then convert this into holding torque which I presume must also be sufficient to resist cutting forces or even drive against them?
I read that steppers perform at their optimum if well matched to the duty(poor matching can lead to missed steps)

Then presuming I could come up with three distinct optimum motors
For the sake of argument say I need 175 Oz in. for the ram , 425 for table traverse and 800 for the table lift, how do I go about finding a board to drive the varying loads and indeed how do I size the power supply to handle all together albeit with a degree of diversity.

Then being a forward thinking sort I wonder if I should have a couple of spare axes just in case I choose to drive the quill for peck drilling, or a rotary axis

There seem to be a wide variety of electronic options out there. ( many from the far east) Do any of you have "tried and tested" experience of any.

It does appear to be a minefield for the unwary.
All suggestions will be valued

John

 

[Tin Falcon]

It would seem logical to also go for electronics which can handle spindle control (0-10V output would control my VFD from 10 to 80 cycles which would give me 200 to 3000rpm so that would seem a desirable option rather than just relay output)

Certainly doable my g540 has that feature.

So firstly how do I measure the torque needed to drive a particular motion?

put the handle4 at 3 o clock position hang a small scale on it tug down notice the reading. you may have to do it a couple of times. force times distance(center of lead screw to center of handle ) say it it is two inches handle to center and it takes 10 pounds of force to move the handle then 10(LBS) x 16 (oz/lb) x 2 = 320 in/oz. the max force to move the axis will be static friction witch should be a reasonable holding torque remember a stepper max torque is when it is stopped.

Then presuming I could come up with three distinct optimum motors
For the sake of argument say I need 175 Oz in. for the ram , 425 for table traverse and 800 for the table lift, how do I go about

finding a board to drive the varying loads and indeed how do I size the power supply to handle all together albeit with a degree of

diversity.

Not a problem with gecko drives you can do a modular setup you can even use steppers and servos in the same system just make sure the drivers can handle each motor.

Then being a forward thinking sort I wonder if I should have a couple of spare axes just in case I choose to drive the quill for peck drilling, or a rotary axis

the 540 has 4 axis ready to go.

There seem to be a wide variety of electronic options out there. ( many from the far east) Do any of you have "tried and tested" experience of any.

http://geckodrive.net/categories.html quality made in USA.


http://www.kelinginc.net/index.html has a good rep my power supply came from him.
tin

 

[Iffy Boat Race]

Before you jump in to CNC do some research.There are lots of options out there for the home machinist.Have a look at stepper motor options,Control board options and most importantly have a look at control software and CAD/CAM packages.
I have been lucky enough to get some CNC time with my work.12 months ago i got the opportunity to become the second operator of a Doosan Puma CNC lathe.
I knew nothing when i started and now i do most of the setting up and running.I have yet to start programming.
Anything CNC can be a challenge and the rewarding.
Be prepared for a learning curve......when you can get it all working and not crashing your tools makes it all worthwhile in the end

 

[Tin Falcon]

IBF advanced and be recognized citizen. an intro in the welcome section would be a marvelous thing.
a bit out you . you shop your interest in model engines etc.
Thanks for you insight.
Tin

 

[John11668]

Thanks to all for the guidance!

The Gecko gear looks good TF and from research is well thought of all around but the carriage to UK is a bit steep.
I am tempted towards a modular set-up sold in the UK which looks initially to be well thought out.
What do you guys think of this (system 4C)

http://www.diycnc.co.uk/

Anyone out there have any experience of it

 

[Tin Falcon]

No experience with that product .
observations looks like a good modular system Modular is good if one part goes bad swap out a card. features look good like opto isolation. the one down side i see is fairly modest power capability.
that unit is rated at 30v max at 2.5 amps or 75 watts per axis . that is probably enough for an x 2 mill or a small lathe.
in comparison my gecko is rated at 50v @3.5 amps or 175 watts per axis. Probably more than needed. I use a 3bv power supply but could have gone with a 48 comfortably. More volts + more speed IIRC my x axis movement is 8 inches how much rapid travel do I need not a lot.
As a second comparison:
Here in the US the Xyltek board is much talked about, but. it is a one board wonder. If one driver goes the whole board is toast also it is rated at 24 volts do not put 24.5 volts to it less the magic smoke is released. this same board is sold under different names. I admit I was tempted to go that direction myself. I have talked to guys that have used them with no problem. other folks have toasted them .

Like I said before the power you need is determined by the heaviest thing you need to move . for instance a mill table w/vise and how fast you want to move it. Rapid feed. just make sure the system you get has enough power to meet your expectations.
Tin

 

[John11668]

On this project I think the plan is just to make things move under CNC, and I am not really bothered if it is a bit slow.
As I will probably be machining steel I expect I will have to go a bit slowly anyway and I rarely use the faster traverse rates the machine has anyway.

So coming back to principals
I have measured the force to rotate the 4inch hand wheel at a number of positions along the z axis.
Varies a bit but generally breaks away before 8 lbs so if we allow for 10 lbs on a 2" moment arm we would seem to need 320 oz.in.
Gearing down say 5/2 with timing pulleys (ignoring losses) and we would seem to need 128 oz. in.
This would seem to be well within the torque capacity of a Nema 23 1.8 Nm stepper, with lots in hand to cover stiction and cutting forces.
http://cgi.ebay.co.uk/Stepper-Motor...ial_Automation_Control_ET&hash=item3a5ee49977

So this bit of kit takes 2.8 amps on unipolar , 4 in bipolar parallel and 2 in Bipolar series. Presuming the drive has a capacity of 2.5 amp and I presume I can set a limit on the board then am I correct in presuming that this will just limit the torque and as I would seem to have plenty in hand there may be no problems.

Which wiring mode would I use mode. Can I use bipolar series ? Would the drive reverse the motor in this mode
Do I cause myself any problems using a larger motor than required

 

[Tin Falcon]

stepper motors are normally run direct 1:1
bipolar series IMHO would be the way to go .
I used 320 oz motors direct connected with flex couplings .
Tin

 

[kvom]

While you may machine steel at 10-12IPM, you will appreciate fast rapids(G0). My mill has a conservative 70IPM rapid, but with tweaking of the PC settings they would be no problem in driving at 100+. For aluminum, milling rates of 30IPM are pretty common.

 

[jpeter]

You can do what I did: Contact Dan Mauch at Camtronics-cnc.com. I discribed what I had and he told me what I needed to convert my Bridgeport. He sold me the stuff. It all worked. Prices were good too. I had no clue as to motor size or type or anything. I'm smarter now but not when I started. Somebody said steppers were generally direct drive. Not mine, I had steppers on a Grizzly mill. 2:1 timing belt drive. Steppers need some type of flex coupling. A friend of mine uses a short piece of rubber hose for 1:1 coupling.