Gib Strip Fettling

[Dr_GMJNParticipant @dr_gmjn]

Would it be possible to use my Fobco Star pillar drill to mill the divots? It has a depth stop. It would save temporarily re-assembling the mill, and is probably just as robust as the mill.

As mentioned, I’m planning on 3D printing some angled jigs which I could clamp into position on the table.

[Clive FosterParticipant @clivefoster55965]

The GHT method that I described can be done without a milling machine and gives best results. Although if you are making a new gib from scratch creating the angled surfaces at top and bottom so the new gib suits snugly into the top of its recess whilst having clearance at the bottom so it doesn’t drag can be challenging.

An oft overlooked point about creating conventional parallel gibs with push screw adjustment is that there are three basic ways of doing the job. It’s important not to conflate them and the repair /production processes involved if the job is to be done properly.

Whether you choose to use a floated gib riding on the screw ends or a GHT style one snugged up into the recess its best to delay drilling and fitting the dowel(s) until after you have got the new or refurbished gib fitted and settled home nicely. The drilling is best done on the machine taking pains to go straight through in line with the adjuster screws. If you go a touch far and end up with a small, shallow hole in the dovetail this will be of no consequence. Annoyingly you will have to pull the gib to make the dowel hole a touch larger so the gib doesn’t bind and have to repeat the adjustment process. But properly doweled and fitted gibs seem to go along time between adjustment. 10 years to my certain knowledge on a SouthBend re-furb.

The GHT method follows the design used on my Smart & Brown 1024 and some other machines of the very highest class. The defining characteristic is that the gib is forced up into its recess and held there by the action of the ball ended, or floating intermediate balls, thereby coming close to the ideal of a complete solid metal unit but adjustable. I suspect the gibs on my 1024 were ground to fit the gap as there is no visible clearance between the gibs and the two parts of the dovetail slide units.

In principle this style of gib is the most rigid as it requires only one oil film thickness of clearance but its very expensive to make as each gib must be hand fitted to its machine. Smart & Brown used two dowels and a greater density of adjusting screws than common practice too. When new the ex-factory price of my 1024 was comparable to that of a modest but completely satisfactory house. Rather a lot for a machine of only 10″ swing so the buyer would, rightly, have expected great pains to be taken in both design and execution to deliver a performance commensurate with the expenditure.

Of necessity more affordable machines generally used a simpler principle with the gib basically floating on the ends of the adjusting screws. Various versions of the design exist. The most basic being the simple pointed screw in a dimple style used by SouthBend. Myfords version using a ball ended screw being a little more sophisticated. Generally inexpensive to make and, if done accurately, of entirely adequate performance over a decent service life.

The number of smaller machine tools made in this manner is probably beyond convenient, or even inconvenient, counting generally delivering performance entirely satisfactory to their users for many years until wear intervenes.

The obvious fundamental weakness is that the gib is unrestrained against longitudinal float so the ends of the adjusters and the locating dimples can wear under the action of direction changes. It is said that direction changes can also induce a wedging action causing the adjustment to self tighten. Theoretically more likely with the Myford arrangement but something whose practical consequences I’m inclined to doubt. The wear is usually recognised by the gibs becoming impossible to adjust accurately as the screws and dimples only mate at whole turn intervals where the wear aligns.

There are inherently contradictory issues with screw end loading and gib tilt. The SouthBend style pointed screw provides line contact with the gib when new so it can accurately take up an angle matching the slide but the loading is high. The Myford ball ended screw style spreads the load but creates an unbalanced lifting couple due to the upper end of the screw being deeper in the gib than the lower which tends to tilt things.

The third way of doing things is to use screws with cylindrical ends engaging with a flat bottomed hole in the gib. This obviously provides much larger contact area at the end of the screw, reducing loads, supports the gib against sideways float and gives a parallel thrust against the gib. Impractical on small machines tho’.

Clive

[Andy StopfordParticipant @andystopford50521]

I had one of these mills with similarly haphazard seating for the gib screws.

I used a Dremel-type tool, with a rounded end diamond burr, to make the seatings deeper and more of a hemispherical shape. I also turned points on the ends of the screws.

Simple to do, and it worked absolutely fine afterwards. The fact that the hemispherical seatings probably weren’t totally in line with the screws didn’t seem to matter, and may actually have been advantageous, since the random out-of-trueness would tend to prevent the gib from sliding back and forth.

[Peter Cook 6Participant @petercook6]

just a thought. ARC stock spares for Sieg mills. They have gib strips for the SX1LP for a few pounds, and my memory says the SX1 and SX2 have the same X-Y table. Perhaps a call to ARC will produce a couple of new strips which may be easier than modifying the existing ones.

[David George 1Participant @davidgeorge1]

On most gib strips they are just a piece of stip metal and you could machine the dimples on either side so where the dimples are damaged you can just flip it over and re-machine the dimples on the other side. The side with the wrong dimples just stone the face smooth and ignore them as any loss of area will make no differance to operation and may help with lubrication slightly by holding oil. I then shim the bottom of the gib strip to lift the bottom face and assemble the slide and if possible I use the slide lock to clamp the gib strip. I then use a piece of silver steel turned to the thread internal diamiter with a conical point turned concentric on the end which I then harden and temper. This I use to punch through the screw holes which then marks the correct position of the dinples. I have a piece of steel bar with an angle machined on to match the slide angle and hold this in a vice but with a gap underneath at the correct angle . I set the drill point of a centre drill over the center dimple and clamp with a piece of round steel and a toolmakers clamp and drill the dimple to a good depth slide along and repeat till all dimples are drilled. Turn points on new grub screws and assemble.

[Dr_GMJNParticipant @dr_gmjn]

Thanks all. Plenty to re-read tonight.

This evening I succumbed to the inevitable and stripped the tables and cleaned everything down.

I’ll de-burr all the ragged edges just to make handling the castings less dangerous, then decide what to do.

I experimented by putting the domed screws in the Y-axis and temporarily assembling and adjusting the slider. Pushing it back and forth by hand resulted in random, quite severe binding. Not sure what it would be like with the original cylindrical ended screws. I’ll try the tomorrow.

Does anyone know why the x-axis has domed screws, and the y-axis had cylindrical screws? Neither axis has the divots machined well enough for it to have any obvious effect when interfaces with the screw ends.

Anyway, I ordered a 3mm ball mill, so I guess I’ll end up sticking with plan A which was to clean up the divots in the pillar drill, in conjunction with a printed angled jig. Is the dovetail angle standard or should I measure it?

I’ll use the end cap method for laterally locating the gibs. I might add small grub screws and lock nuts for fine adjustment.

 

[Dr_GMJN]

On Dr_GMJN Said:

[…]

Does anyone know why the x-axis has domed screws, and the y-axis had cylindrical screws? Neither axis has the divots machined well enough for it to have any obvious effect when interfaces with the screw ends.

[…]

Let me hazard a guess …

It’s because they came from two different shipments of screws !

It appears that whoever assembled this thing [possibly more than one individual] had no real understanding of what they were doing.

”Cottage Industry” at its worst ?

MichaelG.

[JasonBModerator @jasonb]

Though quite a sizable cottage these days. No doubt various machines are built to a price point so you gets what you pay for. Link

[video src="https://shopsource.singoo.cc/1114/video/iKWrGk7ijnFD6EzZ/SIEGAnhuiProductionBase~4.mp4" /]

[Dr_GMJNParticipant @dr_gmjn]

There’s no doubt they’re built to a price, and I suppose in many ways they are very good value; I ended up with a decent 10V using it after all. The issue is having to re-make things that I perceive to have errors caused by the machine. For example I wonder whether the stick-slip of the y-axis (which might have caused grabbing and poor finish on several components) was due to the gib? The x-axis has always been better.

Still not cheap in absolute terms (to me anyway), and it does get frustrating.

[HopperParticipant @hopper]

Yes they don’t exactly give them away so it is reasonable to expect them to work as advertised. You would not expect a Rolls Royce for the price, but a Ford Fiesta that just works would be nice.

Frustrating though that such a handy little machine is let down by such beastly careless assembly at the last stage of manufacture. I have read that there is a shortage of skilled labour in China so this is probably one result.

But, bit of fettling and you will get there. Sounds like the gibs could well be related to the poor finish problems you had. You should be able to feel nice smooth even progress when pushing the table or slides by hand without feedscrews in place once you are done.

Your head movement problems re setting rotary table etc earlier could be similarly caused too.

It looks in your pics like you could now get in and drill those divots in situ. Maybe a long series drill bit or a drill extension for those X axis screws if your pistol drill body is too fat. That may well be why the factory divots are so poor, the assembler just vaguely poked a pistol drill down there, cocked at an angle because the drill body was too fat to align the drill bit properly? Done properly with a bit of care they should be at least better than they were.

 

 

[Michael GilliganParticipant @michaelgilligan61133]

Curiosity about this little machine led me to ARC’s site, where I found both a review and a link to a CNC-ing thread on the old form.

[stay with me please, dodgy gibs have no place on a CNC machine]

Unfortunately that link takes us to the dreaded LEGO blocks 404 error … so this is the best I can do at the moment:

.

.

 

 

MichaelG.

 

[JasonBModerator @jasonb]

This Thread

[Michael GilliganParticipant @michaelgilligan61133]

Thanks, Jason

MichaelG.

[DiogenesParticipant @diogenes]

Just as a ‘FWIW’ – does it make any difference to the Y axis movement ‘action’ if the the 4 base hold-down bolts are loosened off..?

[Dr_GMJNParticipant @dr_gmjn]

Never tried that TBH.

[Dr_GMJNParticipant @dr_gmjn]

I didn’t bother waiting for the ball mill to arrive, I opted to mill flats on the gibs and be done with it. I turned a point on an M4 grub screw and screwed it into all the gib holes in turn to mark them.

I used the x/y table thing as a jig, and milled the flats in the pillar drill:

I drilled a step in the ends of each gib to locate the stepped grub screws rather than mess about with end stops:

That’s the short version, the flats ended up slightly misaligned in four positions so some re-work was needed.

I rebuilt everything and the best compromise between wheel stiffness and play (which is still a very poor one) is about 0.003”. I guess it should be less, but it is what it is.

While doing this I wondered (not for the first time) why the y-axis has to be extended outwards by about 1/3 of its travel before the spindle is over the centre of the x-axis? Seems a bizarre design.

Anyway, sliding the axes back and forth by hand doesn’t now seem to give any stick-slip, However adjusting the gibs for minimal play and fitting the hand wheels puts me almost right back where I started in terms of feel. I would say by the time I’ve re-fitted the DROs, the feel of the wheels will be back to the same horrible sticky mush it was previously, so basically it’s been a bit of a futile exercise.

Suppose it could be something to do with the lead screws, but I’m not getting into that, I’m done with it.

I’ve asked elsewhere a few times, but it can’t hurt to ask again: what’s a good quality mill about the same size of this one? Preferably equivalent quality to the ML7 as a minimum.

or

Are there any machine builders where I could dump this thing and let someone who knows what they’re doing get it as good as it can be? Messing about with machines like this isn’t how I want to be spending my limited spare time.

Thanks all.

[Dr_GMJN]

On Dr_GMJN Said:

Does anyone know why the x-axis has domed screws, and the y-axis had cylindrical screws? Neither axis has the divots machined well enough for it to have any obvious effect when interfaces with the screw ends.

 

Thought I’d covered this in my previous post. Obviously not

Screws with cylindrical ends go into plain parallel flat bottomed holes along the centre line of the gib strip which floats on the screws. The cylindrical end effectively performs the same longitudinal location function as dowels.

Simple idea but the accuracy constraints if it is to work properly with negligible slop are severe. I imagine that in practice only the end pair or even only one end screw is closely fit the into its hole with minimal slop. The others having slimly larger holes so a bit of variation in screw or hole placing can be coped with.

For obvious reasons the holes and screws must be “exactly” (ie to sensible tolerances) along the centre line of the gib and the sides must be clear of the casting. Any untoward contact with the castings of offset in the holes will cause the gib to tilt during adjustment. Less than desirable as the gib will no longer properly contact the dovetail. In a practical world there is enough tolerance for things to be made given reasonable care. But its a deceptively demanding way of going about things.

Screws with domed ends should go into conical holes in a gib strip sized to contact the top of the casting recess. Due to the angle of the gib the top part of the domed end has more “leverage” than the bottom side acting to force the gib upwards in contact with the casting. Which clearly stiffens things up. The friction of the gib sliding against teh casting makes this set-up less responsive to adjustment which must be done with the screw removed so you cancel whats going on. A good deal of reciprocation during adjustment is needed to ensure that the gib has properly settled into position. The inherent stability means gibs of this style can be run with closer clearances than the floating type.

Screws with domed ends have been used with reasonable success on floating gibs working into either conical holes or larger diameter one with a part spherical arc at the bottom. Adding a stabilising dowel makes things behave much better. This set-up is inherently sensitive to wear and generally seems to need more frequent adjustment.

Clive

 

[Dr_GMJN]

On Dr_GMJN Said:

[…]

I’ve asked elsewhere a few times, but it can’t hurt to ask again: what’s a good quality mill about the same size of this one? Preferably equivalent quality to the ML7 as a minimum.

[…]

It has its own limitations of course … but I think you might be happier with a a BCA if you can still find one … a bit bigger, and a lot heavier, but they were built properly.

MichaelG.

[Dr_GMJNParticipant @dr_gmjn]

I’ll be re-building it tomorrow.

what play should I be aiming for in the two axes, and how should it be measured?

Subjectively how much load should I be applying (and where) when checking for play?

I suppose in practice load is never applied in the way that rocking the x-axis with it fully extended would?

[Michael Gilligan]

On Michael Gilligan Said:

On Dr_GMJN Said:

[…]

I’ve asked elsewhere a few times, but it can’t hurt to ask again: what’s a good quality mill about the same size of this one? Preferably equivalent quality to the ML7 as a minimum.

[…]

It has its own limitations of course … but I think you might be happier with a a BCA if you can still find one … a bit bigger, and a lot heavier, but they were built properly.

MichaelG.

Thanks. Not heard of those. I was looking at Tom Senior, but there are a few variations (and difficult to find the smaller ones).

[Michael GilliganParticipant @michaelgilligan61133]

Bedtime reading:

https://www.lathes.co.uk/bca/index.html

Sweet Dreams !

MichaelG.

[Michael Gilligan]

On Michael Gilligan Said:

Bedtime reading:

https://www.lathes.co.uk/bca/index.html

Sweet Dreams !

MichaelG.

 

That’s more like it, although without looking at figures, it does look smaller capacity than mine, especially the x-axis. About the same price as a Tom Senior (I wonder if £2k is what milling machines of this sort of size are in general?), but a quick look on EBay shows far fewer BCAs

[Kiwi BlokeParticipant @kiwibloke62605]

I don’t wish to argue with Clive Foster, and am always open to correction, but I think he over-emphasises the importance of the way the end face of a cylindrical-ended screw contacts the gib strip. It’s the sides of the cylinder(s), contacting the sides of the gib strip’s recess(es) that is responsible for the end-ways restraint of the gib (which is not reliably provided by conical or part-spherical divots), and the avoidance of a ‘ramp’ which allows the strip to be forced into harder contact, if dragged lengthways, which are really important, and which this arrangement provides.

There are so many ways to skin a cat… Just to reduce the signal-to-noise ratio, if a gib strip can be constrained adequately (and I don’t want to define ‘adequately’ – it should now be obvious…), and one is prepared to make a new strip, or turn over the existing one, balls can be inserted between the screw ends and a plain face of the strip. Because the strip will be tangent to the ball, the force the adjustment screw makes on the strip (via the ball) will be normal to its face, which is a Good Thing*. I’ve improved a couple of machines by doing this, and it’s especially good for locking screws. Worries about point contact can probably be ignored, unless the strips are made from Chinese cheese.

* Occasionally, one comes across machines in which the gib adjusting screws are aligned at right-angles to the strip, but its usually impracticable because other bits get in the way. The Superior surface grinder is one, IIRC.

[JasonBModerator @jasonb]

Although a good BAC is nice it will depend on condition, You have had issues with your old Myford so you should be prepered for similar with a BCA. Do look at Z height and tool holding options as many with BCAs seem to have difficulty finding collets and the Z is quite small.

Maybe look at a Wabeco for a better build modern mill but be prepared to pay for it.

Might show the play on my X3 table later, you have already seen that what come off it is OK. Maybe you are just wanting to work to numbers that a steam model does not need, I do get that impression from some of your posts in the Victoria thread and as you say it made a very good Stuart 10V so must have been fit for purpose then. If you are expecting toolroom accuracy from a hobby machine price then you will always be chasing your tail. Simple example of getting what you pay for, replacement spindle bearings for my X3 would cost about £12retail, for the similar size KX3 at trade from Ketan would be a couple of hundread. The KX3 costs a lot more than just the price of some ball screws and stepper motors and a box of wires.

[JasonBModerator @jasonb]

Michael, were you hoping for some words about the gibs in that CNC post of John’s? I don’t think he got much further but would probably have milled slots into the gibs across their width.

[Author]

Posts