Modulathe

[John HaineParticipant @johnhaine32865]

https://github.com/kachurovskiy/modulathe/tree/master

 

[Michael GilliganParticipant @michaelgilligan61133]

Spectacularly good concept, John

… comment based entirely on a quick skim through the ‘Assembly Manual’

MichaelG.

[Nigel Graham 2Participant @nigelgraham2]

A 38mm bore, MT5 spindle is hardly in the “desktop” class, surely.

No guide to maximum feasible bed length the customer can specify.

Actually the whole thing is how an enthusiast somewhere built his own lathe, probably finding nothing suitable commerically made that he could afford, but in publishing the instructions seems naive of just why a decent lathe capable of reasonable-quality metalwork is expensive.

.

No mention of a tailstock, other chuck types or any other fittings. Nor of chuck mounting – I would expect some form of cam locking on a lathe with a spindle that big. Wouldn’t a 125mm dia chuck be a bit small for that spindle? The designer specifies the spindle and chuck, the former a Chinese-made item intended as a spare for several lathes with model numbers starting “WM” (Warco?). Its own advertised price, around £45.

Very small diameter leadscrew for a 5-1/2″ centre-height lathe, but as it is a ball-screw is this 12mm typical on NC machines of this capacity?

Tapped holes look a neat alternative to T-slots for the cross-slide, but you’d need cover or plug all the unused holes to stop them becoming choked with swarf. Really, it’s probably just a cheaper way to make it.

Why the apparent choice of spindle bearing types? That does not inspire confidence in its design. It claims a very large spindle for a supposedly compact lathe otherwise, so surely its entire headstock would be specified to suit? I think I answered my own question when I re-read the section headed “Core Ideas”.

Then read the linked Manual….

Much of the machine is plastic-filament printed. The bed is 160 X 80 X 4mm tubular steel, default length (not sure what that means) 680mm, filled with concrete – nothing wrong with that in principle, indeed it is a proven commercial method, but how do we ensure high-quality working surfaces with this lathe? What does the headstock sit on and the saddle slide along? We are told to use hammering or grinding to remove any bulges in the tube’s surface! So far, nothing about the bearing surface, nor how we guarantee accuracy.

Right, we’ve built something worthy of RR’s tool-room so far. We’ve bought the spindle (above):

[list of parts to be bought, ending with:]

M45x1.5mm precision nut, 15€

Cool the spindle, heat up the large bearing [what temperatures?] and slide it onto the spindle.

[Assembly instructions, including some printed tooling, to..]

Print and install spindle pulley with a 4x4x20mm steel key, tighten with light pressure using a M45 precision nut. Do not buy cheaper non-precision nuts.

The specified size is on the ISO-M Fine list but what makes it a “precision nut” apart from costing £12.22 by today’s price? Why does it need be “precision”?

Sanou 125mm 3-jaw chuck, 62€
Very important to buy a quality SANOU chuck with ground jaws. Don’t get VEVOR chucks.

It says. I reckon Vevor will be relieved, not saddened, to read that. The specified spindle end is a flange with three holes so it may of course simply mean Vevor’s mounting arrangment differs from Sanou’s.

 

Looking further down it is sort of clear the ways are actually linear rails, which is fine; but although the author does his best to explain the importance of aligning them and gives some instructions on this, I think he has rushed into “print” with a prototype more likely to disappoint than satisfy if you intend it for high-grade model-engineering. He has certainly thought it through, to an impressive result, but not quite through enough and his dismissal of commercially-made conventional and CNC machines is hollow as a result. He ends by suggesting improvements, but surely he would have been better making and proving them before publishing?

Effectively it’s a DIY project with a lot of interesting ideas based on modern CNC machine-tool designs (such as the filled-tube bed and the linear guides), but as it is, with a plastic headstock and no convincing radial and axial support for the spindle, I would not guarantee it being capable of high-grade work… or of much work at all despite such a massive spindle and generous centre-height.

Oh, and you will have to design and make your own tailstock, and buy a lot of M5 taper tooling!

 

“Lathes are dangerous” it says at the top. So are any machine-tools, portable power tools, step-ladders, slippery baths, the kitchen stove….

[Michael GilliganParticipant @michaelgilligan61133]

These two replies [both of which are, in my opinion, valid on their own terms] might give a good indication of the rocky road ahead for ME&W magazine !

MichaelG.

[JasonBModerator @jasonb]

Indeed on the one hand people can see past the odd bit that may not be ideal others just seem against it and would probably fall at the first hurdle of not being able to open the files to print from.

The 38mm bore spindle is just a spare part from a 210mm size lathe and you can also get minilathes with a similar bore so certainly benchtop league if not desktop though depends on your desk a bit. Spindle bearings the same as these lathes use too.

Why would you need a lot of MT5 tooling, MT5-3 or 2 sleeve and you can use your old myford tooling.

Tailstock only needed if you want to turn between ctrs, just like the other small “Chucker” lathes and full size CNC any drill is held on the the cross slide, you then have things like automatic peck drilling at your displsal.

[John HaineParticipant @johnhaine32865]

You might say that this type of article is sounding the death knell of the magazine.  Instant publication, free to access.  Github can support Wikis for discussion (though no discussion of this project yet as far as I can see).  Though originally for software Github seems to have an increasing number of hardware projects.

[Ex contributorParticipant @mgnbuk]

The specified size is on the ISO-M Fine list but what makes it a “precision nut” apart from costing £12.22 by today’s price? Why does it need be “precision”?

Precision locknuts are a “thing” John.

They are intended for uses such as pre-loading or securing bearings on precision assemblies such as spindles or ballscrews. The threads and abutment face are ground at the same setup to ensure that the face is truly perpendicular to the thread. The threads are ground to be a close running fit on the male thread of the shaft. A method of locking the nut against rotation but without moving the nut axially when it is clamped is included – typically a brass insert pressed into a bore machined perpendicular to the thread flank in front of a grub screw. The brass plug is inserted before the thread is ground so that it has the thread across it’s end. There may be more than one locking screw/plug equally spaced around the nut.

In use, the nut is screwed on to the shaft and the locking plug screw(s) adjusted to take out any axial play between the nut and shaft, but still allowing it to rotate. The nut is adjusted to set the correct bearing pre-load as required, then locked against further rotation with a tweak on the locking screw(s) that does not cause any further axial movement or cause uneven bearing preload (face camming). Easy, damage free, removal by slackening off the locking screws a turn or so & giving each a tap with a hammer & brass drift to set back the brass plugs & the nut just screws off freely.

These things are the “gold standard” for locknuts & the Chinese items are really not that expensive for what they are. They are night-and-day better to use than the usual cheaper solution of two thin locknuts with a tab washer to lock them. I used the SKF versions (types KMT and KMTA) at work as my preferred option on spindles and ballscrews & have a Chinese version on hand for my Chester mini lathe as a replacement for the standard two thin nuts + tab washer set-up when I get around to playing with it (taper roller bearings also on hand).

I have not yet looked at the rest of the article, but from a spindle bearing preload locknut POV the article writer has made the correct call here IMO.

Nigel B.

[Ex contributor]

On Ex contributor Said:

The specified size is on the ISO-M Fine list but what makes it a “precision nut” apart from costing £12.22 by today’s price? Why does it need be “precision”?

Precision locknuts are a “thing” John.

[…]

Excellent response !

… just unfortunate that the forum software trapped you  🙂

… I presume you were actually quoting Nigel, not John

MichaelG.

[BazyleParticipant @bazyle]

To compliment this make your own CNC mill https://www.youtube.com/watch?v=L8t82OQXefM

This uses a 3D printed framework filled with concrete.

BTW the large bore in small lathe is what the American gun nuts want so increasing the potential audience. A lathe is no good if you can’t shove a barrel down it.

[John HaineParticipant @johnhaine32865]

One aspect that’s a bit sub-optimal is that the headstock is only filled with dry sand AFAICS and he acknowledges that it isn’t that rigid.  On the other hand the bed is filled with ordinary concrete and I believe that this takes a long time to completely set and can distort in the process.  I think it was Mike of Mike’s Workshop (but may remember incorrectly) who described using epoxy concrete to stiffen up the column and base of an X1 mill – maybe a better choice for both the bed and the headstock.

[Bazyle]

On Bazyle Said:

BTW the large bore in small lathe is what the American gun nuts want so increasing the potential audience. A lathe is no good if you can’t shove a barrel down it.

If I ever buy another lathe, a large spindle bore will be high up the list of necessities the same way it was when I went from a 7×10 to the 10×22. It’s so much easier, quicker, efficient and economical to poke a length of material through the spindle and machine the desired part on the end than to cut it into short pieces first. Part of my original workshop clear up a few years ago was to throw the rollcab drawerful of short bits away. There are better uses for a drawer that cost £100…

I have several spacers to make from a length of 30mm tube, and that’s annoyingly that’s slightly too large to fit my WM250

[Ex contributorParticipant @mgnbuk]

… I presume you were actually quoting Nigel, not John

?

Not quite sure what you are getting at here Michael ?

Your quote from my reply is not what I am seeing here – I quoted from John’s comment & italicised it to show that was what my response was to. My reply as written (and displayed on my pc) has a clear seperating line between it and the quote and is not in italics, not as shown in your quote ? Maybe I should have put an apostrophy after “thing” and before John ?

Whatever – I am Nigel and I was quoting John, if that makes things clearer ?

Nigel B.

[John HaineParticipant @johnhaine32865]

Just to be clear that I made no comment about precision locknuts.

[Ex contributorParticipant @mgnbuk]

Just to be clear that I made no comment about precision locknuts.

Indeed you didn’t John – my mistake, sorry for any confusion.

Nigel B.

 

[Ex contributor]

On Ex contributor Said:

… I presume you were actually quoting Nigel, not John

?

Not quite sure what you are getting at here Michael ?

Your quote from my reply is not what I am seeing here – I quoted from John’s comment & italicised it to show that was what my response was to. […]

No you didn’t Nigel

That was my point

You were actually quoting from post #778777 by NigelGraham2

MichaelG.

.

Edit: __ Hang the expense … let’s have a clarifying screenshot

.

[BazyleParticipant @bazyle]

The dry sand is meant to absorb energy of vibrations. It used to be suggested for the walls of loudspeaker cabinets back in the ’80s. I have wondered if it was worth putting bags of sand in sheet metal lathe cabinets, or on drip trays to provide dampening rather than just mass.

[John HaineParticipant @johnhaine32865]

It may absorb vibration but at the expense of rigidity I’d have thought.  After all you don’t fill CI headstock castings with sand though some cheaper machines seem to have quite a lot of casting sand left!).  He does talk about a 0.5 mm sideways deflection of a long bar in the chuck which sounds excessive to me.

[JasonBModerator @jasonb]

Though looking at it the bar is long and slender and some of that could well be down to the chuck.

[Ex contributorParticipant @mgnbuk]

You were actually quoting from post #778777 by NigelGraham2

Had you said that initially then I would not have been confused by your initial “clarification”, Michael – having two “Nigels” in the mix didn’t help here.

Thank you for pointing out my initial error.

Nigel B.

[John HaineParticipant @johnhaine32865]

Not a great finish though that could be down to the material.

When he rotated the spindle by hand I hope the squeaky noise wasn’t the bearings?

That ELS controller – NanoELS v5 – is also linked from the website.  Another open-source design which maybe people should be aware of?  I note that it operates sensibly, withdrawing the tool and feeding it back to the start and then feeding back in for the next cut.  I guess there may be a 1ppr spindle index pulse.

[Nigel Graham 2Participant @nigelgraham2]

Weren’t the Peatol lathe beds filled with concrete? Not ordinary garden-path stuff which would probably attack the aluminum casing, but a special type, perhaps epoxy rather than cement.

At the other end of the scale I do recall reading quite some years ago of a major German manufacturer of very large machining centres making the beds as plate-steel fabrications filled with concrete.

It is also possible to cast basalt – the dark-grey sparkly stuff of your “granite” surface-plate – and this is used for making liners for plant carrying very abrasive slurries. I do not know if this has ever been done for making machine-tool beds etc, with the bearing surfaces fitted to it rather than machined directly into the stone. The melting-point is not far different from that of iron.

[JasonBModerator @jasonb]

The Syil mills use SCHNEEBERGER mineral casting for the bed and columns and I’ve also seen home made machines use epoxy concrete. I suppose it would be easy enought to 3D print the moulds for epoxy concrete and include locations for threaded inserts within the 3D prints so they get cast in place. I’ve seen a Gantry router used to cut boards to make moulds like that for a larger CNC mill. Also seen ganrtry mills built up off a slab of granite with the lengthways gantry supports also from Granite bonded on.

Back to the subject lathe. I wonder if a bit more rigidity could be had if the rails and head were fixed straight to the box section, can’t really see why it needs that thin printed “bed” between the two.

[Author]

Posts