Trying to learn... And looking for a CNC Lathe

John is correct.

The Warco WMT 300/1 I have is a lovely little machine and I have learnt many of its ways. There is a lot of backlash: possibly 0.5mm at the headstock end of the leadscrew and 0.25mm at the other end. It means that I have to machine the same way every time, withdraw the cutter, move back etc. No great shakes, so long as you know AND DON'T FORGET (guess who has managed to wreck several threads when he did..   )  . I expect to use the same method of cutting on the cnc.

I really need a repeatability of 0.025mm and a theoretical precision of better than 0.05mm. I measure most of the diameters of my pens with a 0.02mm precision vernier caliper and try to machine everything so it's correct by that. You can feel a step when the diameter of two mating parts are wrong by 0.1mm even after polishing rounds the step. In fact, it's usually better to machine concentric parts togther in their final position, because slight errors in concentricity are really noticeable.

To give an idea: I machine most of the bits using an ER32 collet. The runout on the collet chuck I have is less than 0.01mm according to my dial gauge (the needle moves, but less than one segment, and it's not much worse than dialling the headstock on its own), however when one concentric part is machined one way and fitted into the end of another part machined the other way in the chuck, it's possible to feel two areas where there is perfect alignment and two areas where there is a step (one step on one part, the other step on the mating part). A pen is being used by the most sensitive part of your body, so exact dimensions are not hugely important but differences in sizes are.

However, that is not the biggest problem. Not by a long shot. The biggest problem by far is the flexibility of the plastics used. I have had, while doing rough turning (with 0.5mm deep cuts - any more & the plastic chips or melts with carbide cutters and doesn't cut at all if coolant is used), on a 105mm long barrel using a dead centre, the ends came out at 12.00mm diameter, and the centre was 13.96mm diameter. Purely due to work deflection over only 105mm. I have since realised that a steel mandrel down the pre-drilled core stiffens the barrel up by a factor of several hundred.

Now I want to use CNC to get exactly that sort of curvature, but in a controlled & repeatable manner. Silly really. Sigh.

Regards,

Richard.

edited to remove unexpected smiley

Edited By Richard Williams 7 on 04/06/2013 11:39:36

For plastics you don't want to be using carbide, typically there is far less back rake and cutting pressures are higher thus pushing the work away from the cutting edge. HSS steel would be better, and grind exagerrated reliefs all around (think a knife blade whittling wood!). You cannot take big cuts with a tool like that as it will tend to dig in and pull the tool into the work (and vice-versa), but for shaving/finishing cuts on soft materials it will give you the most accurate result. Needless to say the cutting edge has to be razor sharp, and it will be fragile. If you recognise its limitations such a tool still can be useful. I use such a tool for cutting PTFE (carbon/bronze-filled PTFE) when forming seals and piston rings.

Yikes!

Richard

Hmmm - around where I work, Satellite parts are made and the whole things tested; no space shuttle parts made for a while, for what (I hope) is an obvious reason. I won't say anymore about that. Lunchtime conversations can be quite in-depth, as one might imagine. And the machine shops and machinists are pretty good - again, as one might imagine.

The main problem as I see it is to how best to "dumb-down" the technology so that others can use it. The fact that Richard and others are contemplating CNC means that the "dumbing down" and "making it work for the man on the street" is happening.

Much like LBSC or Martin Evans made model locomotive building easy for the beginner, we do need beginners guides for CNC - and of course, those more interested in the details can go further.

Another JohnS

Richard - if that's silly, then every tradesperson who tries to excel in their work fits into the same category.

Keep going - I, too, enjoyed reading the info and links posted.

Another JohnS.

John AS,

I have been surprised at how comprehensive & relatively simple the stuff available for DIY convertors is. When I last seriously looked at CNC machinery I was in my final year of my degree course (1987) & my project was to write a translator between the institution's CAM system output and the Bridgeport cnc mill's input. The concept of the G code language was still far off. The machinery was bespoke & terrifyingly expensive - and students were firmly not allowed to touch any of the machinery because it was worth more than any dozen students! The electronics and processes used were enough to frighten me (a mere 'mechanical' engineer) off for years.

I have long felt that home workshop technology is about 50 years behind industrial technology, but in the case of CNC machinery, it might be much less. The driving software (Mach3/ CAMBAM/ EMC2) are far from mickey mouse bits of software and have certainly provided me with a considerable degree of re-assurance that I'd be able to get on with what I want to do (make pens) rather than be hand coding each move. Not that hand coding would have stopped me, just made life more difficult and slowed everything down.

Regards,

Richard.

Far, far less.

Although I have owned commercial CNC machines for 20 odd years if someone at that time had told me that I could easily afford to have a CNC mill, lathe, laser cutter and 3D printer [ which didn't exist then ] in a little shed they would have been burnt at the stake as heretics.

This interest in itself has spawned a whole sub industry where for many the process of CNC and relevant process has become their hobby and interest.

CNC Zone has 311,000 odd members with over 1,233,000 post so far.

I dare say no other site has this level of traffic ?

I will confess I find CNCZone a little daunting, but it is a resource to be mined on occasion.

As a moderator on the Fountain Pen Network I'd hate to have to do the same job on CNCZone, even though the CNC Zone traffic is much lower (FPN has 2.6 million posts, 77000 members and a daily traffic of 35Gbytes - as I write there are 1370 people looking FPN). Just the plethora of forums (325 + about 100 sub fora) is mind boggling, and the chance of missing something relevant is very high.

There is a feel, though, from a lot of the posts that they are from professionals in 1-2 person small businesses trying to tap into external expertise as well as hobbyists.

I agree that the power & speed of introduction of cnc machinery into hobbyist's workshops is mind boggling. However, I think at the moment those who have it are still ahead of the curve, and it may become more common in the future. It is, at the moment, something out of the ordinary to have cnc. OK, it's also out of the ordinary to have metalworking machinery at home, only two of us out of 60 in the engineering company I work for have any form of metal working equipment at home. (There's nothing quite like crippling your own argument just as you make it, is there? )

Regards,

Richard.

Well, I must thank everyone who had a hand in this discussion, it was enlightening and helped me a great deal. I changed my mind completely several times about what to do, but came to a decision in the end. I do acknowledge all advice, though on reflection I didn' follow much - it all aglomerated into a generalised confidence that CNC is not novel or frightening, which was probably the most useful help you could possibly have given. Thanks.

I eventually plumped for a 4 axis chinese 6040 CNC engraver rather than convert a lathe. The price was a bit less than doing the conversion, but that wasn't the main reason for doing so. The biggest reason was that I really am not electronically minded, and if it didn't work first time, I'd be at a loss as to how to cure the problem and would have no idea as to whether the problem was my work, faulty electronics or the programs not talking to each other.

The machine is a moderate sized 600 x 400 mm bed arrangement with an 80mm chuck on the 4th axis. Maximum part height is 117mm under the gantry, which is OK for what I need. I can set the chuck up on part of the bed while leaving much of the bed free for making boxes or thread milling parts that need it.

You will notice there is no PC next to it in the photo - it was taken as I had finished plugging/screwing the bits of the machine together.

The machine can cut aluminium at 100mm/min with a 0.5 deep cut and plastics are also feasible. Timber can go at 5-600 mm/min. The flex is 'adequate' as is the speed; as someone else said, I'm trying to make pens not a Space Shuttle!

As with many Chinese machine tools, the setup process has been an education. Firstly the manual was wrong in several details (not least of which was making the Z axis move in the reverse direction) & didn't mention the 4th axis at all. When I queried the supplier on the number of steps per degree for the A axis, the supplier told me the Mach 3 default (640) was correct. Ho hum, 2 hours later a combination of trial, error & calculation had the correct 177.77777 steps per degree. The bed isn't quite perpendicular to the Z axis and I'll need to fit a few shims under one end of the bed extrusions. However it generally works pretty well.

The learning curve has been .. no, not steep. I'll try & find a more suitable word... Vertical. A cliff. I may soon be at the point where I can get something useful out of it. To begin with I'm going to make a box for the pen, then start on the pen itself. Hopefully I shall get something worthwhile in a month or two - there is so much to learn.

Regards,

Richard

Edited By Richard Williams 7 on 24/07/2013 15:00:55

Edited By Richard Williams 7 on 24/07/2013 15:03:37

Edited By Richard Williams 7 on 24/07/2013 15:04:49

Edited By Richard Williams 7 on 24/07/2013 15:06:11

I have had a fun time in the last week trying to track down a problem.

I have been getting spurious Z axis feed. Initially I thought I hadn't tightened the collet properly. I now know I have, and tracked the problem down to the Z axis feeding downwards spuriously. If I slowed the spindle, it got worse. When the spindle was off, no spurious feed. Ho hum. It's noise from the VFD. There is a choke, so the problem will be the unshielded cables.

I have done a search on te Web & one suggestion, in a thread where Big Shed contributed, was to replace all the cables with screened professional Microphone Cable, 4 core, with each core being 11 strands of 0.2mm^2 and the braided shielding being tinned 112 strand/ 0.12 mm^2. The suggested supplier was Jaycar, who seem to be Australian.

Is there a UK based supplier of an equivalent cable who sells by the metre?

I have looked at ArcEurotrade, but their cable is unshielded, so I'd have the same problem. The cable from RS Components & Maplins seem to be 7 strand x 0.2mm^2, so I'm not sure if it'll be up to the job.

Can anyone suggest a supplier to try?

Regards,

Richard

Try Canford Audio

It doesn't necessarily follow that the problem is unshielded cables. I assume that we're talking about a mains powered VFD, with the choke on the mains input? I doubt there would be a choke on the output of the VFD. Look and see if the cables from the VFD to the spindle motor, and to the Z-axis motor (presumably a stepper) run close together. If so separate them and see if it makes any difference. If it does, then the problem may well be airborne interference, as surmised.

Personally I wouldn't use microphone cable for either the VFD output or stepper motors. Microphone cable is normally a controlled impedance cable and will not be rated for the voltages or currents involved.

For the output of the VFD you would ideally need a shielded cable rated for mains voltage.

Regards,

Andrew

Andrew,

There is a choke on the output of the VFD, not the input. It's positioned about 2" from the VFD where the wires are still separate from the outer insulation.

The X, Z and spindle cables run next to each other in a cable chain along with two water pipes. I have not had any trouble with the Y axis which is separated from the cable chain. The X axis has had some trouble, but less than the Z, and it has a shorter run adjacent to the spindle cable.

I'll try this evening to unthread the cable from the chain to see if makes a difference.

Baldric,

Thanks. I'll take a look.

Regards,

Richard.

Edited By Richard Williams 7 on 30/07/2013 16:09:51

Zapp Automation sell screened 4 core power cable. I have bought stuff from them and they are a good supplier.

I think the "choke" is an in-line filter - a choke on the 3-phase output of a VFD would not work very well! (Some variable DC supplies for commutator spindle motors, such as are fitted to Denford mills, have humungous chokes in series with their outputs but these are smoothing DC not AC.) By the look of your photo, your spindle uses one of the the high-frequency 3-phase induction motors as are supplied from China - water cooled, right?

Frankly I doubt that noise coupled from the VFD cable into the stepper cables would directly generate spurious pulses, but it's worth trying screened cable. Many of this kind of problem are caused by earth loops - too many earth connections.

Thanks John. I'll have a look at them too.

One thing I would like to get a handle on, though, is why the spurious steps would stop entirely the moment the spindle stopped if it wasn't related to noise. And why the number of steps seemed inversley related to VFD frequency, coming to a maximum just before the point where the VFD switched itself off (at almost maximum speed there was no discernable z axis movement). I continued the program with air cuts to the end, and there wasn't a single jitter from the Z axis motor. Had it been a broken wire, I'd have expected the Z axis to continue moving regardless of whether the spindle was running or not.

Regards,

Richard

I'm not saying it isn't noise, but it is perhaps coupled more through earth loops than through "radiation" from unscreened cable. The fact that the speed of spurious steps gets higher as the motor slows down might be because the in-line filter gets less effective at lower frequency out of the the VFD. Why did you mention a broken wire? But I have had very peculiar things happening on my cnc mill when wires have come adrift, with quite the wrong axes moving and so on.

Reading a bit about these 6040 routers and having had dealings with a couple let me tell you of known faults.

Firstly the motor wiring to all the stepper motors is crap, best to replace with decent wiring making sure you get good joints at both ends.

There is no need to use shielded cable on the steppers but the VFD and spindle MUST be run in screened cable.

Only earth one end of the screened cable.

The board fitted to the 6040 is a cobined breakout board and 4 drivers all on the same card and it's really built to a price.

There are some web sites that tell you how to modify them to make them more reliable but to be honest get ready to replace this board with seperate stand along drivers for reliability.

They are very good value for the money but do need a bit of care like a lot of this Chinese kit but once done you do get a decent machine.

John S,

I must admit that was my take on the machine having done some research & I did go into it with my eyes open. Whether the brain was engaged at the same time is a question I refuse to answer...

I am being very careful to ensure that nothing comes to a crash and stalls any of the motors - that's where I expect the most risk to the BoB. I hope I can get it to last 2 years, if it can, then I think it'll have done its job & I won't be unhappy if it then goes futt. Obviously, I'd prefer it not to.

I am not sure which end of the spindle power cable should be earthed. I would expect it to be the end closest to the spindle, then the end furthest away from the Z axis could be radiating with minimum affect on the Z axis.

John H,

I mentioned broken wires because quite a few people who've had these machines seem to start talking about them after a couple of months! I gather the wires are not very good.

I did try the machine having removed the spindle motor cable from the cable chain, and separated it as far as possible from the motors & other wires. The only noticeable difference was that the low speed interference with the Z Axis was greater than before, causing the Z axis motor to feed at about 30 rpm with the VFD speed set at 25 Hz/1500 rpm.

Regards,

Richard.