The saddle (or carriage) of the mini-lathe travels the length of the bed for turning operations and provides the base for the cross-slide to travel perpendicular to the spindle axis for facing operations. Since the saddle on the mini-lathe is a single piece with no readily reversible method of field adjustment; the saddle is fitted to the bed in the factory.
It had been determined by blueing that the inside V of the saddle makes contact at four points on the bed’s front inverted V. The rear contact surface is a flat area with no features that assist in preventing the saddle from twisting. It is the four points on the front that prevent twisting and set the angle relative to the bed that forces cross-slide travel to be perpendicular. If an adjustment in the angle is required; removal of material from opposite points on the saddle’s internal V will adjust the angle of the cross-slide while still keeping the saddle stable on the bed.
The cross-slide itself mounts on dovetails ground in the saddle. Provided the dovetails are straight and parallel the cross-slide plate will travel firmly and tightly provided the cross-slide gib strip fits well.
The last critical piece of the saddle is the surface that the bed gibs are bolted to. It is simplest if this face is parallel to the bed’s underside as that simplifies the shims required. However this surface cannot be proud of underside of the bed. If this surface extends beyond the bed there is no way to tighten the bed gibs sufficiently for the bed gibs to make contact with the under side of the bed. The result would be to have the saddle lift in operation.
It was initially appeared difficult to measure any of these relationships to determine if changes needed to be made since there is no obvious repeatable reference. Where did one start?
After long consideration and discussion it seemed the best choice was to assume that the flat portion of the cross-slide dovetail was flat and both sides of the dovetail were ground to the same plane. This simplifying assumption provided for a quick check on the V of the saddle, the flat portion that mated the rear of the bed, and the surface where the bed gibs mounted.
Consider that if the saddle in turned upside down and the dovetail flats placed on parallels such that the tops of the parallels are level then the following should also be true if the saddle is fit correctly
- A 3/8″ to 1/2″ (10-12 mm) drill rod placed in the V will be level
- The flat area that mates to the rear of the bed will be level
- The mounting surfaces of the bed gibs will be level
In other words, the horizontal flats of the cross-slide dovetail, the four points of contact in the V, the flat on the rear of the saddle and both bed gib mounting surfaces are all in parallel planes and that this can be demonstrated in a single measurement setup.
Using the cross-slide dovetails as the reference also provides simple and secure mounting of the saddle on a mill for any work required.
Returning to the immediate issue of the realbull mini-lathe; having finished milling the lathe bed and lapping the underside bed surfaces the gibs were bolted to the saddle (without any shims) and mounted on the bed. The gibs stuck and jammed into the underside of the bed. Shimming was attempted but the thickness of the shims varied widely (around 0.006″) from one end of the gib to the the other. The gibs had already been lapped and known to be straight. This revealed that the surface the gibs mounted against on the saddle were out of parallel.
It was entirely possible to get the correct set of shims to make the gibs parallel but it was not the correct solution. To fix this meant going back to the shop to mill those surfaces.
- Setup for milling the mini-lathe saddle
And so it was done. The photo shows the parallels being used to raise the dovetails and base the cuts on the horizontal portions of the cross-slide dovetails.
Fortunately for this mini-lathe, no further work was required on the saddle. After this milling operation the gibs mounted with shims around .014″ and the saddle traveled the entire length of the bed with no rocking, twisting or lifting in operation.
The perpendicular travel of the cross-slide was determined by mounting a indicator on the cross-slide and ensuring the forward most and rear most points of the spindle face are equal. Any deviation indicates the cross-slide is out of square as the the spindle had been previously fitted and made parallel to the bed.
The squareness of the cross-slide was double checked by mounting a chuck on the spindle and chucking a bent rod. The end of the rod was placed to inscribe a circle around six inches diameter. The DTI was used to measure the fore most and rear most points of the inscribed circle to ensure there was no difference in the measurements.
While no measurable deflection was seen measuring the spindle face, there was just over a thousandth of an inch measured on the six inch inscribed circle. I think it’s likely that I did not get the exact same point on the end of the rod between front and back and it may be that if the end of the rod were rounded to a true semi-sphere I would get a different, more accurate measurement. I decided to live with 0.001″ and move on.