Wadkin Resurrection: Protractor

I’m back. Yes, this blog lives. Over the past 2 months I’ve been through some tough challenges I do not care to relate in public, and am just returning to ‘normal’ – whatever that is – so hoping to be back to normal operation, both in the shop and blog-wise, over the upcoming few weeks.

—————–

From the previous post in this thread, you may remember that I had the sliding table on the saw re-planed flat, and along with that the front miter gage was also planed to make it geometrically correct, and was fitted with a new pivot bolt. Additionally, the worn detent holes in the table were plugged and re-drilled with a tapered cutter, and a second position for the mitre gage was made in the mid-length point of the sliding table.

As to why I would take the time and trouble to do this? Well, I like my stuff to work properly for starters, and I notice when it doesn’t. A bowed sliding table means that rip cuts with stock which rest partly upon the sliding table side (i.e., most rip cuts) end up riding up on the bowed of the sliding table, taking the cut out of 90˚ or whatever else angle was desired. Some might tolerate the disfunction, but I could not. Though I was considering, and still do consider, buying another saw altogether, I felt that at the very least I should try to put the saw right and see how I liked it – – and if I still want something else after that then I can sell it in good nick, as some Brits might say, leaving me in better repair than I found it, and, not unimportantly, definitely salable for more than it would be otherwise.

Anyhow, part of that repair work, as mentioned previously, was getting the front miter fence corrected geometrically and positionally, so as to do its job as intended. As the term ‘front miter fence’ suggests, there was also a ‘back miter fence’ available for the Wadkin PP dimension saw, however it was optional and seems to have become rarely optioned, so it is now a quite rate accessory. You can see it in Fig. 6 (bottom left) in the brochure for the machine shown below:

The back miter fence is employed in tandem with the front miter fence to effect cuts for miter joints in polygons of any number of sides. This type of double miter fence was seen on other saws of this type, but seems to be a rare accessory regardless of make.

Later Wadkin saws had what i might call an amalgamated type of front/back fence:

This would work, but only for frames which are square (i.e. 4-sided frames), so it is a more limited sort of device. Forget it as far as other polygons are concerned – – well, at least in terms of using the device in one set position, and not having to shift over for complementary cuts.

One factor playing into that is that the casting for the Wadkin back miter fence was not robust enough to withstand drops to the floor, and breaking of the guide slot was not an uncommon result. So, of the back miter fences which were supplied, there has been a certain attrition over time, and of course it is not uncommon when machines change hands that parts, manuals, and accessories get lost in the shuffle. I’ve tried finding an original manual for a Wadkin PP saw and had no luck so far, for instance. Where’d they all go?

The pairing of front miter and back miter, when in the 90˚ relationship, produce of course a quarter-circle, and thus the N. American term for the pairing of front/back miter fences is “quadrant”. For some strange reason, in recent years this term is now being applied to just the more commonly-found front miter fence alone, though it then becomes a bit nonsensical a term in my view, as a front miter alone does not a quarter circle describe.

Notice in Figure 7 above another type of fence, termed by some a protractor, what is otherwise termed in N. America at least, a ‘miter gage’. The brochure describes it as:

“For work involving only a small movement into the saw, fast production can be obtained with this sliding fence supplied to order. Tables are tee-slotted both sides of the saw to receive this fence.”

Like the back miter fence, the protractor is a rare accessory. One fellow in England who frequents a Canadian Woodworking forum dedicated to vintage power tools, mentions have searched in vain for 6 years for the protractor, to no avail. And he lives in the UK, where Wadkin stuff in general is going to be a lot more abundant than it is in N. America.

Fortunately, some other members of that forum have gone ahead and cast reproductions of both rare accessory fences, and I now have obtained each, in rough-cast form.

The front/back miter fence set greatly facilitates cutting miters, so it is obvious as to why it would be nice to have that set. The reason for having a protractor however is less obvious, however if you have a saw which tilts only one way (like most, but not all saws), then angled cuts can then be taken on both sides of a blade. Otherwise the single tilt capacity is a limitation. So, I’m planning to use the protractor in the t-slot which is in the main table of my saw, to the right of the saw blade, and continue using the front miter fence on the left side of the blade, on the sliding table.

Another advantage to having a sliding fence on the right of the blade is that the two fences on each side of the blade can be linked to one another with a connector to create a sled-like cutting arrangement. Anyone who uses a table saw knows the advantages of a cutting sled.

I noticed that some American-made dimensioning saws in the distant past (early 1900s~1920s) came with such capacities, and there were even protractors for both sides, one with an angled portion removed so as to clear the blade, and a horseshoe-shaped cast connector:

The above photo comes from a web page devoted to the topic of Oliver miter gauges. The bridge between the two fences in interesting. Here’s a close up picture of one that is for sale currently:

However I’m not sure what advantage it provides relative to screwing a suitably clean flat and square wooden fence right across both gauge heads.

After speaking with my machinist friend in New Hampshire, the same fellow who planed the sliding table for me, we determined that in order to keep costs down for the protractor work I should tackle the rough milling of the castings. Besides the cost savings (which are more or less fictional if I consider that my own shop time is being used), I wanted to have a try at milling cast iron, as it was not something with which I had any experience.

Rough castings I received are made oversize by around 10% and with draft, which is a taper to the pattern to enable the pattern to be easily removed prior to the molten iron being poured. This left me, especially with the protractor casting with its semicircular beveled protractor edge, with a tricky piece to machine, in that there were no easy ways to fixture it to the table.

The base of the rough casting was 0.9″ thick or so, which, given that the finish dimension was to be 0.75″, left me with ample material to play with. I initially fixed the bottom of the casting down onto a pair of blocks and took a light skim on the upper surface, only to come face to face with the fixturing challenge. The hold downs that I used to hold the casting to the table were in the way of the cutter, so had to be moved around from time to time and this lead to a imperfect surface.

Take 2, then.

I got a piece of MDF, and after playing around with positioning the casting a bit I used some PL300 construction adhesive to affix the base of the casting to the MDF:

Casting now in place, and I let it set up for three days as things turned out:

The casting for this miter head is massive by any standard, certainly multiples greater in size than the typical miter gauge you see on a cabinet saw.

This worked very well, and I was able to deck the upper surface without issue:

I gave the top edge of the fence a little attention while I was at it:

If you look closely at the above photo, you can see that the rightmost portion of the casting was not quite cut as it was a hair lower than the rest. I was striving to keep the height of the fence portion of the casting to 3″, and did not wish to take any more off the upper surface until I had take some time to consider the best way to proceed.

At this point the fence portion was sitting at around 3.06″ in height, and I needed to skim the bottom of the casting no more than 0.06″ to stay at or above target dimension. I had to scratch my head for a while to figure out the best way to fix the part in place upside down to re-mill the base surface

Eventually it occurred to me that I could hang it off the table to one side and swivel the milling head around and adjust the ram projection to put the cutter where it needed to be – like this:

I was still wrestling with the issue of holding the piece in place, and knowing that I would have to move the hold-downs from time to time, I put the hold downs together out of separate pieces (T-nut, double-threaded stud, washer and locknut) This enabled one hold down to be removed at a time, from the top, instead of having to lift the piece off of a conventional hold-down, in which the stud and t-nut are one-piece.

With the face of the casting which lay against the table now being a flat plane, it was no problem to shift the hold-downs around as required, occasionally clamping with additional hold-downs from the perimeter of the casting.

The milling, so as to take advantage of the mill’s power feed in ‘x’, was done along the short dimension:

In case you are wondering, I’m using a 1/2″ shank router bit from CMT to do the cutting.  After this work, I will retire it. Cast iron is not normally something one would tackle with a router as the speed is way too high and the cutting will be unsatisfactory, not to mention risky if the cutter grabs the material in freehand use of the tool. It’s still possible to do it, but great care must be taken and light passes. For cast iron, the rpm of the cutter needs to be low, along with the feed speed ideally.

The router cutters worked perfectly, though they were not ideal otherwise.

After decking the surface, the resulting cut on the bottom of the gauge was clean and flat and evidenced no rocking when the part was placed on a granite reference plate. I was happy with the way it came out. I was also left with the fence height of 3.025″, which was perfect, leaving my machinist a little bit of work taking a final pass or so to reach the finish dimension of 3.000″

With that, I flipped the casting over once again, measured the remaining thickness of the base (0.88″), and got to work cleaning it to size. Target thickness is 0.75″ ultimately, but I finished out at 0.80″. With a cutter like this, I had to be taking lighter cuts of around 1/32″ to work cleanly and without chatter or risk of tool breakage.

“One day I’ll have more tooling” is ever my mantra. Maybe you’ll make it yours one day?

With the base of the casting flattened and decked, I turned my attention to the front face of the fence, which was strongly tapered. I happened to remember that a steel fabricator near my shop had mentioned having a 90˚ angle plate for milling, and he had offered to let me use it if need be. So give there was now a need, I popped over there and borrowed the angle plate, which happened to be just the right size upon which to mount the protractor:

A view from the rear:

Once I had the piece suitably leveled out, I could commence milling the draft off the casting:

The casting had a lot of draft as you can see in the view of the end of the casting, so I spent at least 8 passes back and forth alone on that portion of the fence front.

Starting to move out towards the upper edge a bit:

About halfway along:

Nearly there:

Two more passes left a clean surface.

As it turned out, the angle plate was not actually square. Once I had decked off the surface I discovered this unpleasant fact upon checking the result with a try square. Never assume something is square, always check – my bad.

There was no heavy price to pay in this case however as I simply shimmed under the plate until it was 90˚ to the table, re-leveled the piece using an indicator, and then took a couple of light final passes to get everything dialed in to where it should be.

I had a pencil line marks out on the backside of the casting to give a rough reference line as to how much material was to be removed off the front face of the fence:

All done as far as the work I am tackling:

I put a coat of Boeshield T-9 on it as it will likely be a while until it is completed and, well, I don’t care much for the rusted look:

The casting is slightly oversize in all dimensions, to leave the machinist something to work on for final refinements.

Note that there will be a part II to this thread, at least, as this protractor is also intended to have a digital readout.

That foray into cast iron milling was rewarding. I like learning about new materials and gaining experience working with them. I feel like I got a better finish than a lot of places I have had machine cast iron for me in the past. It’s hard to find a good machine shop it seems.

While the milling was underway, back and forth, back and forth, I did some looking at the situation with the steel bar which I will need to obtain and configure of the protractor head. Checking the Wadkin out, bad news was all I found however.

I was nosy-ing around, and thought I’d inspect the existing t-slot and infill bars on the main table. The main table has a dovetail slot running crosswise, into which mounts the rip fence, and so the infill bar is composed of two sections which terminate at the dovetail slot. The slot is not quite mid-way up the table, so the nearer infill bar is significantly shorter than the one on the far side of the dovetail slot.

Pulling the nearside bar out I discovered something interesting when checking it for thickness. At the point where it meets the dovetail slot, the number is 0.365″ or so:

In the middle of the bar the thickness is less:

And at the end of the bar which meets the table edge, the number is less yet again:

Not what you want to find, eh? The taper on the bar amounts of some 0.015″, which is a lot. Checking the longer infill bar on the far side of the dovetail slot, the situation was the same, being thickest at the dovetail slot end and tapering a bit more than one hundredth of an inch over the length.

Reflecting upon this situation, given that the infill bars are flush to the table surface when they are mounted, that the slot depth also deepens towards the dovetail groove, and that I also had a similar issue with the rebate for the wooden table edge on the main table (which needed tape to shim it in the middle where it was a bit low,) I could work out what had gone on during manufacture. It seems that after they made the table casting, they machine the table for the dovetail groove, the rebated surface, and the t-slot. Then the casting must have sat a while, or had sufficient stresses that it ended up cupping across the width. Then the infill bars were screwed down and the entire surface was machined flat (it remains flat), and out the door the saw went. Since the protractor was not a part of the original spec of the machine, nobody would have noticed anything save for the imperfect fit of the wooden table lip (which I suspect many operators would not notice in the first place). I’d noticed that table lip issue but some tape underneath the wood supplied adequate correction.

I would speculate, casting design-wise, that the crosswise dovetail slot on the upper surface may be the largest contributor to the casting not being stable and cupping down in the middle around that slot. It’s one thing that the casting cupped and as far as the table surface goes, the secondary machining with infill bars in place made everything nice, but they should have gone over the t-slot and rebate ledge again to make them flat too. That they didn’t re-machine indicates a less-than-thorough approach. They’re not – or weren’t rather – Hofmann I guess.

Based on some communication I have had with someone who owns an original Wadkin protractor, from measurements he found that the bar is meant to be 1/100″ proud of the table surface. This is what I thought likely, as it would give the most drag-free operation of the protractor head.

But with a t-slot which dives down in the middle and then climbs back up again, as I have, the t-slot is not going to work as intended, in terms of having a bar with engages with the t-slot.

That leaves me with two options:

1. Get the table t-slot and rebar re-machined.
2. Use a protractor sliding bar which does not have lips to fit the t-slot

The size of the table casting exceeds what my trusted machinist friend in NH can deal with, so that means sending elsewhere, and that could mean shipping and the uncertainty of working with a shop that may not do quality work.

Option 2 is certainly cheaper and easier, and for the most part the lateral guiding of the bar in the slot is what determines accuracy, BUT, the factory put a t-slot in there for a reason. It costs more to put a t-slot in than a simple dado, so they must have felt it necessary.

The reason that springs to mind is that the t-slot increases safety as the miter gauge can not lift out during a cut, and it would make the miter gauge less likely to fall off the table if left in place with the head overhanging the table lip.

I’m sufficiently persuaded that the t-slot is a good thing that I have elected to get the table t-slot and rebate re-machined flat. I’ll probably take the top to Marina Industries in East Hartford CT, where, though I have had mixed experiences in the past, definitely has the machine capacity to do the work, and it is a simple enough constrained sort of job that it should be problem free. We’ll see….

All for now. In the next post on this topic the back miter fence is tackled.