Inconsistent Bed Way Dimensions

Since completing the tapered gibs, I found the carriage snug on the headstock end of the bed but loose on the tailstock end.  When the gibs were adjusted for the tailstock end, the carriages was quite noticably snug at the headstock, too snug - could bog down the motor in power feed, and I sure didn't want to crank through that either.  Funny that this didn't show up with the stock gibs, but maybe that is because they had to be kept so loose it just couldn't be discerned.

Measuring parallelism on the rear flat way is no sweat.  In doing that I found 2-3 thousandths difference along the ways.

Measuring parallelism between the inverted V-way on top and the flat way surface on the under side of the front way is tricky.  I felt it was not reliable to expect the top, flat surface of the inverted V-way to necessarily be parallel with V surfaces.  The V surfaces are what the carriage runs on so that needed to be the reference surface for any measurements.  I devised a way to measure this and found 3-4 thousandths difference along the inverted V-way, being thickest toward the headstock, thinnest at the tailstock end.  Roughness of the machined (non) finish probably didn't help this.  There was also primer/bondo exposed under there which could be contributing.

Inverted V Way Measuring Block

I felt it wise to make a "V" block that could placed on the inverted V-way and provide a flat upper surface from which to measure the dimension to the flat underside of the bedway.  This mean milling a 90* groove in a small piece of steel followed by turning it over and milling the top surface.  It was necessary to hold it, resting on a 1/2" reamer blank on a set of parallels to ensure the top surface was parallel with the V-groove.  The finished product rests snuggly on the inverted V-way and provides a stable platform for measuring way thickness and thus parallelism.

Measuring V-way Parallelism

Using the V-way block, it was easy to reliably measure parallelism of the inverted V-way.  It was thinnest near the tailstock,  which became the reference dimension, and gradually increase through middle and headstock.  These measurements were taken after I'd started filing the underside of the bed and so do not reflect the full extent of difference.  Even so, at this stage the photos shows about 0.0025" difference.  I compared measurements from the top flat surface of the inverted V-way and confirmed it was not parallel with the V-way surfaces.  I also didn't like how the micrometer anvil was only half on the top surface with the rack installed, although the rack is removed for filing of the bottom flat.

FIXING THE NON-PARALLEL BED WAYS

Making a File Holder

How to fix the bedway not being parallel?  I've been interested in learning how to scrape, so I could restore an old Atlas lathe, but having not invested the time and energy to learn, scraping is not an option.  My two options were milling or filing.  Milling was out of the question as I don't have local access to a mill large enough to do the job and the setup on that would be a job in itself.  So, filing seemed like the only real option.  It seemed important to me to ensure filing of the bedway underside maintained a flat and perpendicular to the vertical side, in the hope of it being parallel across the way as well as along it.  Recalling a file holding fixture I use to sharpend edges on my skis, I decided to make a similar fixture suitable for the lathe.  I wanted the file held exactly at 90* with a good bearing surface to run along the vertical surface of the bedway just below the inverted V way.  Starting with  3/4" square x 4" long keystock, I milled a 1/4" wide slot, offest from center a bit, and milled the top face to clean up surface.  Then I drilled and tapped 3 holes for mounting screws for M6x1 SHCSs.  The finished holder can take any of the three sizes of files I have; only the small and medium were needed for this job.

Filing the Bottom Bedway

Maintaining flatness of the file, and alignment with the sidewall is key in this operation.  The file holding fixture holds the file at pretty close to 90*, so it is up to the operator to maintain that angle in use.  Holding the file fixture firmly against the sidewall while applying upward pressure to do the cutting may seem easy, but it is easy also to tilt the file inadvertently (obviously exaggerated for demonstration), which could have adverse affects.  To protect the leadscrew from filings, I put a paper tube left over from Christmas wrap over the leadscrew.

The finished under surface is quite a bit smoother than the stock bed (sorry forgot to get a photo of it, but we all know it).  I took off the rack for filing on my lathe, but it isn't stricly necessary.  The one concern I have is that the newly filed surface is too smooth to hold oil as the gib slides over it.  This is where scraping/flaking would be nice, but...   Isn't it interesting how the lathe was painted its hammerite green AFTER the rack was installed.  But I haven't found any paint in the rack teeth.

Results?

After several iterations of filing and mic-ing, I got to the point were I felt it was good enough.  Measurements of the tailstock end, middle and headstock end show only about 1/2 thousandth difference.  That has got to be good enough, right.  It seems that way to me.  The carriage moves smoothly end to end now.  I cannot discern any movement and indicating it shows no more than 1/2 thousandth wiggle.  I'm happy and will try using the lathe as it is for a while to see if there seems to be any difference (improvement) and if there are any problems that arise.

Thanks for listening.  Rick.