Thinking of jumping into CNC milling. How to start ?

[SparkyLabs]

Hello. I run my own one man band company, I design and make small volumes of electronics or have the boards made out where quantities permit (above 25).

Most of what I do needs to go into a box of some kind that invariably needs holes cutting. Machinists ans box manufacturers have long lead times and of course want tooling lcharges. So I'm figuring that if I get a small CNC myself I can have the benefit of production quality sam[les and prototypes and small quantities are viable. for example I have a project currently where i need to make 50 and tooling is £95+ VAT yet each box is only 80p to machine and I'd have to wait 7 weeks.

I have set my eyes on a Proxxon FF 500 / BL-CNC machine which also cmes in a manual version. CNC is £3663 + VAT and the manual one is £1392 + VAT, so they want around £2200 for motors, controller and I guess the biggest cost is software.

My question is, is this a good starting point ? It seems a reasonable amount of money although that is mighty expensive software.

Should I consider any other options ? is there open source software worth considering ? (there is some question as to wheather the machine comes with software).

[m_c]

For the £2k, you're getting ball screws (the standard machine will only have acme lead screws), and lots of extra bits that allow motors to be fitted, along with a controller and some form of software.
Most ready to run CNC mills of this size are pretty expensive, as they're not produced in the same quantities as the manual machines, and there are increased support costs.

Personally, as you say electrical boxes, I'm assuming plastic/die cast enclosures?
In which case I'd look at router style machines, as they'll be far more suited, provided they have enough room under the spindle.

Software to run machines varies from free to expensive, but it all depends on what controller the machine uses. LinuxCNC is open source, and reasonable good. Windows has lots more options. Mach 3 or 4 is one of the more universal options, but there are lots more that are specific to certain machine controllers.

However you also need to factor in software to generate the required G-code that the machine software accepts, which can range from free to extremely expensive, depending on what you'd like to do.

[SparkyLabs]

I see, I thought the manual one would be the same and that they just swapped handles for motors but as i said I am a total noob here. Yes I understand the potential value of all the bits and I thought it is a good machine and a good price but as I have not ventured into this sort of thing yet I thought I'd ask.

Yes I am doing ABS cast boxes. What worries me with a router is the height indeed. The boxes I am currently looking at need machining on the side, they have draft angles so by the time you have one fixed in a vice on the bed and the tool in the machine you sudenly find you need 150mm of working height for a 120mm box. I'm not sure if this too tsll for a router. I'd also like to be flexible for the future. Currently I have a clarke CMD10 but it is painful to use.

[magicniner]

The CAD/CAM learning curve is not one to be underestimated, I bought a desktop CNC mill then spent 18 months before I could reliably produce 4-axis code for it ;-)

- Nick

[SparkyLabs]

Well I would prefer a ready to go solution. if this is what is costs then that is fair enough I just wanted to make sure I was going the right-ish way about it.

[Clive S]

Are you just wanting to drilling holes or wanting to cut out shapes etc. As drilling could be done on a manual mill.
A picture is worth a thousand words of what you are trying to achieve.

[SparkyLabs]

Holes and shapes. But even getting holes right on my manual mill is tedeaous because it has no readouts and the backlash is terrible plus my lack of skill. I am looking to do production quantities on this machine to get it to pay for itself along with good fast prototyping.

[SparkyLabs]

There will also be plenty of cutting down of potting boxes, the bigger they get the taller they get and this can become a problem as what I need are more like trays so the only thing I can do it to cut down existing boxes.

[magicniner]

what I need are more like trays so the only thing I can do it to cut down existing boxes.

That sounds like a job for a bandsaw with a suitable guide to get them within a couple of millimetres and a quick pass with the mill to finish ;-)

[John S]

Might pay to have a look at one of the blue A3 style laser cutters on Ebay for about £1500

I used to mill boxes out for the division master style hand held controllers and one case too about 15 minutes with some rejects where the plastic had welded to the cutter and ruined the box.

I now do these on the laser in about 2 minutes with no rejects

[SparkyLabs]

Do you have any links to the laser cutter. What are the limitations of laser if any ? I could of course end up cutting metal too. With plastic I am hoping that a nice sharp cutter not too much speed will prevent melthing. I've not had any problems hand milling with melting, just the positional accuracy that take so long to get right that it would not work for repeats.

[SparkyLabs]

That sounds like a job for a bandsaw with a suitable guide to get them within a couple of millimetres and a quick pass with the mill to finish ;-)

Well I am based from home so don't have room to set up a complete machine shop. I need a bit of a one size fits all machine, granted some jobs are not the most suited but if they can be carried out with no severe downsides then that is fine.

[magicniner]

I have a bandsaw that's just 24" tall with an 18" x 12" footprint that can be used on a Machine Mart bench grinder stand, it has a plethora of uses, all of which save time and some of which save material ;-)

[SparkyLabs]

Well here and now I need to be able to first cut holes and slots and other shapes in the side of cast aluminium and ABS boxes. As I would not be able to afford any further machinery after that I'll have to use the mill for the box cutting down. Of course large quantities can be done by the box manufacrurer if I am not equiped.

[Clive S]

I would suggest that you try and contact some one in your area to see a mill working so you will have some idea as to what is involved before you start spending the hard earned dosh.

I for one would be more than happy to show you if you are ever up here (Stockport)

[SparkyLabs]

Well I also have a full time job where large CNC machinery is used. I'm not sure how specific the workflow is to various machines and how much is a standard process. I do 3D CAD with no problem as that is the day job, it's really a case of how I take that data and get it into a program that controls the machine.

I know the proxxon software allows you to import a DXF, that is something I can generate with no problems.

[JAZZCNC]

Yes I am doing ABS cast boxes. What worries me with a router is the height indeed. The boxes I am currently looking at need machining on the side, they have draft angles so by the time you have one fixed in a vice on the bed and the tool in the machine you sudenly find you need 150mm of working height for a 120mm box. I'm not sure if this too tsll for a router. I'd also like to be flexible for the future.

What you need is Custom built router.! (Wonder by who.:whistle:)

This will give you more scope than mill and will easily handle plastics and aluminium. You'll get much more productivity because will be able to mount several boxes at time on fixture jig and cut quicker.
The faster spindle speed of router suits cutting aluminium better than slow spindle you'll get on small Mill (4000rpm). Cycle times will be much faster and will allow for other jobs like engraving etc where higher spindle speeds are more desirable.

Custom built router will knock the spots off Small mill for this type of work. The extra cutting area is always welcome and if done correctly will give much more scope in cutting anything upto aluminium.

Honestly think you'll regret buying that Proxon CNC and will soon outgrow it or find it's limits.

[SparkyLabs]

oh, so do you build such routers ? what happens about software in this case ?

I have to admit that I'm still taken with the idea of a good general purpose machine. steel machining is not totally out but not a priority at the moment.

[JAZZCNC]

oh, so do you build such routers ? what happens about software in this case ?

I have to admit that I'm still taken with the idea of a good general purpose machine. steel machining is not totally out but not a priority at the moment.

Yes I do. I'll PM you my number then if want to chat about them or if just want advise in general feel free to give me call.

Regards cutting steel then I'll always advise you go with Mill but if your talking general purpose cutting of all materials then the size type of mills your looking at will struggle more than well built router.! . . . Hence why I said think you'll regret buying Proxon Mill

IMO if you want to cut Aluminium and harder then buy Mill. However be aware that for aluminium because of slower spindle rpm's it will require HSS tooling and will take much longer than it would good strong router with Carbide tooling.
For softer materials upto Aluminium/Brass and 3D type work then Good Router will always win by the fact it allows more scope and faster feedrates.

Re: Software.

Don't get too hungup on software or be fooled by Bull company's like Proxon peddle. Most of the software provided is basic in what it can do. So what happens is you still end up buying third party Cam software which then spits out G-code you load back into the control software which runs machine.

To run CNC machine you basicly have Control software which does what it says on the tin.? It controls the machine and tells it where to go by processing G-code file which is loaded into it.
The G-code file is created by External CAM program which takes either 2D Vector Graphics(dxf etc) or 3D models and using options provided you create cutting toolpaths which then get processed into G-code file.
The DXF's or 3D models can come from CAD or if using CAD/CAM software created directly inside same software.

So to answer your Q's about what happens with software on machines built by me.

I provide the Control software which in most case is Mach3 which will be all setup and ready to run the G-code file. The CAD/CAM side I leave to the customer to buy/decide because depends greatly on what type of work your going to do to which suits best.

I'll gladly advise and often I'll spend time helping new users with the software. Because I've got lots of experience I've used most software and it's little like CAD in they all work in very similair ways so often can get people confidently producing G-code quickly.

If you are used to CAD then you won't have any trouble with CAM. It's basicly chosing the type of cutting stratergy ie: Drilling, profile,pocket etc then selecting lines/arcs or Surfaces for path you want cutting then selecting tool and filling in few parameters.
When all toolpaths are created you then output them via Post processing file which creates G-file which suits the machine control your using.
This Post processing stage is seamless inside Cam software but very important part and often over looked. Without correct post processor you'll get G-file that is garbage to Control software.

[SparkyLabs]

OK well might be worth having a chat with you. Essentially I design the eclosure cuts in 3D CAD (solid edge). My circuit board design software will spit out a 3D model of my board loaded with models of connectors etc that I put in. I assemble the circuit board model into the box in solid efge and can therefore make the holes in the box model with reference to the circuit board components. So from that I can produce a drawing and save that as a DXF. From that I need to get onto a machine that can cut the holes.

With making fixtures it's a toss up with fixture cost. I'd need to hold boxes up on their sides in two positions regarness of the draft angle they sit on, for 50 boxes that I can get machined for £95 tooling + £0.80 per box spending pounds on fixtures soon becomes pointless. I take it a router has a similar bed to a mill so that stuff can be bolted down ?

[m_c]

Fixtures for boxes should be relatively easy, and reasonably universal.

The hardest part will be coming up with some form of clamping strategy to give good support. If it was me and I was planning on doing lots, I'd be looking at some form of large open vice like setup, which allows the box to be stood on end/edge, and clamped on what would be the top/bottom (or front/back depending on if you class the lid as top or front).
Add in an end stop for locating the box against prior to tightening, and you've got a repeatable clamping setup.

For doing the clamping, some form of toggle clamp setup would allow for quick changes, but would need someway of being adjusted for different box sizes. If you really wanted quick changes, an air cylinder and valve would make things quicker, and cover a larger range of sizes without adjustment.

[magicniner]

Fixtures for boxes should be relatively easy, and reasonably universal.

The hardest part will be coming up with some form of clamping strategy to give good support. If it was me and I was planning on doing lots, I'd be looking at some form of large open vice like setup, which allows the box to be stood on end/edge, and clamped on what would be the top/bottom (or front/back depending on if you class the lid as top or front).
Add in an end stop for locating the box against prior to tightening, and you've got a repeatable clamping setup.


A two piece machine vice can offer a lot of flexibility with infinite variability in jaw spacing and the option to mount on riser blocks,

- Nick

[JAZZCNC]

For doing the clamping, some form of toggle clamp setup would allow for quick changes, but would need someway of being adjusted for different box sizes. If you really wanted quick changes, an air cylinder and valve would make things quicker, and cover a larger range of sizes without adjustment.

Like this you mean.!! . . . . Also Shows good example of why router is better suited than mill.


https://www.youtube.com/watch?v=HBrgn62OmHs

OR if prefer in aluminium.


https://www.youtube.com/watch?v=lBjpH2HQYPc

[m_c]

Like this you mean.!! . . . . Also Shows good example of why router is better suited than mill.


That's maybe a little too advanced for Sparky, but certainly shows how the commercial guys do it.

One thing I did mean to mention, was you mention you get charged a fixture cost. It should really be referred to as a setup cost, as what you're really paying for is the setup time for the job. That includes checking your drawings, working out what tools are needed, converting them to whatever code the machine runs on, and setting the machine up to run your job (load the code, setup the required fixtures, and make sure the correct tools are loaded). Once the setup is done, it's just a case of loading boxes and hitting a button.
The actual fixture cost for somebody running these kind of jobs regularly will be very minimal per job. The fixturing shown in the videos Jazz posted will cost 4 figures, however that will cover lots of boxes with minimal changes between box sizes. It can be thought more of as a machine cost, rather than a job cost.

[JAZZCNC]

The fixturing shown in the videos Jazz posted will cost 4 figures, however that will cover lots of boxes with minimal changes between box sizes. It can be thought more of as a machine cost, rather than a job cost.

From Datron Yes but here in the real world same could be made for not lot of money. But this may better showtype of fixture I was meaning. Also one to see the difference between Mill and router.!!


https://www.youtube.com/watch?v=BWB2pEFZH-4#t=143.701

Almost slow motion but not quite.!


https://www.youtube.com/watch?v=FV59Yx2njY0

[SparkyLabs]

I don't particularly need high speed or large batch manufacturing. I'm only wanting to do prototypes and very small quantities. I think too many boxes stacked up could increase the risk of error but even on a mill a few at a time dould be loaded. I have a box face of 90x30mm and a mill with a travel of 100*290mm can do a few.

[SparkyLabs]

ultimately flexibility will serve me more than speed. If I find that I need to do a particular job in large volumes and that a specific machine is best for this then specific machinery may be justified.

[m_c]

Those videos were mainly to show you how things can be done, and to give you ideas on how to mount boxes.
We know you're not likely to ever use such a big/complex setup, however it should give you ideas on how you could mount a single box, while making any fixtures as universal as possible.

If you look at the setup in the first videos, they use an extruded aluminium T-Slot bed to mount pneumatic clamps on. Using a T-Slot bed gives you easy adjustment, and the pneumatic clamps mean you get a wide clamping range so they don't have to be setup exactly. They just have to be setup to give enough clearance to get the boxes in, while making sure they clamp before running out of travel.
The pneumatics has the major advantage that a single valve controls the clamping of all the boxes.

To simplify that, use a T-slot base for the flexibility, and then use a two piece machine vice setup like magicniner has mentioned (first google hit - http://www.bison-bial.co.uk/vises/6522-200 to give you an idea of what we mean). Add in a block to the side of the vice to position the box prior to tightening, and you've got a repeatable setup for if you need to multiple runs.
With that kind of setup, you should be able to cover a very wide range of boxes.

[SparkyLabs]

Any comments on the Wabeco F1200 CNC ? this is like the proxxon controlled with nccad

[magicniner]

"by connecting the WABECO controller to a PC and installing
the nccad turning software the customer turns the conventional
WABECO milling machine into a CNC milling machine"

I hope they do better with mills than with proof reading because you won't get far with a mill if you use turning software! :-(

[SparkyLabs]

Well the specs are slightly better than the proxxon (that also has dubios english and if you read the nccad website you will faind yourself reading worse gibberish, I guess german machinists don't have great english), they are offering an ex demo machine for around the price of the proxxon but it looks like as a machine it is built to a higher standard. One reveiwer on axminster (UK distributor for proxxon) recomended the cc f1200 over the proxxon machine he had bought second hand.

Proxxon offer 0.05mm repeat accuracy,

webaco offer 0.015mm "positioning accuracy"

I'm noty sure what they mean by "truth of rotation of the tool spindle" of 0.01mm

[SparkyLabs]

oh and the webaco has a standard MT2 spindle, not sure what the proxxon ff 500/bl-cnc has.

[SparkyLabs]

The manual has better english: http://www.emcomachinetools.co.uk/image/data/f1200toccf1210hs.pdf

They meant concentricity of the spindle is 0.01mm

[magicniner]

standard MT2 spindle

You can get ER Collet Chucks on straight shanks in smaller sizes and thus have an ER Collet Chuck without a huge increase in tool protrusion from the spindle, it's a shame they didn't go with BT30 though, that would have supported a tool changer if they implemented a decent draw-bar system.

[m_c]

truth of rotation, will be spindle run out.
Might have to add that one to the dodgy chinglish translation thread.

By the looks of the proxxon, it's got an ER20 spindle.
Although a MT taper will give a bit more flexibility, I doubt in this size of machine you'll have enough spindle torque to handle anything that an ER20 collet couldn't handle. ER20 can handle up to 13mm cutter shanks, and I'd doubt this small a machine will make much use of any cutter bigger than 4-5mm.

[SparkyLabs]

Well the wabeco is 1.4KW versus 400W of the proxxon so I guess they meant it to be a beafier machine. The only pitty is that maximum spindle speed is 3000 rpm but then like I said high speed work is not a concern because I'm only prototyping and doing small batches.

[magicniner]

Well the wabeco is 1.4KW versus 400W of the proxxon so I guess they meant it to be a beafier machine. The only pitty is that maximum spindle speed is 3000 rpm.

There's a 2kW spindle option listed in that linked PDF that shows max rpm as 7500 which sounds much more useful.

[Clive S]

he only pitty is that maximum spindle speed is 3000 rpm but then like I said high speed work is not a concern because I'm only prototyping and doing small batches.Yes but with small cutters you need higher rpm that is why we tend to use the Chinese 24K rpm spindles. The water cooled one are very quiet.

[SparkyLabs]

Well I spoke to jazzcnc today and done some quick thinking.

I think I will just add digital readouts to my crappy Clarke CMD10 to make life simpler and save for the cc-f1200 HS with 7500 rpm.

Sent from my phone so mind the autocorrect.

[magicniner]

My desktop CNC milling machine came with under 4krpm top speed from a DC motor which was very disappointing in use, I converted it to 3 phase which gives me up to 7krpm, for smaller work I built a simple speed increaser which gives me up to 30krpm but ideally buy something which covers as large a range of work as possible both speeds and envelope because jobs will come along which you could do "if only" ;-)

[Clive S]

Well I spoke to jazzcnc today and done some quick thinking.

I think I will just add digital readouts to my crappy Clarke CMD10 to make life simpler and save for the cc-f1200 HS with 7500 rpm.

Sent from my phone so mind the autocorrect.

Have you read the specs. ( re-adjustable free of play trapezoid screws in all axes ) no ball screws and only max 600mm/min rapids

edit I see that they are optional extra for about £1000

[SparkyLabs]

OK, no not re-read it since I spoke to jazzcnc but yes ball screws are a must. I'll stick with the manual machine and look into the CNC later.

Sent from my phone so mind the autocorrect.

[Davek0974]

Sent you a PM sparkylabs, may be of interest ;)

[SparkyLabs]

Is that the one you sent this morning yes read it. Else it's Tapatalk playing up.

Sent from my phone so mind the autocorrect.

[Davek0974]

Yep, just the one ;)

[SparkyLabs]

OK. Well I'm still not sure which way to go and had a good chat with jazzcnc this morning which was most helpful. I think if I put DRO's on my current manual mill it will make it useable for prototyping while I learn more about machining and CNC and save up and then buy something decent if it makes sense for light manufacturing.

Sent from my phone so mind the autocorrect.

[magicniner]

Longer spent evaluating and planning is rarely lost time.
There's a huge range of things that a decent DRO with multiple zeros renders relatively simple on a manual mill.

[SparkyLabs]

Well my whole looking into a CNC mill started with my frustration over the manual machine I use. I looked at DRO's but it became apparent that for some you can start spending hundreds of pounds although I'm a bit confused about the claims of you been able to cut an arc or circle with one of these, typically the ones on eBay have this claim.

So I didn't want to spend lots of money on the current machine knowing that I will outgrow it and started looking at just getting a CNC. A work colleague found me the proxxon unit however this has turned out to be a bit of a Schrodinger's cat as while the mill exists or apparently exists I will never know if the CNC version exists until I try to order one and see if it actually turns up. My first attempt at getting a quote got me a quote for a manual machine twice in fact this happened. And from this discussion it is clear that there is far more to this than I had realised so I should probably learn about what the hell I am doing and the equipment before buying something that turns out to be no good. I can only but spend the money once.