Shopsmith Forums

thunderbirdbat wrote:

JPG wrote:

thunderbirdbat wrote:The main table is adjusted perpendicular to the saw blade/sanding disk. The aux tables are adjusted to the main table. Does this mean the the saw blade/sanding disk is not perpendicular to the way tubes/floor? If this is the case then the slight angle would not allow the aux table to be swapped as the angle would be opposite at the other end.

The floor is irrelevant. The blade/disk face in a parallel plane to the spt mount bores(center axis) is what matters. Parallel means both left right and vertical. If so. the main table adjusted to the blade/disk should also align with the both aux tables. All that is left to adjust on both aux tables is the front/rear tilt to match the main table(horizontal plane of all three table top surfaces).

I suggested the exclusion of the main table(at this point in the process) so as to eliminate it as an intermediate reference.

The current goal is to determine why swapping is not successful, so concentrating only on the tables/posts/mounts should help reveal the reason why.

How do you know that the mount bores are parallel with the blade/disk? A few thousands out of parallel will be noticeable over the length of the SS. When aligning the aux table you also adjust the side to side height with a straight edge to be level with the main table. In my adjustment procedure step 81 and 82 adjusts the side to side level. "Hold a straightedge on the infeed (81)/outfeed (82) top surface of both the worktable and extension table, as shown in Fig. B-102 (81)/B-103 (82). AS needed, adjust the top nuts located on the infeed (81)/outfeed (82) side of the extension table, in order for the straightedge to be level across both the worktable and the extension table infeed (81)/outfeed (82) sides." If the bores were parallel to the blade/disk this would not be required. Also what is the manufacturing tolerance for these parts? The tolerances could also account for the requirement especially when multiple parts are combined.

I agree with everything you have said and there was a time that I believed ALL of this pain was the result of the Main Table not being flat and/or not be level (ie in the plane at 90° to the blade). I am no longer in that camp. I have chased this far too many times and have used several different methods to verify the table position.

Right now I am assembling a test bed that will set upon my work bench; a surface whose flatness I am satisfied is good enough for this project. The test bed shall consist of two Base Assemblies tied together with the tubes that were cutoff when I created the Shorty. After verifying the geometry of this test bed, I will use it to align two extension tables and to then swap them. The expectation is that the swap will be unsuccessful. This is an attitude, I know but everything that I just described I have done several times with essentially the same failed results.

OK You have seen the same thing that almost all of us have seen. Your premise in the other thread was that because both end castings with the double tilt are the same it will make a difference.

Here are a few things I have thought about. You want to take the legs off an put the upper assembly on your known to be flat bench. The bottom of the bench end castings are not machined to match anything so that is an unknown reference surface.

The bench legs are stamped steel. Hardly a precise part.

The holes that bolt the legs to the end castings are probably not as precise as we are talking about and they are oversized for easy assembly.

When you add all of this and probably more it doesn't work no matter how hard you try.

I encourage you to keep trying because I absolutely have the same attitude as you about trying to find the answer if there is one.

However, in this case I just have two extension tables an have moved on.

jsburger wrote:OK You have seen the same thing that almost all of us have seen. Your premise in the other thread was that because both end castings with the double tilt are the same it will make a difference.

Here are a few things I have thought about. You want to take the legs off an put the upper assembly on your known to be flat bench. The bottom of the bench end castings are not machined to match anything so that is an unknown reference surface.

The bench legs are stamped steel. Hardly a precise part.

The holes that bolt the legs to the end castings are probably not as precise as we are talking about and they are oversized for easy assembly.

When you add all of this and probably more it doesn't work no matter how hard you try.

I encourage you to keep trying because I absolutely have the same attitude as you about trying to find the answer if there is one.

However, in this case I just have two extension tables an have moved on.

Those are two points I hadn't thought of and which seem persuasive to me, unfortunately.

Sorry to come in late on this one. I think that there are so many variable in this alignment situation you may never mitigate all of them in the end. The floor inconsistencies combined with the SS inconsistencies mean that every time you move the machine you will introduce a variable. The only possible solution which most of us have probably done is to label the ext tables so they stay consistent to the particular SS and also left and right. Then I have just adjusted the ext tables to be in plane with the main table. Even then they take periodic adjustment. My garage floor was laid by a diabolical soul and is about as flat as a tank obstacle course. At one point I put marks on the concrete so I could return to the same spot which actually did help.
Paul

Interesting that the base/leg screws have been mentioned.

The original Mark 5 bases used counter sink head screws that likely provided more consistent alignment of the base to the leg.

There are too may variables here. Step one should be to ensure the bench and way tubes are totally parallel in all planes. Make sure the legs are securely fastened to the bases and loosen all the tubes. Take off all the tables and mount the lathe tailstock on the right side. Adjust until the drive center on the headstock is precisely aligned with the cup center on the tailstock. Note the height of the tailstock and remount it at the same height on the left side. Put the drive center on the SPT spindle and check for alignment with cup center on the tailstock. With all the tubes loose, use shims between the legs and the floor to align the drive center and the cup center. Then tighten the tubes and re-check on the right side for alignment. It might take a few iterations but you should be able to get the centers to align on both ends of the Shopmith. When you do, you will know that the tubes are parallel in all planes with a fair amount of precision and then you can move on to the tables alignment.

This is an interesting approach - using the drive centers. I gotta think on this.

Shown here with a non-double tilt. http://www.shopsmith.com/ss_forum/viewt ... 70#p124670

What I like about the procedure as I have described it, is its simplicity. No measuring tools are needed nor does it rely on subjective or difficult to repeat actions by the operator. You simply line up the points of the drive center and cup center which I argue will be within a few thousandths of an inch. You would have to have some fancy laser measurement equipment to do as good.

The problem is when you take the shims out from under the legs and start rolling it around on an uneven floor you run the risk of knocking it out of alignment. Its a lot to ask of the bench tube clamps to withstand repeated drops from a 1 inch height down onto an uneven floor and still maintain alignment.

OK I took an hour or so to play with several things on my machine today.

First the lathe center idea. On mine the height is right on but the left to right is off by 1/16". The problem I see there is that the tail stock has to flip 180 degrees when mounted on the other side, so what was the front is now the back etc. This leads me to think the error is somewhere around half of what you see. Since the method of adjusting the location is in a circular pattern this becomes even more involved as it will change things in two dimensions......

Next I put previously adjusted right-left extension table on and checked it out in terms of sliding the bars only and they still do that fine even after a couple of years. I did not attempt to do a rip fence alignment test.

I've never had the issue of tables not being the same height front to back, my issue had always been getting the rip fence alignment so perhaps that is now out of wack but it was more effort to check then I had the time to do.

Ed