Before I am hung, drawn and quartered, for operating a lathe without guards, here is the proof that I have been sensible.
Guard over the X axis pulleys. I like to watch the wheels going round and round, hence the transparent top. Also note the cover over the exposed ball screw.
Cover over the Z axis pulleys and belt, again transparent. If I wore a watch it would be transparent.
I also installed an ER40 collet chuck. I will be using this for all work with diameters under 26mm.
Before I get onto a brief reflection about scale, the photo below shows 2 cannon barrels.
The big one was what impelled me to converting a manual lathe into a CNC lathe. That time consuming, costly, and ultimately very satisfying project, started because the CNC lathe which I used to turn the big barrel could only handle the job by doing it in two stages…. doing the breech first then the muzzle. That was due to the big barrel being too long for the lathe, at 300mm (12″).
The small barrel was a test for the CNC converted lathe just finished, being the first complicated shape which I have made. To save on material, I made it at exactly half the scale of the big one, ie 150mm long (6″).
Comparing the two barrels reminded me, that if an object is twice as big as another, in all 3 dimensions (height, width, depth), it is 8 times as heavy. And any projectile, and weight of black powder, would also be 8 times the weight. But the wall thickness of the explosion chamber is only TWICE as thick.
My point is, that if scale is maintained, the smaller the cannon, steam engine, boiler, whatever….. the less likely it is to explode.
Not that these cannons will ever be fired. Just hypothetically.
The wiring of the lathe is complete. (Except for limit switches. They can be added at any time).
Mach 3 is configured. The wireless hand control is installed and working. Ezilathe installed and waiting for input.
Some covers to be made.
Hook ups in progress. That’s the faulty VSD on top of the electronics enclosure. The CNC engineer lost his hair trying to figure out the problem.
Still some testing and fine tuning required.
But nothing much will happen in the workshop for the next 3 weeks.
Another couple of advances in the conversion. Today I installed the lead screw cover and the cable protector to the cross slide stepper motor.
The cable protector was easy and straightforward. It flexes in one direction only, and is fixed at the ends after the cable is threaded through it. The length is adjusted by adding or removing links. It was placed so that coolant liquid will drain out of it, and to minimise the accumulation of swarf. The cables themselves have a thick covering and are well protected. The link protector will not kink, further protecting the cable.
It was cheap. About $AUD20 for 2 meters, posted from China. I used about 1.1m.
Showing the stepper motor cable protector, and the lead screw protector (one half of it. The other half is on the other side of the carriage.)
The lead screw protector was another story. It is a spring steel coil, about 50mm wide, and as it is compressed the coils fit inside each other. I made a big mistake in allowing it to spring open before I had installed it (there were no instructions). It immediately opened to a length of over a meter, in coils about 50-60mm diameter. No big deal, I thought. I will just compress it back to its original configuration. Big mistake.
It was what I imagine coiling a live, oily, biting, boa constrictor would be like. (OK, boas constrict rather than bite. How about an anaconda, or a big eel.)
I fought it for about an hour. And eventually succeeded. Minus a few bits of my skin.
So I did not allow the protectors to expand again until after I had them on the lead screw.
This is what they look like. Pretty cool IMO. They just expanded into position when I removed the restraining clips.
The lead screw stepper motor and protector. The Estop box above will get some ends to exclude swarf.
It was not cheap. The best price that I could find was from South Korea. $AUD200 inc postage. But it is excellent Japanese quality.
The wiring is happening, but the variable speed drive seems to be dead. It has been sitting unused on a shelf for 2 years, so no point asking about warranty. Took it apart to check for broken wires, fuses, burnt out components etc, but nothing visible. Will order another one. About $AUD200. An unexpected expense.
These lathe CNC conversion posts are probably becoming a bit tiresome, but just in case there is someone out there who is interested, I will continue until the job is finished.
The latest was to make and install a spindle speed (and position – thanks David M) sensor. It consists of a disk with a slot cut in the periphery, attached to the main spindle. And an opto-electronic sensor which is connected to its own electronic board, thence to the breakout board and VSD.
The disc with the slot at 8:30 and the sensor at 9:00. I must have chosen the wrong cutter or turning speed for that disc aluminium… looks a bit rough. (note added 13/7 Stuart T says that I should have used coolant-lubricant).
View from above. Any clearer? That gear is now superfluous except as a spacer.
So there is one electronic impulse per spindle revolution. That is enough to measure the RPM’s. Essential for cutting threads.
The beauty of this system is that there is no gear selection or changing, and ANY thread pitch can be selected… metric, imperial, BA etc… any odd ball thread that your heart desires.
The HTD (high torque drive, I am informed by many readers) pulleys and belts and taper lock fittings. Unfortunately I could not find a taper lock to fit the small pulleys, so when it is all finally, definitely, absolutely, correctly, positioned, I will Loctite them in position. Protective covers yet to be made. I quite like to see the mechanicals in action, so I am intending to make the covers from clear polycarbonate.(Lexan) .
Adjusting the lead screw.
The 48 tooth HTD pulley has been installed using a taper lock.
Then some time was spent adjusting the parallelism of the lead screw. That requires quite a few movements of the carriage along the 600mm thread. Each 360 degree turn of the lead screw advances the carriage 6mm, so you can understand that I became a bit impatient with all of the repetitive hand actions to move the carriage from one end to the other.
So this was a solution to that issue. That HTD belt is the one that was too long, so I was happy to find a use for it. The variable speed battery drill shot the carriage end to end in a couple of seconds.
All is now adjusted parallel.
A few more little installation issues, then for the wiring.
Today I fitted the lead screw.
No big deal, I sense that you are thinking. After all, the ends are machined, the bearings fitted, and all waits in readiness.
True, but there is a strict sequence of events. And since it has been 3 or more weeks since it has been together, I had to rediscover the sequence, by trial and error. And each bit of the fitting is very heavy, very delicate, very tricky. So it took me several hours to get to the final photo in this blog.
But first a view of the inside of the newly machined apron.
The lead screw fitted. The cross slide screw is also fitted. Note the red E Stop panic button fitted to the left. Next job is to fit a support bearing at the right hand end of the screw. Then to check and adjust parallelism of the screws. A rough check showed that they are within 0.25mm
Hooray!
Today I collected the lead screw after the ends were machined by Statewide Linear Bearings.
I decided to drive the 100km each way to pick it up, in preference to using a courier. I wanted to ensure that all of the small bits were there, and also just to make sure it was handled properly. Mostly freeway, listening to Dan Carlin on the Persian-Greek wars, so it was a pleasant way to have 3-4 hours to myself. (If you do not know about Dan Carlin, Google him and download an episode. If history at school had been like this, we would all be history addicts.)
This is the lead screw, ends machined, and support bearings fitted. 1100mm long. 28mm dia
All good, except that the nut was back to front. That nut is pre-tensioned, which means that the 2 halves are separated by a precisely machined washer. I was nervous about removing it and replacing it the correct way around. However I had previously asked the ball screw expert about that aspect, so armed with the technique I made up a sleeve of the correct size, removed the nut and replaced it. No balls fell out. So all good! The above picture shows the nut in its correct position.
The nut. Looks expensive? Is expensive. And beautiful.
The machined driven end. $AUD250 machining there. But it is perfectly done.
And with the support bearing installed. A pulley for the HTD belt goes on the distal bit of shaft.
The CNC lathe has 3 belts. There is a V belt from the 3 phase motor to the main lathe spindle. Although I changed the motor and the pulleys, the old belt fitted, which was good. No hassle. But the stepper motors driving the lead screw and cross slide screw, and their pulleys were all […]
The CNC lathe conversion has been happening, despite no posts on the blog.
I have mounted the electronics enclosure, and mounted the various components inside. No wiring yet.
This stainless steel tool box is the electronics enclosure. It fits the space quite nicely, and is adequately big. The back gear cover to the right will be retained, although the back gears have been discarded. The main switch and emergency stop will be mounted somewhere on this cover.
The Variable speed drive (VSD) sits on top. That will control the spindle speed. The transformers, stepper motor drives, and Breakout board (the heart of the system) are positioned inside. Plus cooling fan and filters. Ready for wiring.
Drilling the apron to attach the cross slide ball screw bearing. One chance only at this one, so the setting up took a couple of hours. The apron is clamped to a large angle bracket on the milling table. M6 threading followed.
The end result. The bearing as attached to the apron and the ball screw is in place. I machined the end of this ball screw to fit the bearing, cut a thread (M10x1), and machined the end to accept the pulley. All good. There is 0.25mm adjustment available if required, but it all seems pretty correct. The bearing sits on a carefully machined block which is 7.85mm thick. Still waiting the lead screw machining.(!!)
I sometimes feel a bit ashamed when I have visitors at my workshop.
The reason is that when I am in the middle of a project, I really concentrate my energy on the decisions, the machining, working out how to fix the mistakes…
…. and tidying up as I go, is near the end of the list of must do’s.
Consequently, tools tend to be put aside at the spot where I have been using them. And off cuts of steel or brass or wood or whatever, lay where they fall.
And as mentioned in a previous post, I have a policy of leaving swarf on the floor, to discourage wildlife from slithering into my workspace. (see the old post about the tiger snake between the lathe and the milling machine). And if you are not Australian, look up tiger snakes. They are just about the most dangerous reptile on the planet.
So my workshop is not the tidy, organised sort of workspace which you might expect from a retired gynaecological surgeon.
But occasionally, the mess becomes so extreme, that I cannot find tools, I trip over stuff on the floor, everything is really dirty, and it is dangerous and embarrassing when visitors call in. And some of those visitors have workshops where you could eat off the floor.
So yesterday I spent a whole day tidying, sorting, putting away tools, throwing out rubbish, and sweeping the floors.
What about the tiger snakes I sense you asking.
Well, here in the antipodes, we are in the depths of winter, and it is bloody cold. And all sensible cold blooded reptiles are asleep in their homes. So for a few months it should be safe to sweep up the swarf. Here’s hoping anyway.
Continuing the installation of the ball screws, and stepper motors.
I have completely removed the digital read out module and glass slides, and they will not be reinstalled. Not sure what I will do with them. They are only a year or two old, and in good condition. I will probably put them on Ebay. Same with the old gearbox, carriage apron, and electric controls.
Here are some pics of the plates and blocks which support the ball screws and steppers.
This is the steel plate at the headstock end, bolted to the bed. And the block with the holes is cast iron 42mm thick, to support the leadscrew and leadscrew stepper motor. It was machined out of an old piece of machinery, hence some unintended holes. Being cast iron it was fairly easy to machine, but incredibly dirty. Turned everything in the workshop black, including me. (whoops. Unintended not PC)
This is the block which replaces the gears and controls of the apron under the carriage. The thick block is cast iron, and the stepper motor support is 20mm thick steel. Very heavy.
This plate is hidden under the carriage. It secures the lead screw nut.
The slot in the carriage had to be widened and deepened a bit, in order to accomodate the slightly fatter and taller cross slide nut. See the next photo to see the setup for milling the hole through the carriage.
A rather confusing photo. The carriage is clamped to a large angle plate on the mill, and I am enlarging the hole which accommodates the cross slide ball screw. It was at the limit of what my mill could manage. An intermittent cut, with a lot of tool stick out. Not the best way of doing the job, but it worked OK.
Plastic covers attached to the stepper motors, and toothed belt pulleys fitted.
Checking the centres between the pulleys, using 2 wooden wedges to push the pulleys apart.
The underside of the carriage. The hole and channel at the left side of the picture was machined to accept the larger cross slide screw
So you can see that I have been busy since the last post.
At present the lead screw is at Linear Bearings in Melbourne, having the ends machined to accept the driving pulley, and support bearings. I did consider doing this machining myself, but decided to leave it to the professionals because of the high cost of the item and the hardness of the material.
I am still waiting for the replacement ball nut for the lathe cross slide to arrive.
Meanwhile, I have been busy machining the supports for the lead screw.
Drilling the holes for the support bolts for the lead screw nut
And gradually drilling the hole to 49mm!
That is a 49mm drill! First time that I have used it! Thank goodness for the FS Wizard app, to give me some idea about feeds and speeds. Following this I used a boring head to enlarge the hole to 55mm.
Gradually enlarged the hole in 20mm steel to 55mm diameter. and here is the lead ball screw, sitting roughly in its proper position.
So this is where I am at. The lead ball screw is sitting approximately in its correct position. Considerable adjustment required. And I am yet to turn the ball screw ends to their correct dimensions.
My first hitch occurred today.
I was very excited to receive the nut for the cross slide ball screw. If you have been following these posts you might recall that the ground ball screw for the cross slide came from Taiwan, and arrived in 3 days. But I had to order the nut from a seller in USA. The nut was advertised as new old stock, but with no packaging. That was OK, but the postage cost for such a tiny item was ferocious.
It was the last item to arrive from overseas. However when I looked at it, it was obviously NOT new.
The cap screws holding the ball recirculating tube were different from each other, and the washers underneath were too big for the screws. Obviously not the way that TNK made it. Somebody has had the nut apart. And the ball retaining tube was very scratched and loose. Again, not TNK standard.
But no biggie. If it works — fine.
So I turned up a retaining tube to remove the previous nut from the ballscrew, and it came off without any drama.
But when I tried to fit the “new” nut, it just would not go on. Tried reversing the direction. No go. Bugger bugger.
Somebody has altered or changed the “new” nut. Maybe installed balls which are too big, or maybe damaged the entry thread. I do not know.
What to do. I have been waiting 2 weeks for this to arrive.
First, Ebay email to the seller. See what the response is. Ask for a refund. The postage was almost as expensive as the nut. If unsatisfactory response, they will get the worst Ebay feedback ever.
There is one other seller of these nuts on Ebay, also in USA, and 50% more expensive, and the postage is also 50% more expensive. (how DOES ebay come up with the postage charges. It seems more related to the cost of the item rather than the weight-size etc.) But the nuts are in original packaging. And I want to get going with this, so fuck it. Pay up and get it.
I will report in a later post. (if the bad language in this post seems to reflect my state of mind, well, yes it does.)
PS. Next day. After sending photos of the issue, the seller accepted responsibility, and I am getting a full refund. That restores my faith in Ebay/Paypal. I hope that the next one is in better condition, and comes a bit more quickly. Sorry for the bad language.
This is a list of the components which I have accumulated to convert a manual lathe to a CNC lathe. It is not quite complete, but close.
The electronic components, not including computer and parallel cable and manual pulse generator.
Two stepper motors. Nema 34, 1200 inch – oz. With rear covers.
A Gecko microstep drive for each stepper motor
Cable and connectors for the stepper motors
A transformer-power supply (48 volt)
Another transformer-power supply (5 volt).
3 phase 1.5kw motor (top) to replace the single phase motor (bottom)
Timing belt gears 24 and 48 tooth, 5M. Order belts when size is definitely established.
FK20 lead screw bearing and Ball screw covers
The electronic heart of the system- the breakout board. A C11R9
The index pulse board and sensor. A C3.
Manual pulse generator, wireless.
Variable speed drive, identical to this one on the mill
An electrical enclosure, to hold the various components. This stainless steel box was originally an item of medical kit. Here I am checking out one possible location. Not yet definitely decided to use this.
Finally in the electronics section, I will need a computer, loaded with Mach3 and Ezilathe. Surprisingly, it does not need to be a particularly powerful PC. And there are advantages in using an older operating system such as XPpro. I think there are a couple of those in the attic. If not, I should be able to pick one up for under $100.
3. Various structural items. Most of these will be 20mm thick x 200mm wide steel, cut to size and shaped and drilled. I will take pics of these as I make them. I was planning to have them water jet cut, but the shapes a fairly simple so I will cut and machine them myself as I go.
So, that is most of the items for the job. I have spent about $AUD1100 on the ball screws and nuts, about $AUD1500 on the electronics and electrics, and maybe another $AUD500 on pulleys, belts, steel, taper locks, fasteners etc. I will add it all up accurately at the completion. The lathe was cheap, bought sight unseen a few years ago. So all up, I should have a CNC lathe for under $AUD4000. And many pleasant hours of design and machining. And a great learning exercise. Stay tuned!
I am in the process of collecting all of the components for the conversion. Parcels are arriving from South Korea, Taiwan, China, USA and Australia. Next post I will take a photo of the bits, before I commence assembly. I have spent a lot of hours on the computer drawing up the positioning of the new components, deciding which components to get, then communicating with the sellers and making the purchases. Not to mention hanging around home when parcels are due. If I duck out for 10 minutes, that’s when the delivery van arrives. And of course he leaves his little card “sorry we missed you”.
This post I will show you some of the drawings of the proposed conversion. Of course the first step is to strip the lathe of all of the old manual controls, gears, motor, Digital read out, carriage apron, lead and cross slide screw, electric control box and contents. (taking photos of every component insitu in case of a change of mind, and restoration required later).
Then measuring the bed and carriage in minute and accurate detail, and drawing it in CAD.
This is the lathe side on and end views, showing the new lead ball screw in red and green. The cross slide ball screw is also there, but not well seen at this scale. The 4 ball screws at top left are the possible combinations for eventual installation. The second red one is the position eventually decided. The green ones are with the nut re-installed in reverse direction. I really did not want to remove and re-install the nut, because it is pre-loaded, and I do not want to risk disturbing the setup.
Many drawing versions are required, and as the components arrive, I find myself making changes. This is definitely not the final version. The two carriages show the carriage in its extreme positions on the lathe bed.
This was an early sketch of how I thought I would arrange the cross slide motor and lead screw nut.
This is a fairly accurate drawing of a cross section through the cross slide. Black is existing. Red is the new nut and ball screw. You can see that removal of some of the cross slide bed iron will be required. After looking at this I decided to move the nut and screw up a couple of millimeters.
The Internet has been very useful in showing what fittings are available. These are a few of the catalogues and tables which I have downloaded. Some sites even provide CAD drawings so their components can be inserted into my CAD drawing to see if they will work together.
I have learnt a lot about ball screws in the past few days. And I have purchased 2 ball screws and nuts on Ebay. For those relatives and friends who follow this post, who have no idea what I am talking about, the “ballscrews” are what determines where the cutting tool on the lathe is positioned. Very crucial. (can something be “very” crucial? It looks a bit like “very unique”. )
Ball screws are the usual positioning screws for CNC machines these days. The alternative is Acme or square thread screws, but the few manufacturers who used to use these have all switched to ball screws (as far as I know). Even Wabeco, the respected German lathe manufacturer no longer specifies any threads except ball screws.
Ball screws require less power to turn due to friction being a fraction of the alternatives. Ball screws are silent. If they are noisy there is something wrong. They are generally more accurate than square or Acme threads. They used to be many times more expensive than the older types, but since China/Taiwan has taken over most of the manufacturing, often using equipment sourced from US/Europe/Japan, the costs have plummeted.
And the backlash is minimal or zero. Another name for ballscrews is “antibacklash screws”.
Backlash is annoying on a manual lathe, but it is very important on the cross slide of a CNC lathe. Interestingly, it is less important on the CNC lathe lead screw, because most CNC lathe machining on the leadscrew occurs in one direction only, towards the headstock. However, the cross slide is machining in both directions, in and out, so the absence of backlash is necessary to maintain workpiece accuracy.
Ballscrews have grades of accuracy varying from C0 -C20. The bigger the number, the less accurate the screw. In general, it is recommended that industrial lathes should be C3-C5, which means zero backlash, and accuracy of about 0.001mm. That degree of accuracy is quite expensive, costing thousands of dollars per screw, and only attainable in ball screws which have been finished with precision grinding. The alternative is ball screws which have been “rolled”. These are much less expensive, costing hundreds of dollars per screw, depending on the degree of accuracy. The best rolled screws can attain an accuracy of 0.01-0.02mm (C7), down to 0.1mm (C20). These are approximate figures which I have gleaned from several manufacturers’ specifications.
So from scale drawings of the lathe bed and carriage and cross slide, I worked out that I needed the following…
This is a photo of a ball nut and screw very similar to what I have bought for the cross slide. THK brand, 14mm diameter screw, 4mm pitch, and BNT nut style. Note the rounded channels that the balls occupy.
(ps. see the comments section for discussion about axial accuracy and backlash. A super accurate C5 ballscrew has axial play (backlash) specification of 0.020mm, but the double ballnut configuration will reduce that number substantially. “zero backlash” apparently does not really mean zero.)
After removing most of the lathe gear which will not be required after the CNC conversion, the lathe is looking a bit naked.
The carriage apron, the lead screw, the back gears, the drive rod and control rods have all been removed. Also the cross slide screw and handle. The cross slide itself is temporarily removed, but available for measuring for fitting a ball screw.
I have now made accurate measurements and drawings of the lathe bed and carriage, in order to choose ball screws and nuts for the lead screw and cross slide.
The lead ball screw is easy. There is plenty of room and machined surfaces for attachment. I see no particular problems there. Just time, careful machining and expense. Chinese or Euro-American? As usual, there is a big price difference and maybe not such a big quality difference as previously. Looking at 25 or 32mm diameter, with 550-600mm of thread.
The cross slide ball screw is another matter. The current cross slide square thread screw is 14mm diameter, and I would like to use a ball screw about the same size. The problem is that a ball screw nut is considerably bulkier than the existing square thread nut, so some machining of the cross slide will be required to make space. The cross slide dimensions are already fairly tight, to maximise the swing over the carriage. I do not want to weaken the cross slide too much. So it is all a bit tricky. Time to consider options. And to get another opinion.
No lathe pics, so here are some of Turkish wall smashers.
This one was given to Queen Victoria by the Turkish sultan. It was made about a decade after the fall of Constantinople. It was cast in 2 halves. There is a giant thread connecting the halves. I imagine that the strange square holes are to allow levers to be inserted for the screwing by many strong men. No double entendre intended.
This one could have been used to make the breach in the wall at the fall of Constantinople 1453. That stone ball is 600mm diameter. With no trunnions or other supporting mechanism the barrel was probably dug into the ground for support. That would allow repeated shots at exactly the same point in the walls. 8-11 shots per day. It was made for the invading Turks by Orban, a christian who had previously offered his services to the defending Byzantines. The Byzantines whose empire by this time had been reduced to a tiny fraction plus the city itself, could not afford his services. The rest is history.
The carronade is finished, as far as I intend to take it. At some future date I might make pulleys and ropes etc, but at this stage I am putting it on the mantlepiece. (mantelpiece groans). Some detailed pics in a future post.
I have commenced my next project.
I have a CNC lathe but it will accept work up to only 125mm diameter and 125mm long. It was not big enough for the long gun, and barely fitted the carronade. And I have some ideas of further larger projects (field artillery pieces, and possibly a model of a Turkish wall smasher like the ones which allow the Turks to conquer Constantinople. That one was almost 6 meters long, and fired stone balls of 600mm diameter!!! So even at 1:10 I need a bigger lathe.
OK, so I could use a manual lathe, but that is not the point. A bigger CNC lathe would be fun. And I have a Taiwanese one which I think would be suitable for conversion. It is 600mm between centres, (just big enough for my Turkish smasher), and about 300mm swing. It is not pleasant to use as a manual lathe due to very noisy spur gears. So I have decided to convert it to CNC.
The steps are:
I have made a start. Removed most of the unwanted manual components from the lathe today. It felt very threatening and unnatural to be wrecking a perfectly good lathe. See the photos. At this stage I am taking lots of photos in case I have a change of heart and restore it to its original state. But I will press on. Watch this blog. I expect that the conversion will take a couple of months, by the time components arrive from overseas.
The lathe prior to CNC conversion
After removing the lead screw, apron, gear box, cross slide back gears etc etc. Looks a bit naked. Not much remaining.
Some videos and pics of some stuff made on the Boxford.
CNC is great for multiples. These are oil cups with ME threads.
Steam engine link
Ball end handle for a small lathe
The finish on the distal end was suboptimal.
First step in making Watts parallel motion links for the beam engine.
Base for a Jan Ridders Stirling engine
The steam control cock and butterfly valve. The body of the top valve and both handles were CNC’d.
If you have found this little series of blogs about the lathe CNC conversion interesting, and would like to see a similar blog about how I got an ancient CNC mill working, let me know. Leave some feedback.
johnsmachines 