Member postings for Tony Jeffree

Posted by EtheAv8r on 29/06/2011 13:00:09:

Edited By EtheAv8r on 29/06/2011 13:02:21

Sorry about that, John...

 

Regards,

Tony

Posted by John Stevenson on 28/06/2011 22:31:09:

Not the boys that got "Santa's Slaves" tattooed on their chests by a dyslexic tattooist, by any chance? <bg>

 

Regards,

Tony

Posted by Steve Garnett on 28/06/2011 22:05:45:
Oh. I'm using release 5!

I am also using FF #5 and all the images show up just fine...

 

Don't you just love PC's?

 

Regards,

Tony

Posted by EtheAv8r on 28/06/2011 11:20:32:

If you are reasonably familiar with PCs and Windows, getting the hang of a drawing package is fairly quick. Once you have the drawing in CAD, you save the drafing as a DXF, import that into the CAM package, where you select how it will treat the various parts of the drawing (for example, if you had drawn a circe, you might intend that to be a simple circular toolpath, or the outer edge of an "area clear" operation). The CAM program generates the Go-code, which you load into the CNC control that interprets the G-code and translates that into coordinated moves of the various axes.

 

Kit-wise, the expensive way is to buy a purpose-built CNC mill (for example, from Tormach, Taig, or Arc Eurotrade). The cheaper way is to start with a manual mill and convert it; Taig can sell you a "CNC-ready" mill that has the necessary motor mounts, to which you would add your own motors, drivers, etc. An example Taig conversion is documented here:

 

http://www.jeffree.co.uk/pages/taigcncpt1.htm

 

and here:

 

http://www.jeffree.co.uk/pages/taigcncpt2.htm

 

although these days I would probably use Mach 3 instead of Desk CNC, so the details of the electronics would change a bit.

 

There are ways of making the end result usable both manually and under CNC control. The simplest way is to use stepper motors that have double ended shafts, and transfer the hand crank to the back end of the motor. However, with Mach 3 there are several options for adding hand controls (so-called "pendant" controls) that give you manual control of the mill via the motor drive system.

 

Regards,

Tony

John -

 

Makes the point very nicely.

 

Another example of a 1-off that was done using CNC, and which would have been very time consuming indeed using manual techniques, involving a steep learning curve in hand-engraving just as a starter:

 

http://www.jeffree.co.uk/pages/Chapter-ring.html

 

Again, not a single line of G-code was written during this process. As observed in the text, I could have made a more traditional-looking job of the numerals; with a larger table available to me on the mill (or maybe a small CNC router), the answer would have been to draw out the numerals "properly" in CAD rather than using a rotary axis and the in-built Roman fonts.Also, the CAD/CAM tools available to me have much improved from the rather primitive SuperCam package I was using back then, so life is even simpler now.

 

A later example is this brass plaque that I engraved using a CAD drawing for the outline/screw holes and DeskCNC's in-built stick fonts to engrave the text:

 

Again, doing the same by hand would have been impractical both in terms of time and personal ability.

 

One-off work is just as legitimate a target for CNC tools as is repetition work, and IMHO the former is much more likely to be useful in the average ME's workshop.

Regards,

Tony

Posted by Peter G. Shaw on 24/06/2011 15:43:45:

Peter -

 

If there was absolutely no backlash in the leadscrew drive, you wouldn't need the flip-up tool; you could reverse the drive without retracting and all would be hunky dory. The whole point behind the flip-up tool is that it allows you to reverse the drive in a machine that does have backlash. However, if you change the way it pivots from flipping up (as in the original design) to swinging left (as in Chris's suggestion), it doesn't give you the desired result at all.

 

Regards,

Tony

Posted by chris stephens on 24/06/2011 12:23:58:

Chris -

 

I think the flaw in this particular variant is that if there is any backlash in the leadscrew, when you reverse the spindle the tool would want to rotate to the right (i.e., against the stop) and would therefore not flip out of the way. So this would effectively be the same as the situation you would be in if the tool wasn't pivoted at all.

 

Regards,

Tony

Posted by harold on 09/06/2011 09:13:54:

It doesn't really exist. Photoshop can do some wonderful things these days...

 

Regards,

Tony

Posted by David Clark 1 on 09/06/2011 12:58:07:

 

Fraid so David.

 

But then, John told me how he did it the other day - and a very useful technique it is too. When he first mentioned the problem to me, my immediate reaction was "Use Moglice", but that wasn't how he did it either.

 

Regards,

Tony

 

 

I always thought saga was an old English name for rice pudding...

 

Regards,

Tony

Posted by David Clark 1 on 08/06/2011 09:19:22:

I hope they aren't going to limit it to iPad - there are plenty of Android-based pads (and phones) out there too. And Android-based machines are starting to outsell IOS-based machines, so the market for putting it on Android will actually be bigger.

 

Regards,

Tony

Posted by David Clark 1 on 02/06/2011 21:25:04:

Hi David -

 

Is there another way?

 

Regards,

Tony

I'd send the drawings to John Stevenson attached to a suitable number of beer vouchers

Regards,

Tony

<...continued from previous post...>

 

Tony

Dennis -

 

Given that you are someone that has spend their career in automation, and has read my (and presumably others') articles on CNC in MEW, your post indicates a level of ignorance of what is available to today's CNC constructor/user that is frankly alarming.

 

I completely agree withyour aversion to programming one-offs using programming languages of any kind. I have also served my time as a programmer, and am familiar with enough low- and high-level programming languages to last me several lifetimes. Fortunately for me, and for most others that want to go down the CNC route, you don't need to learn G-code or any other programming language for that matter, to run a CNC machine; in fact, although I have owned and used CNC machines for a fair number of years now, I have yet to write a single line of G-code, and sincerely hope that that situation will continue.

 

The very first CNC system that I used was a Taig CNC mill; the CNC software provided with it was Dos-based, a program called SuperCam that provided both CAM functions and the "CNC control" functions needed to translate a toolpath into step-and-direction signals for the stepper drive hardware. While SuperCam could be programmed using a very, very limited subset of G-code, the program was primarily designed so that the toolpath would be specified in the form of a CAD drawing; you provided SuperCam with (for example) a DXF file that contained the output from your favourite CAD package. Absolutely no need for programming expertise in the sense that you are talking about, and it positively encouraged the user into the good habit of making an accurate drawing before cutting metal. The one serious drawback with SuperCam was that it assumed that the CAD drawing directly represented the path that the cutter would take, so you had to draw the tolpath 1/2 a cutter diameter smaller than the finished part to allow for the width of cut, and there was no easy way of defining an "area clear" operation other than by filling the area to be cleared with equally spaced lines 1/2 a cutter diameter or so apart.

 

The next system that I used, Desk CNC, again on a mill, is in some ways similar in approach to SuperCam in that it is an integrated CAM/control package, but rather more cleanly presented (Windows-based) and capable of allowing for cutter diameters, doing area clear, etc. etc. very simply. DeskCNC is much easier to use than SuperCam, because you draw the part actual size, and the same drawing can be used with different diameters of cutter if appropriate - for example using one type/size of cutter for roughing out and another for a final cleanup pass. Desk CNC also implements a much more complete subset of the G-code language if you really want to use it. I don't.

 

The third system that I use is Mach 3, this time for my ML-7. Mach 3 is as good an implementation of a G-code interpreter as you will find on the hobby market; however, again I haven't needed to use G-code with it so far, because Mach 3's "wizards" provide a very simple way of setting up the usual kinds of operation that you do on a lathe - plain turning to a shoulder, taper turning, threading, filleting, ...whatever, on a screen where you fill in dimensions in boxes, press a button, and Mach 3 generates the equivalent G-code program . So no need to program G-code for the vast majority of everyday uses.

 

There are also a good number of CAM packages available at prices that shouldn't upset the hobby user too much; in essence these take a CAD drawing, usually in DXF format, and generate G-code output that can then be fed into DeskCNC or Mach 3 (or whatever) to define the toolpath.

 

Yes, if you really want to, you can do the macho thing and program the G-code directly, but why would you do that, with the CAM tools that are readily available to do the job for you?

 

Given the above, using CNC for one-off part machining can be quick and easy, even for simple parts that would be straightforward to machine by hand on a mill or a lathe. However, there is one aspect of all this that hasn't been mentioned so far; even making a one-off part usually involves repeated operations, unless the part can be machined to size in a single pass. Sometimes that repetition can be tedious in the extreme - for example, taking repeated cuts to reduce a large diameter workpiece to size. Put your CNC system on the job and all of that handle cranking goes away. Obviously, if cranking is what floats your boat, then CNC isn't for you. Personally, I think it is much over-rated. Each to their own.

 

So yes, I completely agree with your comment that "It takes time to master any programming language and even longer to get good at it, and stay abreast of updates and debugs." Can't argue with that at all. But it is utterly irrelevant to this discussion unless you *need* to master the programming language. For the vast majority of uses of CNC in the amateur workshop, you don't.

 

<...continued in next post...>

...continued from previous post...

 

Tony

 

On the contrary, John, I take absolutely no delight in *anyone at all* deciding to unsubscribe from the magazine, so please do me the courtesy of not putting words into my mouth.

 

I came to engineering as a newbie, like yourself, with no former knowledge of metalworking to spak of; my present state of knowledge and skill, such as it is, is the result of lots of reading - books, magazines such as MEW, on-line forums - lots of trial-and-error, and lots of help freely given by people that know more about the subject than I do. Over the years I have written a good many articles for MEW; my primary and overriding motivation in writing for the magazine is to give something back, in appreciation of the help and support I have been given in the past and that I continue to receive. Inevitably, what I write about is what I do in my workshop; increasingly, that includes a fair amount of CNC-related "stuff", and will continue to do so. I have found CNC invaluable in my workshop activities (and almost exclusively for one-off pieces, not for repetition work, by the way; the idea that CNC is only useful for repetition work is often trotted out, but it is simply not true) and I know, frothe feedback that I have received, that what I write about CNChas been of help to others that want to go a similar path. I am not arrogant enough to believe that all model engineers should be using CNC; whether they do or not is a personal choice. However, for it to BE a personal choice, it has to be available and accessible to them; hence, while I believe that the one issue of MEW in question was wrong in its balance of CNC vs other stuff, I do not want to see the Editor browbeaten into removing CNC (or other relevant new technology for that matter) from the pages of the magazine just because of the personal prejudice of a small number of voiciferous but blinkered individuals.

 

Contrary to some comments that have been made in this thread, CNC is increasingly accessible to even the most modest buget; as John S has pointed out, building a small CNC router can be achived for as little as £100 if you are prepared to apply some skill and imagination, and once built, you have a tool that you can use to machine components for something better. A while back, I converted my Myford ML-7 to CNC (and wrote up the conversion for MEW); my lathe had formerly been fitted with a Myford quick change gearbox, and following the conversion, I sold off the gearbox and various other bits (the banjos and gears and the topslide) that were now surplus to requirements, and ended up with a CNC lathe plus a decent amount of surplus ££ in my pocket for other things. The point? Many of our readership would consider it perfectly reasonable to lay out the £400 or so that a QC gearbox would cost on the used market, or the £ lots more that Myford would charge for a new one, but somehow, laying out the ~£300 it cost me to do a CNC conversion is out of reach of their pockets. That makes no sense to me at all.

 ...continued in next post...

Posted by EtheAv8r on 01/06/2011 09:58:08:

Posted by Brian on 30/05/2011 16:46:58:

Thank You David, can we all asume that CNC is now excluded from MEW in the future?

regards

Brian

Edited By EtheAv8r on 01/06/2011 10:03:39

Very well put - thanks.

 

Regards,

Tony

Posted by Steve Talbot 1 on 31/05/2011 10:35:17:

Well Steve, if we take your argument to its logical conclusion, there are some readers that only have a lathe and there are some that only have a milling machine, so we should split the mag into "Model Engineers' Lathe" and "Model Engineers' Mill" or similar. That would do a major dis-service to the vast majority of the readership that actually want to have a magazine that covers all aspects of the (model) engineers' workshop.

 

The number of articles may well have reduced over the years, but my impression is (without doing the analysis) that the articles in recent years have been longer. I'm sure someone with time on their hands will do the necessary research and confirm or refute this.

 

I would wholeheartedly agree that articles in MEW, on whatever subject, should be comprehensible and have some engineering content. I know that whenever I have written for MEW I have attempted to make the article as understandable to the uninitiated as I can.

Posted by Tomfilery on 30/05/2011 17:30:24:

Hi Tom -

 

At 55, you can definitely count yourself as one of the young farts

 

Good to see a vote for a broad cross-section of technology, old and new; without a reasonable amount of attention to the new technology, I fear we will all become old farts before our time.

 

For my part, I will continue to contribute articles to MEW based on the stuff that I do in the workshop (when I have the tme and patience to do so). That will continue to be a mixture of old and new; my most recent article (Worden grinder construction) was definitely old technology, but even that caused more than a modicum of controversy amongst those that seem to have had a total sense of humour bypass. Ho hum.

 

The article I am currently working on will probably cause the luddites a good deal of indigestion, as it relates to simpler alternatives to running full CNC on a lathe. Does that bother me? Frankly, no. To me, there are two kinds of engineers (model or otherwise) in this world; the ones that already know it all, and for whom there is nothing new that can be remotely interesting, and the ones that (regardless of their level of skill) recognize that there is always something new and interesting to find out about, that will enrich their knowledge and understanding of the craft, whether or not they choose to make use of it themselves. As far as I can tell, it is mostly the former kind of engineers that seem to be having a problem with the amount of, or even the presence of, CNC articles in the mag; frankly, I am surprised that they bothered to subscribe in the first place, and losing them from the subscription list would be a net gain both to them and to the remaining readership.

 

Regards,

Tony