Where Does All the Precision Come From?

[dusty]

I am starting this new thread in an attempt to redirect the discussion that was started in "Truing, As Best You Can, Shopsmith Table Tops". A question was asked, maybe not directly but asked, just how much precision can one expect to get when using a Shopsmith. I contend that you can achieve whatever level of precision you really need BUT you must know where the sought after precision comes from and why you lose that level of precision when you do.

Fully understanding these many and varied factors gives you the knowledge needed to decide whether you want to expend the effort or not. Be forewarned, it does not come without effort but at the same time once achieved it is almost effortless to maintain.

Before doing any adjustments, I am going to ramble on a bit about what I believe is the heart of the Shopsmith and why precision is possible. Everything, well almost everything, that contributes to the Shopsmith being the wood working tool that it is is embedded in the Headstock. It doesn't matter which Headstock you might want to talk about.

The relationship of the blade to the table is what we are talking about when we discuss "Shopsmith Alignment". Unlike nearly all other bench saws on the market, we reposition the table to achieve that relationship. The blade, by default, becomes the reference surface against which all other adjustments ultimately are made. The blade attaches the the main drive shaft using an arbor. There is an inherent angular relation established at the mating. The relationship between first the blade and the arbor and then between the arbor and the main drive shaft. Once established, that angular relationship does not change. It is 90°, the magic number; the angular relationship between the main drive shaft and the plane of the blade.

The drive shaft is part of the Headstock and is not adjustable. With the exception of maintaining bearings there is little or nothing you can do to help this relationship. But there is one more item I want to mention because we are going to come back to it before this discussion is over. That item(s) is the Way Tube(s). The Headstock mounts onto the Way Tubes and in so doing establishes another angular relationship. The angular relationship of the Way Tubes to the plane of the blade to the axis of rotation of drive shaft. When the Headstock Lock and the Quill Lock are locked, that relationship is what it is. There are no adjustments.

Add a Carriage,

Add a Tie Bar and Tube Assembly,

and then a Table and maybe we can beginning thinking about alignments.

Hmmm, We never, ever talk about the Carriage or the Tie Bar and Tube Assembly being adjusted so I guess maybe they don't even play into this discussion. They are just there, sorta like the four legs and the shop floor.

[JPG]

A word about the headstock, carriage and way tubes.

The wedge locks do more than just clamp in position along the length of the way tubes.

The headstock wedges also lift up the headstock thereby registering the headstock bores against the bottom of the waytubes.

The table carriage wedges pull the carriage down so the carriage bores are registered to the top of the way tubes.

Opposite, but effectively equal as long as they are both clamped.

This is why it is imperative to always lock prior to any measuring.

Both the headstock and the carriage are clamped at their 'right' end, but do have bores at the left end. The way tubes are a close fit in the bores, so any deviation due to the unclamped end being 'free to move' will be quite small, but not 0.

I think the headstock clamp being above the way tubes is a design compromise due to difficulty in locating it below. Since working load will be downward in horizontal mode, it seems it would be better if the clamps were below.

In vertical mode the working load is primarily along the way tubes and the location becomes moot.

[dusty]

Everything that you say here is dead on but just how much slop do you measure in the Carriage and Headstock bores?

I am also thankful that you made the point that whenever measurements are being taken these specific locks should be secure. If they are not, that alone could contribute to some error. However, it is my opinion, that this slop allows only for vertical movement and that any lateral movement that might occur is non-consequential.

[Gene Howe]

I always lock the carriage before adjusting the fence and/or table. Especially in the drill press mode.

[JPG]

I always lock the carriage before adjusting the fence and/or table. Especially in the drill press mode.

Doing so with the headstock in drill press mode is also recommended!!!


Sorry Dusty, the urge was too much to resist!!!

[sehast]

Here is an alignment procedure that worked for me on my 520. I like to start with the most fundamental things first and work up from there. That starts with the two bench tubes connected to the base on the right side and the headrest on the left. These connections do have play as pointed out by JPG which can result in a twist of the bench tubes or a lack of squareness of the tubes to the base or headrest. A twist is avoided by ensuring the two tubes are in the same plane. I used a digital angle gauge mounted on top of a precision straight edge. My straightedge is within .001” over the full 2 foot length but the digital angle gauge is only +/- .05 degrees which results in no better than +/- .010” per foot accuracy. Understanding those limitations lay the straight edge across the tubes front to back on the right end and zero the angle gauge and then go to the left side to see if the reading is still zero. If adjustment is needed loosen the tube lock bars to bring the tubes into alignment. To measure for squareness mount the extension tables in the base and headrest and then check that the table base tubes are 90 degrees or perpendicular to the bench tubes. If necessary adjust the tube lock bars. Now we have the bench tubes square with the base, headrest and the extension table tubes.

Next step is to ensure the way tubes are parallel to the bench tubes and square to both the extension table base tubes. You could use the same technique that I used for the bench tubes using the straight edge and the angle gauge but I think JPG’s technique is more accurate (http://www.shopsmith.com/ss_forum/viewt ... 56#p193956). That involves chucking a bit on the spindle and checking that it aligns precisely with the center of the extension table mounting base on the right side. Then switch the chuck and bit to the upper shaft on the left side of the headstock to confirm that the bit aligns with the center of the extension table base on the left side. Careful adjustment of the set screw in the headrest and the lock screw should bring the way tubes into alignment. Next double check with the angle gauge that the way tubes are parallel to the bench tubes and perpendicular to the extension table tubes and the main table tubes. Also mount a saw blade on the headstock and check that it is perpendicular to the way tubes using the angle gauge.

Now to move on to the tables. First I checked for table flatness by using my straight edge and an assortment of stainless steel shims .001 thru .010”. I confirmed Dusty’s observation that the Shopsmith tables are just not very flat. My main table had up to .01 to .015” variation and one extension table and the floating tables were even worse. The only exception was my right side extension table which was within an acceptable .005”. Short of replacing or grinding them flat there is nothing that can be done here but maybe just knowing where the low and high spots are will help with workarounds. Next I aligned the miter slots in the main table parallel to the saw blade using the standard technique describe in several previous posts. Also then set the main table perpendicular to the saw blade using the angle gauge. For table height alignment I like Dusty’s technique. First the 5’ connect tubes need to be inspected to find the straightest/flattest part of the tube. I used a 55” aluminum saw guide I have for breaking down plywood sheets. I just butt the tube up against the guide and slowly turn it to find and mark the straightest part the tubes. Then make sure the marks align with the top of the table rails. In my case this is a very necessary step because my 5’ connect tubes are only straight over a small sector of the tube. Next I went out to the local box store and found a flat piece of 4’ X 2’ x ¾” MDF to place on top of my work bench. If you have a nice flat bench you can skip this step. Then I placed each table upside down on the MDF side by side just like they will be mounted on the Shopsmith. Insert the 5’ tubes and loosen all the table rails. Then tighten the rail thumb screws to secure the connect tubes. I got a couple of concrete blocks to weight down the tables so they fully contacted the sheet of MDF and then tightened all the table rails to the tables. Re-install the tables on the Shopsmith and loose the bolts holding the extension tables to their bases. Install the 5’ connector tubes and tighten the thumb screws. Next I aligned the fence to the miter slots on the main table using a dial indicator. I was able to get it within .001” but .005” would probably be good enough. Then I mounted the fence on the extension table to align it to the miter slots on the main table using the dial indicator. Tighten the bolts holding the table to the base once alignment is achieved. After several tries I was able to get it to within a repeatable .005”. I swapped the two extension tables left to right and guess what, they aligned.

What I learned. There is too much flex and inaccuracies in the 520 table system to totally rely on them to pull the tables into alignment. Like building a house, you start with a sound foundation.

I am sure this is not a optimum solution and I have undoubtedly made some errors so please comment to help us find a better way.

[dusty]

Very good presentation of an excellent procedure. Yes, you started with a strong and reliable foundation and followed a well structured procedure. I believe that what you have done could be done with any Shopsmith out there that has not be somehow mangled. Obviously, straight tubes and rails are needed.

Thank you, not only for your detailed explanation of how you aligned your machine but also for taking time to share it with the other members of the forum.

[dusty]

A word about the headstock, carriage and way tubes.

The wedge locks do more than just clamp in position along the length of the way tubes.

The headstock wedges also lift up the headstock thereby registering the headstock bores against the bottom of the waytubes.

The table carriage wedges pull the carriage down so the carriage bores are registered to the top of the way tubes.

Opposite, but effectively equal as long as they are both clamped.

This is why it is imperative to always lock prior to any measuring.

Both the headstock and the carriage are clamped at their 'right' end, but do have bores at the left end. The way tubes are a close fit in the bores, so any deviation due to the unclamped end being 'free to move' will be quite small, but not 0.

I think the headstock clamp being above the way tubes is a design compromise due to difficulty in locating it below. Since working load will be downward in horizontal mode, it seems it would be better if the clamps were below.

In vertical mode the working load is primarily along the way tubes and the location becomes moot.

A more appropriate title for this thread might have been "Where Does All the Slop Come From"?

I was a bit surprised when I finally got around to measuring the fit of the Way Tubes into the Carriage bores. Now because the Carriage and the Headstock are expected to slide on the Tubes, the fit cannot be snug. Don't worry! It isn't!

I can slide a .005" to .006" feeler gauge in between the tubes and the bores when the locks are loose.

[sehast]

I have always been concerned because when I lock the headstock I can see visible movement of a bit I have chucked. However, when I make runout measurements whether it be a bit or a saw blade I don't see any significant difference from the unlocked measurement.

[dusty]

I have always been concerned because when I lock the headstock I can see visible movement of a bit I have chucked. However, when I make runout measurements whether it be a bit or a saw blade I don't see any significant difference from the unlocked measurement.

I have yet to verify how much of the movement is translated into the actual cut but there has to be some. However, if you do all of your alignments and then project setups while making your measurements with the locks secure you should witness "0" relocation.

Besides, in how many of your projects (wood working that is) do a couple of thousandths make a difference.