Shopsmith Forums

In post # 9 I attempted to explain that my jointer fence had a slight twist.

One morning at the Senior Center BS/coffee session I was explaining the jointer fence problem. One of the group was a printing press maintenance mechanic and explained that he could probably fix the twist. With nothing to lose I hauled the fence over to his shop, making sure I arrived before afternoon cocktails. After placing the fence on a marble slab in his shop to determine the amount of twist he clamped one end of the fence in a vice and used a large 2 or 3 foot adjustable wrench on the other end and gave it a little tweak. Checked it on the slab. It was about as close to perfect as one can get. A .0005 feeler gage would fit under one corner but there was no rocking. As a disclaimer I do not endorse this method of straighting cast. Just stating what was done to my fence.

[quote="BuckeyeDennis"]The tables on my jointer are flat to within my ability to measure them. The fence is the problem. I first noticed that it had a twist when attempting to set it to 90 degrees with a machinist square. When set square back at the quadrant end of the infeed table, it was not square at the cutter-head end of the infeed table. And vice versa.

Against a twisted fence, a flat board can rock. I did, however, discover that I got more consistent results when using the guard in feather-board mode. Presumably because it gave a more consistent center of pressure against the fence.

I happened to be jointing 3/4" stock, but the jointed board edges will reflect whatever angular errors result from fence problems, irrespective of board thickness. Over a 12" wide panel glue-up, I found that these small angular errors wanted to cause significant panel cupping.

So I intend to make the fence flatness problem go away. I just don't need the aggravation! ]
My bad! I missed the key word(fence). I also placed my understanding on the concavity you described.

So we begin anew with different understanding.

A twisted reference surface is useless. So the only thing(other than bending it:eek:) is to attach a shell that replaces the reference surface with a flat surface.

The fence is quite short(height) to begin with.

It is unclear to me the extent of the 'twist', so I do not know if resurfacing it is practical. 0.006 concavity should be fairly easy, but that does not describe the 'twist' at all.

As anal as I can be about alignments, I fail to see or understand the negative impact of a .006" error in in jointer fence or table alignment.

I wish all of my joints were off by .006". Yes, I know that it is really a pleasure to join up a number of pieces and not be able to see the joints without close examination. It just doesn't happen to me all the often.

I have not checked my jointer that closely. Maybe I will. Maybe. But if I do and I find it to be off that much or more, I certainly am not going to quit using it or even "fixing it" after all these years.

Good luck with whatever you decide to do.

Stumpbuster wrote:In post # 9 I attempted to explain that my jointer fence had a slight twist.

One morning at the Senior Center BS/coffee session I was explaining the jointer fence problem. One of the group was a printing press maintenance mechanic and explained that he could probably fix the twist. With nothing to lose I hauled the fence over to his shop, making sure I arrived before afternoon cocktails. After placing the fence on a marble slab in his shop to determine the amount of twist he clamped one end of the fence in a vice and used a large 2 or 3 foot adjustable wrench on the other end and gave it a little tweak. Checked it on the slab. It was about as close to perfect as one can get. A .0005 feeler gage would fit under one corner but there was no rocking. As a disclaimer I do not endorse this method of straighting cast. Just stating what was done to my fence.

If I may paraphrase, back in post #9, you believed that the slight twist in your jointer fence was the root cause of some poor joints, and so you used the conical disc sander to correct the jointed edges (with beautiful results, judging from the photos).

What difference do you see now that your fence has been trued up?

JPG40504 wrote:My bad! I missed the key word(fence). I also placed my understanding on the concavity you described.

So we begin anew with different understanding.

A twisted reference surface is useless. So the only thing(other than bending it:eek:) is to attach a shell that replaces the reference surface with a flat surface.

The fence is quite short(height) to begin with.

It is unclear to me the extent of the 'twist', so I do not know if resurfacing it is practical. 0.006 concavity should be fairly easy, but that does not describe the 'twist' at all.

At the time, I didn't think that I had the means to measure the twist independently of the concavity. But after pondering your post, it struck me that the machinist square and the feeler gauges are all I need. Duh!

And even my small surface plate ($25 on sale at Woodcraft) is large enough to check the infeed table for flatness, should there be any doubt.

Another bad on my part. I responded after reading today's initial post without reading(remembering) all that preceded it in this thread. Doing so would have eliminated the first 'bad'.

I gotta side with Dusty in that perhaps you are expecting perfection with the joints of wood which is not an absolutely accurate medium to begin with.

[quote="JPG40504"]Another bad on my part. I responded after reading today's initial post without reading(remembering) all that preceded it in this thread. Doing so would have eliminated the first 'bad'.

I gotta side with Dusty in that perhaps you are expecting perfection with the joints of wood which is not an absolutely accurate medium to begin with.]

I'm not actually expecting perfection. Although I do confess to being heavily influenced by the machine-tool world, where I have spent the last 15 years of my career. "Precision" in the machine tool industry these days typically means tolerances of about 0.001" or better. With my Shopsmith equipment, I find that I can realistically shoot for about 0.005" .. i.e good enough that the error doesn't matter in wood. And the wood is much prettier than the metal.

A concept from metalworking that I believe applies equally well to woodworking is that of "machine capability". Let's say that you need to make a part with a tolerance of 0.005". If your machine has a repeatability of 0.003", you can probably make acceptable parts. BUT, the machine must be set up nigh unto perfection, and if the material varies, the tooling wears a bit, or maybe even the temperature changes, you are out of tolerance. You CAN make good parts, but a skilled operator has to be monitoring the process constantly, and tweaking the adjustments as required. Thus, strictly speaking, the machine is not "capable" of holding that tolerance all by itself. In the industry, the technical term for this is "expensive" . And your competitor who invested in a machine that repeats to 0.0005", but can make good parts all night long, with no one attending it, is going to eat your lunch.

For me, woodworking is a hobby, and so the economics are irrelevant. But I simply hate having to rework and correct parts. At least the third time around. (The first two times are fun and educational opportunities to diagnose the problem and develop a process fix. ). Hence, if I find something about my equipment that can contribute more than about 0.002" of error to my workpiece, I will try to correct it. Why? Because it will make it much easier, and thus more enjoyable, to machine the finished wood part to 0.005". Or to 1/64", for that matter, if that is your personal standard.

None of which should be construed as advice for any other recreational woodworker. Heck, I love to do freehand turning, where the only criteria is how well the shape pleases your eye! For me, I simply don't want mechanical fit details to detract from the pure aesthetics of woodworking. I'm slowly learning how traditional woodworking craftsmanship techniques can address these same issues. But in the short term, I inevitably build upon what I already know.

The 'perfect' joints that I have seen were the product of highly skilled individuals who knew how to make them with the imperfect materials at hand.

They I am sure have in the past made many less perfect joints.

So do not expect to short circuit that path of experience.



We share the engineer's disease. We always consider less than perfection correctable with greater understanding of what caused said imperfections.

Sometimes those causes are beyond our control and we be frustrated by that.

Worry about those things we can control.



Hand made objects need imperfection to to distinguish them from a robot's creation.

JPG40504 wrote:The 'perfect' joints that I have seen were the product of highly skilled individuals who knew how to make them with the imperfect materials at hand.

They I am sure have in the past made many less perfect joints.

So do not expect to short circuit that path of experience.]can[/B] control.



Hand made objects need imperfection to to distinguish them from a robot's creation.

Absolutely! Therein lies much of the artistry to which I aspire.

Over time, I have compiled a small folder of personally inspiring woodworking photos. A common theme is an artistically highlighted knot, or a well-crafted butterfly joint holding a natural split in check. But in none of these remarkable pieces will you find a misfit joint that distracts the eye from the artistry.

BuckeyeDennis wrote:If I may paraphrase, back in post #9, you believed that the slight twist in your jointer fence was the root cause of some poor joints, and so you used the conical disc sander to correct the jointed edges (with beautiful results, judging from the photos).

What difference do you see now that your fence has been trued up?

I must confess I haven't used the jointer for glue ups post fence repair.

We are on our northern migration and will not get back to my SS until fall.

PS
It was 102 deg. in Boulder City NV when we left we are now in Oregon and they are predicting lows in the 30s. There is something wrong with this picture.