What's New


Workshop Stuff


Moving the mill

Spindle Noises

ShumaTech Digital Readout

ShumaTech DRO Continued

DRO-350 Repairs

South Bend 9" Lathe

South Bend 405 Lathe Bench

Grizzly Mill Revisited

Surface Grinder Rebuild

Surface Grinder Continued

Grinder April 6, 2008

Grinder April 20, 2008

Grinder August, 2008

Grinder September, 2008

Grinder November, 2008

Grizzly G3103 Mill

Grizzly G3103 Mill
Rebuild - Part 2

Moving the Shop

Moving the Shop 2

Bringing Home a Sheldon 12" Shaper

Sheldon 12" Shaper 2

Sheldon 12" Shaper 3

Sheldon 12" Shaper 4

Sheldon 12" Shaper 5

Sheldon 12" Shaper 6

Sheldon 12" Shaper 7

Sheldon 12" Shaper 8

Sheldon 12" Shaper 9

Way Alignment Tool



Email Jim


Sheldon 12" Shaper - pg. 2
July 23, 2015

After getting the shaper home, I decided that before I put power to it, I would clean up the surface rust, and check and lubricate the mechanical systems. Once I had the shaper cleaned and properly lubed, I would make some test cuts so that I could assess its condition before deciding if I want to strip it down to its individual pieces and do a ground up restoration. The shaper has some cosmetic issues, but after a few day's inspection, the machine seems to be in pretty good mechanical shape.

An artist's rendering of the Sheldon 12" Shaper from an advertising brochure. 

I was a little concerned about what I first thought to be a major cosmetic injury to the shaper. There's a 6"X 9" rectangular hole in the front of the base to the right of the Sheldon emblem. On this shaper, there is currently a piece of medium density fiberboard (MDF) covering this hole with a toggle switch attached that controls the power for the variable frequency drive (VFD) and an electrical junction box attached to the MDF that routes the three-phase motor wiring to the VFD. Unless the base was damaged in that area, I couldn't think of a reason that someone would put a hole in the shaper's base. After some more reading and looking at pictures, I determined that the hole was supposed to be there and what was missing was the original switch plate
. It appears that the earliest of the Sheldon/Vernon shapers didn't have the switch plate. Since I haven't yet found a source for determining the manufacturing date of my shaper, the hole and accompanying switch plate may help me to put a date to this shaper. (Update) I have now gotten confirmation that serial number 8-9214 was manufactured in 1957. My shaper's serial number is 8-8956, so it was manufactured a bit earlier. How much earlier still remains a mystery.

I figured that the best way to clean the shaper, since I wasn't going to strip it down completely, was to start at the top and work towards the base. Hopefully this will keep the dirty solvent from getting the cleaned areas dirty again. I got started by removing the tool head assembly at the end of the ram. The tool head includes the vertical tool slide, which enables you to set the depth of cut and the clapper box, which allows the tool bit to pivot away from the work on the backstroke of the shaper's cutting cycle. There was a fair amount of surface rust and dried up lube on the whole front of the assembly. My observations lead me to the conclusion that this shaper hasn't seen a lot of recent use. The lube on all of the sliding surfaces had gotten thick and sticky. I couldn't move the clapper box easily until I soaked it in some mineral spirits.

While I was reassembling the tool head, I noticed that the screw that sets the gib depth for the tool head slide barely engages in the bore that supports it. I thought that this was a bit strange on an older machine. As the gib and dovetail ways wear, the gib adjusting screw would need to be seated further into the bore so that it would push the gib deeper into the ways. Instead, the portion of the screw that fits the bore sat above the bore. Why the gib adjusting screw was standing a bit proud of the supporting bore was puzzling. I checked the dovetail ways for nicks and raised metal and found none. The gib shows very little sign of any wear and appears to be original, but I still don't understand the position of the gib adjusting screw. At some point, I am going to need to disassemble the tool head again and check it on my surface plate.

The down-feed tool slide, tool head holder, and clapper box have been cleaned and oiled.
The tool post and tool holder are in pretty good shape. I received 2 holders and want some more.

Back to cleaning. I brushed on more mineral spirits to the left side of the machine and let it soak in while I got to work removing the vise. The vise has a few spots on top of the jaws where some previous user had let the tool hit the vise. The marks aren't deep and the removable jaws are in fair shape otherwise.
The moveable jaw is a little loose, but I will deal with that once I see how it does at work holding. The vise appears to be original, which is very nice. Missing vises on shapers seems to be all too common and they are not easily found. I got the vise cleaned up and started to work on the table. The table is in very nice condition aside from a couple dings on the top edges of the sides that have raised bumps on the table's top surface. I used a small granite straight edge to locate the raised areas and carefully filed them down to just above the surface of the table. I then switched to a fine grained sharpening stone to remove the last of the bumps so that I didn't take off much metal from the table.

Once I had the edges of the table flat, I used the straight edge to check the remainder of the table. I did this by performing pivot tests with one of my small straight edges. The pivot test is nothing more than setting the straight edge on the table and pushing on each end of the straight edge 90 degrees from its long side - a motion similar to spinning a propeller on a rubber-band powered airplane - but the motion only lasts for an inch or two. On a flat surface, the straight edge should rotate or pivot around a point between 20% to 25% the length of the straight edge from the opposite end you are pushing. This is approximately where the Airy point would be. Airy points are used in metrology as the two points to support an object in such a way as to minimize bending or droop under its own weight. The points are symmetrically arranged around the center of the object and are separated by a distance equal to approximately 5/9ths of the object's total length. I have read that the actual divisor is 0.577, not the 0.556 that you get from 5/9ths. The equation for determining the Air points is to multiply the length of the straight edge by 0.577, then center the result on the straight edge. For a 12" straight edge this equals 0.577 X 12" = 6.924". Subtract 6.924 from 12 = 5.076. Divide 5.076 by two equals 2.538. So the Airy points are 2.538" in from each end of the 12" straight edge.

Using a small straight edge to detect some dings and raised areas on the shaper table.
With the help of the a couple files and a fine stone, I knocked down the ding on the near side of the table.

The observation that a straight edge will pivot around the Airy points is something that I have noticed during many pivot tests while scraping flat surfaces. The pivot point may not exactly coincide with the Airy point, but it should be close. If the straight edge pivots at any point other than the opposite Airy point, then there is a high spot somewhere on the surface and the surface being checked isn't flat. In general, if the straight edge pivots around the center, the surface you are testing is convex and if it pivots from the opposite end of the straight edge, the surface is concave. A piece of swarf or grit under the straight edge will make the straight edge pivot on that piece of swarf, so clean tools and table are a must. If your straight edge is longer than the table and the table is flat, the straight edge will tend to pivot at about 1/4 of the table's width on the opposite side from where you are pushing. By being able to interpret the way a straight edge pivots on a surface, I find that I can easily get a pretty good read on how flat a surface is. In the case of this shaper table, it is pretty flat.

In the left picture above, you can see a divot to the right of the straight edge. The straight edge pivoted around this divot because the metal around the divot had been pushed aside and up when the divot was made. I stoned this area flat and you can see a shiny ring around the divot where I removed some metal. At some point, I will ink up a surface plate and do a more complete job of checking the table for flatness, but at this point, I just want to get the machine's table clean and reasonably flat to test the shaper's operation and get a general idea of its accuracy.

With most of the dings taken care of, I try a pivot test to search for more high areas.
Since the granite is pivoting around a point about 25% of the table width, the table is pretty flat.

Three nights of cleaning were needed to remove enough grime to be able to see the cream colored paint that covers the inside of the base. It doesn't look like the inside of the shaper had been cleaned in a very long time. I went through about a gallon of mineral spirits, wore out a 4" paint brush and shortened the bristles on a soft stainless steel brush just trying to get the mixture of oil, grease, dirt, and sawdust, out of the base. On the plus side, I did find a couple zerks (grease fittings) that I hadn't seen before. I gave each of these a couple pumps of some good quality NLGI #2 GC wheel bearing grease. The only manual I have for this shaper shows exploded parts views and has a little information on lubrication. The information specifies NLGI #2 bearing grease with a melting point of 200 degrees F. NLGI refers to the National Lubricating & Grease Institute. The #2 refers to the thickness of the grease. The fact that the manufacturer specifies bearing grease makes it type G. In automotive applications, type G is for wheel bearings and type L is for ball joints and other suspension components. Grease also has a rating for severity of service (A, B, or C) A is for the least severe service and C is for the most severe. For this shaper, since it won't be fording rivers or participating in the Dakar Rallye, any of the severity ratings will do, though I happened to have type C on hand.

To summarize the shaper's lubrication needs, there are only 2 types of lubrication that are specified for this Sheldon/Vernon shaper. The NLGI #2 for all of the Zerk fittings and oil for all other lube points. For the oil, the specs from the parts manual are "Use a neutral mineral oil, viscosity 250 to 300 seconds at 100 degrees, approximately SAE 20". This is a pretty close match to Mobile Vactra #2 way oil which is ISO grade 68, SAE grade 20, 350 seconds at 100 degrees F. If you buy the "house brand" at McMaster-Carr, the SSU (Saybolt Seconds Universal) rating is 300 seconds at 100 degrees F. That's a perfect fit. Either one would be a darn good match.

The outside isn't that dirty, but it isn't exactly that clean either. Wait until we look inside.
Inside the base, it's another story. That's a 1/4" of dried oil, grease, dirt, and sawdust on the floor.

I finished with as much cleaning as I though was necessary to inspect the machine and get it ready for a test run. The shaper is not clean by a long shot, but it's a lot cleaner than it was. I could now put some of the information I had learned by reading the metal_shaper Yahoo group. Since this group is for all shapers and not just for the Sheldon, rather than start at the first post and read everything up to the present, I searched for Sheldon and am reading those posts first. I read about checking the sliding block in the Scotch yoke and how the clearance between the block and the sliding ways can cause noises and worse. In one post I read, a Sheldon owner had been able to slip a 0.004" feeler gauge between the yoke and block. The general consensus was that four thousandths was not that much wear. I checked my block to yoke fit and couldn't fit a 0.002" feeler in regardless of the position of the block. If the less than 0.002" is a true reading, it would help confirm my suspicion that this shaper hasn't seen a lot of use. There are still some oil pockets (flaking marks) visible on the ram and column ways which also adds to the idea that the shaper has only seen limited use.

It took lots of mineral spirits and scrubbing to get the base floor to this point.
The Reeves drive. To get all of the surface rust removed, I'll need to pull this assembly out.

As I noted, the shaper is missing the original switch plate from the front of the machine. There is a 6.25" wide by 9.0" tall cut-out to the right of the Sheldon/Vernon name plate. This has been covered over by a piece of MDF with a small toggle switch and electrical junction box that the previous owner had hooked up to a VFD to power the 3-phase motor from single-phase 240. He had run electrical conduit out of the box and across the right side of the base. He did a decent job of wiring it, but I wasn't really happy with the conduit running over the top of the machine's base. It also appeared that some other past owners had one had run the power wires through the base and out the rear of the machine. Another had run the wires out the side This was accomplished by cutting rectangular pieces out of the side and rear panels. Again, not my idea of the best way to do a clean job, however, I preferred that approach better than having the conduit run over the base. For the time being, I will rewire the machine with the conduit hidden inside the base and exiting through the cut-out in the rear panel until I can figure out how the wiring was meant to be routed.

Rather than keep the 7" by 10" MDF in the front of the base, I cut some 1/4" thick 6061 aluminum sheet to make a new switch plate. I radiused the corners, bored a 1" hole for the new switch, gave the new plate a good sanding, and gave it a few coats of primer. I followed that by a giving it a couple coats of clear enamel. Primer alone wouldn't last very long in an oily environment. If I decide to strip and paint the shaper, I'll sand off the clear coat and shoot the new color over the primer.

The old switch plate. This picture was taken after unloading the shaper off of the trailer.
The new switch plate. The small plate below the Sheldon plate is from a naval ordnance station.

I installed an emergency shut-off type switch in the new panel that will hook into the VFD and serve and the power switch. This is a big red "push to disconnect" switch. The on/off switch controls the WEG CFW-10, 7.3 Amp variable frequency drive that also came with the shaper. In addition to the on/off switch, the shaper has a manual clutch which disconnects the ram from the Reeves drive. Since the clutch lever is located on the right side of the machine along with the other controls, it should be always within reach when operating the shaper. The clutch is the best way to stop the ram if something should go wrong, but kicking the big red button is now another option.

Since the shaper has a Reeves variable speed drive to control the ram speed and also has a reduction gear for "back gear" type speeds, the previous owner had set up the VFD to just perform a soft start - slowly ramping up the frequency from zero to 60 cycles per second. For simply turning the VFD and shaper on or off, there's only one switch needed. While in theory, I shouldn't really need to use the variable frequency function of the VFD, having the ability to tune the frequency and thus control the speed can be helpful to get rid of vibrations when you are trying for the best possible finish. With this in mind, I mounted the on/off switch 2/3 of the way up the new panel. This left room for a potentiometer to be mounted below if I needed it for speed control. The frequency can also be controlled by the keypad on the VFD.

The cut-out that needed to be fixed and the new piece of sheet metal that was cut and bent to fit into the existing hole.
The panel has been sanded to remove the paint. The piece of sheet metal has been fit in place and clamped for silver soldering.

As I said, one of the side panels and the rear panel had been notched. I am assuming that this was done at different times to route the power wires. As I was going to run my conduit out the notched rear panel, this one didn't bother me. The notch on the side panel, which would be more visible, bothered me. I figured that now was as good a time as any to fix the side panel.

I had some 16 gauge sheet steel that happened to be the same thickness as the door panel. I cut a piece and filed it to fit, then made the bend using a shop vise as a metal brake. I bent the piece over the rounded edge of a piece of scrap metal so that the bend would match the bend in the door panel. Since I was going to silver solder the patch in place, both the patch and door needed to be clean metal. I sanded the door and patch piece and made sure that the patch fit tightly, then applied lots of flux and did a little silver soldering. To finish the job, I filed down the welds, hit the door with a couple grits of sandpaper with the orbital sander, then gave it a few coats of paint. The door came out nice and you would be hard-pressed to find where it was repaired. By the way, I really like the art deco handles on the access panels. A couple of them have some dents, but if I decide to rebuild the machine, I think I can take the handles apart and work them back to their former glory. However, I am getting ahead of myself.

The silver soldered repair has been filed and sanded to blend it in to the panel.
A couple coats of primer and a couple coats of clear and the door is whole again.

With the cleaning done for now, it was time to build some adjustable feet, move the shaper into its new location, and wire it up.

I considered bolting the shaper to the shop floor, but I am unsure if where I am putting it will be its final location. Once I make enough room in the separate side loading stall that I use for tractor and implement storage and also houses my woodworking table saw, I would like to move the miter saw and eight foot bench that it's on into that area. One room for wood and one for metal is my eventual goal. I am getting really tired of cleaning sawdust off of my metalworking machines.

I have built a few sets of adjustable feet for other machines, but not for a machine that tends to walk itself across the floor like shapers have a habit of doing. I thought I'd try using some of the one inch thick "horse stall mat" on the bottom of the metal pads to hopefully give them some traction on the slick shop floor. I got the mat from a local farm supply store a few years ago and cut a long strip to attach to the bottom of my smaller snow blade when we still lived in a home with a paved driveway. It did a good job of protecting the driveway from scrape marks. I now have the remainder of the mat on the shop floor in front of the lathe and nibble off pieces when I need some for a project. Since the shaper will put a side load on the feet, I thought I would not only glue the mat to the bottom of the 3/8" thick steel pads, but I'd also run the all-thread through the steel pad and into the rubber mat to try and keep the metal pads from sliding off the mats. I roughed up the bottom of the pads and cleaned the tops of the mats with brake spray and then glued them together with construction adhesive. We'll see how that works.

The feet are assembled from 2" X 3"X 3/8" steel and 1/2"-13 all-thread.
The feet are assembled and the mats have been glued to the bottom of the metal pads.

With the feet assembled, it was time to move the shaper into location. I needed to spin the shaper 180 degrees to get it headed in the right direction, then roll it across the floor using a Johnson bar and pipes. When I got the shaper to its new location, I used my make-shift gantry crane to lift it, remove the skids, then set it down on the new leveling feet. As I said in the last article on moving this shaper, this gantry has been one of my better purchases. It took less than 15 minutes to lift the shaper, remove the skid, install the feet and place it back on the shop floor.

With the shaper in place, it was time to level it. My shop was originally built as a four stall garage with a narrow room separating the three stalls I use as my metal/automotive shop from the side loading end stall I use for housing my smaller tractor and woodworking stuff. Since the building was built as a garage, the concrete floor has a slight slope toward the garage doors to help move any water to the outside. This slope is a little steeper near the doors. As I began to level the shaper, I found that the rear (garage door side) was lower than the front by at least a half inch. Having to crank down the rear adjustable feet means that the rear of the shaper is going to be supported by more of the 1/2" diameter all thread than I had thought. I am already thinking that I should be using wedge-type adjusters to give the shaper a more solid base. I guess I will find out when I run the shaper for the first time. Making up some wedge-type adjustable feet might be a good first project for the shaper.

Rolling the shaper into position. I need to hook up that pellet stove (right) before winter comes again.
The make-shift gantry crane made easy work of lifting the shaper the needed four inches.

To level the machine, I set my 0.0002" per 10" resolution box level on the shaper table and did my initial leveling diagonally. I didn't know if the shaper's tilting table was set plumb to the machine or whether the table had succumbed to gravity and was low on the operator's end, but it's a lot easier for me to level a machine by adjusting the diagonally opposed feet rather than side to side and front to rear and the table was the only place large enough to place the level diagonally. Once I got the adjustments close, I switched to using the machined casting that supports the left side ram dovetail for the front to rear level and using the cross rail for the side to side level. Yes, I tightened up the gibs before I did this. With the shaper leveled within one graduation (0.0002") in both planes, I checked the table level. The tilting table was about two graduations low on the right side even though the pointer was lined up on zero. The pointer is adjustable, so this is an easy thing to correct. A bit more concerning, but not unexpected, was that the front of the table was lower than the rear by about four and a half graduations, which is about 0.001" per foot.

I will have to double check my "Machine Tool Reconditioning" book by Edward F. Connelly, but I don't think he has a chapter on shapers. However, from studying his other examples with movable tables, I am guessing that the table actually wants to be a little high on the front to compensate for wear and the deflection that the table will see when the shaper is in operation. Milling machine table ways are scraped in or ground to be up to 0.001" high on the front of the table (Y axis). However, since there is a support under the front of the shaper table, I may be wrong and the table may want to be set parallel to the ram. This is all just a mental exercise at this point. Before I start thinking about making changes in adjustment, I need to try squaring a block under power, as I do not know if the ram is even in the same plane as the casting I used to set the front to rear level. I also don't know how worn the ram bearing surfaces are. Hopefully some test cuts will give me a lot more information than I have now.

Update: I was able to get a copy of "Testing Machine Tools: For the Use of Machine Tool Makers, Users, Inspectors and Plant Engineers"  - 1978 by Georg Schlesinger (Author), F. Koenigsberger (Author), M. Burdekin (Author). This book has a section on testing shapers. The test requirements for the ram being parallel to the plane of the table top is listed as being between zero to 0.015mm per 300mm,  which works out to be from about zero to 0.0006" per foot with the provision that the table should be raised in the front. This confirms my suspicions that the shaper setup is similar to a milling machine. I am now pretty confident that I will need to do some work to get the front end of the table to be a bit higher, but I still have a lot to do before I reach the point of actually making those adjustments.

The new feet are installed and it's time to get to leveling the shaper.
Checking the level along the top of the cross rail. I am using a 0.0002" per 10" box level.

I finally got the shaper wired up and turned it on for the first time. I checked to see which direction the Reeves drive pulley was turning. It turned counter-clockwise when looking at the pulley from the left side of the machine. This is the correct direction to allow the ram to move slower on the cutting stroke and faster on the return stroke. If it would have run backwards, it is a simple matter to swap any two of the three leads from the three phase side of the VFD to the motor, but I got lucky. I turned the speed control wheel all the way to slow and pulled the clutch lever out. The ram engaged smoothly and the machine was under way. I turned the speed control to speed up the ram strokes and noticed a slight ticking or knocking noise coming from the area of the sliding block or top link once I got to about half speed and above. I would check into the noise little further soon, but wanted to check the so called 'back gear' operation first. The speed reducer was called a back gear by Sheldon, but it really isn't in the true sense of the term. It is comprised of a couple of planetary type reduction gears and I had read that most all of the Sheldon shapers were noisy when running with the speed reduction engaged. I enabled the reduction gears and ran the Reeves drive from the slowest to the fastest setting. Yeah, the reduction gears are a bit noisy. They are straight cut spur gears in a housing that seems to amplify the sound of the gears meshing. It sounds a lot like my South Bend lathe in back gear. It has a definite whirring sound, but nothing that I was concerned about.

Checking the front to rear level on the casting that encloses the left side ram dovetail ways.
The operator end (front) of the shaper table slopes down about 0.001" per 12".

I grabbed a chunk (4"X 3"X 6") of cast iron from my scrap box and mounted it in the vise. I set the stroke to be about an inch longer than the 4"width of the chunk of iron and set the ram to start its stroke about a half inch before the stock, then manually turned the Reeves drive pulley until the ram passed the stock and started its return stroke. I made sure that this was about a half inch past the end of the stock. Running the ram by hand let me know that I wouldn't crash the tool bit or tool head on my first attempted cut. That wouldn't be a good way to christen the shaper. I honed the finishing tool bit and installed it in the tool holder and was going to try a 0.005" depth of cut with a 0.010" feed. I thought it best to take a light cut on the first attempt. I was quite impressed with my first cut on a shaper. The shaper is very quiet, much like a lathe. Just a sound like frying bacon as the bit peeled off a thin layer of cast iron. The finish on the cast iron was very smooth. It puts the finish produced by my milling machine to shame. It is also fun to watch the cuts being made. My wife had come out to the shop to ask me a question and saw the shaper cutting. She found the rhythmic action of the ram pushing the tool bit through the cast iron almost as interesting as I did. After running the shaper for a half hour or so and getting used to sounds it made while working, I decided that it was time to check the sliding block fit again.

The bronze sliding block is held in place by one 3/8"-16 tpi bolt. The parts diagram shows that the bolt is supposed to be secured by a lock washer, a machined flat washer, and an 1/8" cylindrical pin that prevents the bolt from backing out. I pulled the pin with a pair of end cutters and unscrewed the bolt. If the parts list I have is to be believed, the lock washer was missing. However, given the length of the pin, the lock washer would need to be very thin to allow the pin to lock the bolt in place. I am thinking that there was no lock washer on this bolt despite what the parts manual shows. I think that this manual is for an older model than I have. 

I measured the width of the sliding block where it fits between the ways of the Scotch yoke and came up with a measurement of 1.7265". I used an inside bore gauge to measure the inside of the Scotch yoke rails and came up with 1.7280". With a 0.0015" difference between the width of the block and the width of the yoke, there doesn't appeared to be very much wear on the bronze block. I measured the block in many different places and still came up with 1.7265". There was also next to no wear on the yoke and what I assume to be the original factory scraping was still very evident on the cast iron surfaces that the bronze block slides against.

Testing the axial play between the bronze sliding block and the sliding stud.
The pointer on the DTI has moved counter-clockwise a little over six thousandths.

I measured the cylindrical bore of the sliding block and came up with 1.127". The crank pin that the block rides on measured 1.125". I am guessing that when the shaper was new, there was a little less clearance, but 0.002" didn't seem like that much play. What I did find was that there was about 0.006" axial play where the sliding block bore slides along the crank pin. The Sheldon parts list calls this pin the sliding stud. With the bolt and washer installed, I could move the block axially - left to right when facing the front of the machine - about six thousandths of an inch. I noticed that there was a circular wear pattern on the rear of the sliding block where it contacted the sliding stud. I measured the wear and found that the recess was around 0.005". I figured that the best way to take up the play was to add a couple shims behind the sliding block to remove some of that 0.006" play. I ordered a shim assortment from McMaster - 3088A937.  As long as I was placing an order, I figured that I should take care of the oiling felts. The sliding block has 4 round pieces of 5/16" diameter felt cord that spread the oil from the sliding block to the Scotch yoke rails. These felts were hard as a rock and might be original to the machine. It did not appear that much oil would get to the yoke rails in the felt's current condition. I ordered some 5/16" felt cord and also ordered a 1/4" thick felt sheet to make up some new ram way wipers. The old felts actually measured 3/16" thick, but they were compressed by years of use. By using 1/4" "F1" felt, I hoped to get good contact between the felts and the ram dovetail ways. Replacing the oiling felts is always a good idea.

I received the shim pack and used a micrometer to measure out 0.005" of shims. I installed them, reinstalled the sliding block, and measured the free play. It was now at a little more than a thousandth. I pulled the sliding block back out and cut some of the new 5/16" cord to replace the hardened felt oilers.  I oiled up the new felts and installed them and then reinstalled the sliding block. I removed the four wipers. Two from the front of the ram and two from the top dovetail ways of the cross rail. To make a pattern to cut the felt, I used the old oily felts on top of the new felt sheet. I pushed them into the new felt and they transferred their dirty oil in the shape that needed to be cut out. I added a bit of width to each of the sides of the felts to allow them to be compressed by the ways. Before I installed the new felts, I cleaned off the dirty oil with some automotive brake spray. Time to do some more cutting.

The 5/16" felt cord has been threaded on to the spring. Once the pieces were installed, I cleaned them with some brake spray.
The 1/4" F1 felt sheet has been cut by pressing the old oily felt into the new piece, then cutting around the pattern.

For my first cut, I had used the tool bit that had come with the machine. I had just honed it and put it to work. The bit was 1/4" wide X 3/8" deep and was ground in the pattern of a cast iron finishing bit. This bit was ground so that it presented a cutting edge at 90 degrees to the direction that the ram traveled. I had been reading about shaper bits and holders and wanted to try directly mounting a bit into the tool post without using a tool holder. I bought some rectangular 1/2" X 3/4" X 4.5" long HSS tool blanks and ground a bit that had a 5 degree shearing angle to the cutting edge, a 5 degree top rake, as well as 5 degrees on the front and side reliefs. This bit gave me an even nicer finish to my cast iron block. For the next tool bit, I will try a larger shearing angle for use with steel.

My first project was to square up the cast iron block on four sides so that I could assess how accurately I could plane surfaces. I had found about 0.001" drop over 12" in the front when I had checked the table with a level. After carefully mounting the block in the vise on some paper scraps, tapping it down while tightening the jaws to make sure the work was fully seated in the vise, and then proceeded to cut the first of the four faces. When I completed all four faces, I moved the block to a surface plate I found that I had about 0.0015" difference in height over the four inch front to rear span. That's quite a bit in only four inches. However, the next measurement was a lot better. When I measured the height of the side to side cut, it was only off by about 0.0002" over 6 inches. This measurement is close to what I observed with the level. The next step is to realign the tilt of the table using my level. I will also remove the vise and put it on the surface plate and check it for being parallel between the rails and base. Once those steps have been completed, I will plane the surface of the block again and check the results. Since I have very little seat time on a shaper, I am trying to work one step at a time to see what happens with each change I make.

The first cut was made with a straight cut bit with about 3 degrees side and vertical clearance.
Recutting the top of the cast iron with a slightly shearing tool bit mounted directly in the tool post.

Well, forget the last statement about the accuracy of the cut. Tonight I stripped down the swivel vise and put it on the surface plate one piece at a time. The block only came out as square as it did by chance. For a machine that is probably 60 years old, I shouldn't be surprised by the outcome, but I was. Nothing on this vise is square with itself or the shaper table. The rails of the vise that support the movable jaw aren't parallel with the bottom of the vise and the supporting ring on the top of the swivel base is not parallel with the base's bottom surface. When I pulled off the fixed jaw's removable face, the rail under it was about 0.004" higher than the rest of the rail on the right side and 0.002" on the left. The rails were also not a uniform height from the base. The rails varied in height by a few thousandths of an inch. I have never seen the rails on a vise wear this much, so all I can think of is that it has been poorly resurfaced at some point.  I knew that I needed to do some work on the vise jaws, but I didn't figure I would have to rebuild the whole vise. I have some work ahead of me. On the plus side, I haven't done any scraping since we moved into this home a few years ago and I've been wanting a scraping project. I can only do home repair projects for so long before I want to get back to working with metal. Now I have an excuse to pull out the scrapers.

The last time I rebuilt a milling vise, it was an import and I found that trying to get the swivel function to repeat with any semblance of accuracy was hit or miss. I found that there's enough flex between the vise and the swivel base that I don't use it unless I have a specific need and am prepared to lose some accuracy. I now just mount the vise directly to the table and keep the swivel base wrapped up in a cabinet. Unfortunately, I can't mount the shaper vise directly to the table without the swivel base. The table T slots are on three inch centers (six inches between the swivel base mounting bolts) and the vise mounting holes are on five and a half inch centers.  I will say that even in its current state, the accuracy of this swivel vise is a lot better than the import was before the rebuild, but putting the swivel base on the surface plate and checking the ring that supports the base of the vise was out by 0.0014" as I ran my 0.00005" resolution DTI around the cylindrical mounting boss. I was hoping that there was a ding or burr on the machined base that was causing the issue, but that wasn't the case. I inked up a surface plate and marked the bottom of the swivel base and found it to be a pretty nice job of machining. No obvious high spots from the base being dinged or dropped, but there were some low spots that I'd need to deal with. If I had to guess how the bottom surface of the base was machined, I would guess that it was turned on a lathe. There are concentric circles that radiate from the center to the outside of the base. The ink pattern from placing the bottom of the base on a surface plate with a thin coat of Canode marking fluid reminds me of a circular moire pattern. There were some darker markings on the four outside corners - just past the cut-outs for the four hold-down bolts - but scraping these down did very little to change the relationship between the base and the ring that supports the vise. Time to get out the Biax and do some power scraping.

On the left side of the base you can see the pattern from the original machining.
After a few passes with the Biax scraper. The base shows color across the whole surface.

After a few passes with the Biax scraper, the bottom of the swivel base was showing color across the whole surface. The next step was to refine the surface. I'm thinking that 20 - 25 points per inch will be good for this surface. I cleaned the base and put it back on my big surface plate and checked the ring that the vise mounts to. As I suspected, it was still uneven. Since the ring sits lower than the center boss that centers the vise, it wouldn't be easy to machine the ring with the tools I own. The ring has a two inch segment that is about a half-thousandth lower than the rest of the ring. This area lines up with the the circular area shown in the left picture below (right side, below the bolt hole). The only reason that I can come up with for this is that the two mounting bolts weren't tight enough at some point and the vise was rocking on its swivel base. Since the movable jaw end of the vise is unsupported, there would be a lot of force being directed at this end of the vise base. I will use a file to get the ring close to an even height, then scrape the bottom of the vise flat and use it as the master to scrape the ring flat. When this is completed, I will machine the vise rails parallel with the vise bottom and base.

Speaking of vise swivel bases, I've read is that having four bolts to mount the swivel base to the table is one of the things that distinguish a shaper vise from a milling vise. Milling vises normally only have two bolts to hold the swivel base to the table. While the four bolts certainly secure the base better than two, there are still only two bolts holding this vise to the swivel base and it is this connection that seems to introduce a lack of rigidity. I have to admit that I don't quite understand the point of having four bolts on the base and only two on the less rigid connection. While reading about shapers and shaper vises, I found a drawing from a 1940s introductory course on using the shaper. The manual was written to familiarize an apprentice with shaper operations as a part of war materials production during WWII. One of the drawings shows a vise that has a couple of bosses that help to support the portion of the vise that overhangs the circular mount. (Expand the picture below right.) I think that these bosses would add some more rigidity to the vise to swivel mount junction and help to prevent the wear I have found on the bottom of the Sheldon's vise. I'm thinking that I could grind some blocks to attach under the movable jaw side of the base to increase the rigidity.

The vise bottom isn't flat. Some color on the upper right and a bit on the lower left edge. The circular mount area below the bolt holes shows some wear. This will need to be addressed.
Drawing of a pair of bosses on a shaper vise from a 1940s war effort shaper user's manual. I like the idea of these bosses to limit movement and the wear shown in my vise picture on the left.

You'll notice that in the picture above, the Acme screw has not been removed from the vise body. There is what I assume to be a tapered pin that holds the bearing collar on the screw. It appears that at some point in the vise's life that someone has tried to get this apart before. Both ends of the pin have been mushroomed above and below the surface of the collar. I removed what I could of the mushroomed heads with a file, then used a small drift to see if I could get the pin to move. I wasn't able to get it to move at all. Rather than take chance to doing more damage and having to drill out the pin and ream the hole, I looked for another way to remove the movable jaw. I found that by removing both jaw faces, then screwing the movable jaw all the way toward the fixed jaw, the movable jaw disconnects from the screw. If you then tilt the movable jaw toward the screw, it will lift out. Since I can lube the screw bearing without removing it, I will just leave the screw in place as I scrape and machine the vise bottom and rails.

After another night's work, I had scraped the swivel base to around 20 points per inch. I also did a little more work with a fine file and stone on the ring that supports the vise. I now measure about 0.00025" difference between the high and low spots on the ring. My filing has left the ring a little less parallel with the base bottom surface than it should be, but I will now put it aside until I get the vise bottom scraped flat. I will then use the vise bottom as the scraping template to allow me to scrape the swivel base ring to fit the vise bottom. Scraping the ring will allow me to bring it back into full contact and parallel with the base.

I have been giving more thought to adding bosses to the screw end of the vise. I am now thinking that once the vise rails have been machined to be parallel with the swivel base, I can place the whole refurbished vise on the surface plate and use gauge blocks to determine the thickness of the bosses. There is enough thickness to the vise body to allow me to drill and tap some blind holes to the bottom of the vise to mount the boss. I am thinking that one piece of cast iron to straddle the bottom of both rails with the center relieved so that there are two feet to contact the shaper table. This will give me three points of contact - swivel base and two bosses/feet - which should be about as stable as any design I can think of.

The next step is to finish scraping the vise base, then to decide how I am going to machine the rails of the vise. Since the fixed jaw of the shaper vise is not removable like a Kurt style milling vise, it is a little trickier to machine the rails. The vise rails are nine inches long and my 24" surface grinder only has six and a half inches of cross travel. Due to the fixed jaw, I cannot grind the rails on the long axis of the grinder. I suppose I could grind it in two setups along the shorter axis of the grinder where the grinding wheel would be able to cut along the edge where the rails meet the fixed jaw, but having to do two setups is less than optimal. I could also grind the rails, then finish them by scraping or I could mill the vise rails and finish them by scraping. There is also the possibility of planing the rails on the shaper. That may be the best way to go, but I will need to do some work to get the table level so that I don't introduce more error into the vise. I'll have to think about this a bit more before I tackle machining the vise.

That's it for now. I am looking forward to getting the vise sorted out and getting to know this shaper a little better.

Shaper 1
Shaper 2
Shaper 3
Shaper 4
Shaper 5
Shaper 6
Shaper 7
Shaper 8
Shaper 9

© Fager July 23, 2015