Band Saw Misbehaving

[algale]

Let us see if I understand what I have just read.

The back of the 5/8" blade bears against parts that are not intended to be tolerating the blade.

Does not this provide an explanation for unexpected results?

The back of the blade should always ride against the rear guide bearing and the lower guide bearing, and against the upper backup bearing ONLY when under cutting load. It should not be contacting anything else.(period)

I probably caused your confusion. The problem still exists and the blade is not riding against anything it shouldn't (and hasn't for a long time). I believe the reason I was not having a problem back when Caleb was (a year ago or so) having a similar issue is because the blade was being restrained from twisting by the grooves accidentally cut in the support brackets.

Currently, the blade is riding only against the autotrack bearing and the lower backup bearing. In fact, quite a lot of force is being exerted against both bearings. Maybe the amount of force is normal. But it seems to me that if too much force is applied (perhaps through excessive rearward cant on the upper wheel?), the rear of the blade has no place to go but sideways and since the groove in the autotrack bearing prevents this, the blade will twist where it goes through the insert as the back of the blade slides along the lower back up bearing.

Currently, the twist at rest is not so severe that I cannot square the table to the blade, but as soon as I try a cut with the miter gauge, the back of the blade seems to slide across the bearing and I get a dramatically non-square cut (this is on flat stock, 1/4 in thick by about 2 inches wide). Increasing tension does not seem to help.

[dusty]

I probably caused your confusion. The problem still exists and the blade is not riding against anything it shouldn't (and hasn't for a long time). I believe the reason I was not having a problem back when Caleb was (a year ago or so) having a similar issue is because the blade was being restrained from twisting by the grooves accidentally cut in the support brackets.

Currently, the blade is riding only against the autotrack bearing and the lower backup bearing. In fact, quite a lot of force is being exerted against both bearings. Maybe the amount of force is normal. But it seems to me that if too much force is applied (perhaps through excessive rearward cant on the upper wheel?), the rear of the blade has no place to go but sideways and since the groove in the autotrack bearing prevents this, the blade will twist where it goes through the insert as the back of the blade slides along the lower back up bearing.

Currently, the twist at rest is not so severe that I cannot square the table to the blade, but as soon as I try a cut with the miter gauge, the back of the blade seems to slide across the bearing and I get a dramatically non-square cut (this is on flat stock, 1/4 in thick by about 2 inches wide). Increasing tension does not seem to help.

Am I correct in saying that the level of force against the lower backup bearing (when not cutting) is created ONLY by the effects of the cant of the upper wheel?

[algale]

Am I correct in saying that the level of force against the lower backup bearing (when not cutting) is created ONLY by the effects of the cant of the upper wheel?

With apologies to the Rolling Stones ("You Can't Always Get What You Want"), with the Shopsmith band saw I do believe You Always Get What You Cant. Too little cant and the blade may walk off the wheels. Too much cant and I believe you can force the blade to ride at an angle. With a thin blade, this may not be noticeable or a problem. With a wide blade, it can be.

This morning I spent about 3 hours mucking about with my band saw. It was only when I took off so much cant off the upper wheel that the autotrack bearing was barely being kissed by the blade that my problem remedied itself. There is still solid contact with the lower back up bearing. So to answer Dusty's question, change in cant does not appear to affect force on the lower back up bearing to the same degree as the autotrack bearing (I find this result surprising; I would have predicted the blade would apply about equal force to both bearings as cant is increased or decreased).

Once I did this, I was able to square the table to the blade and get square cross cuts and adjust the rip fence to get parallel rips (within about 3 thousandths per 18 inches). I also got much better resawing results.

While I may tinker a bit more, I do think the situation is at the point where it is no longer a problem to me, at least until further notice.

[JPG]

I think it would be helpful if we understood what causes the blade to move towards the backup bearings. I do not(yet), but have observed that rotating the wheels in reverse produces the same motion. When the wheels are rotated in the reverse direction, the blade drifts towards the back side of the upper wheel. There being no backup/guide bearing on the front side, the blade will actually walk off the rear side of the upper wheel.

AIUI, both wheels are tapered across their periphery(thicker at the back). This is the reason the cant is required. The lower wheel axis is essentially horizontal. That taper places the flat side of the blade at a slight angle from horizontal at the bottom of the lower wheel. This leads to the slight 'lean' of the blade towards the back.

Now since the upper wheel is also 'tapered'(I assume the same amount), in order for the blade to lay flat against the upper wheel, it must be canted to an angle from horizontal twice the taper angle(the blade rides against the upper side of the upper wheel and the blade rides against the lower side of the lower wheel).

Drift across the face of the wheel I think is caused by unequal tension/angle of the blade to the face. This is where I am struggling. If one side is 'tighter' than the other, which way will the blade drift? I am inclined to think towards the lesser tension side, but am not sure of that.

If that is accurate, than more cant leads to greater drift towards the rear.(increases tension at the front of the blade and less tension at the rear.

But this seems counter intuitive as greater tension at the tooth side of the blade should diminish the wandering of the cutting edge. ??????

I would test this with the belt sander, but it is not 'here'.

It is interesting that CS advised to alter the cant angle(although direction seems nebulous).

Not sure any of this meandering is helpful, but may ignite a spark elsewhere.

[JPG]

Since you posted your success as I was meandering, it appears drift occurs towards the lower tension side.

I be curious how much a straight edge laid against he face of the top wheel clears the nearest(upper) edge of the lower wheel on your now 'corrected cant' bandsaw.

[algale]

Since you posted your success as I was meandering, it appears drift occurs towards the lower tension side.

I be curious how much a straight edge laid against he face of the top wheel clears the nearest(upper) edge of the lower wheel on your now 'corrected cant' bandsaw.

I'll try to get some measurements. Do you want the measurement taken with the blade on and tensioned? Incidentally, per the mother ship, cant is not measured this way. Rather, the official method for determining proprer cant involves mounting and tension a 1/4 inch blade. Cant is correct per the mother ship when the back of the blade is riding 1/8th inch from the back of the upper wheel as measured at the 12 O'Clock position.

As for your prior post, I found the meanderings difficult to follow. To begin with, I'm not sure I agree with your observation that the wheels are tapered. If they are, it is very subtle. I also don't see any possibility of the blade walking off the rear side of the upper wheel (even when spun in reverse) as long as the autotrack bearing is there. When I spun the wheels in reverse, the location of the blade on the wheels was unaffected; but the angle changed dramatically; rear of the blade immediately drifted from right to left resulting in a far more severe angle (albeit in the opposite direction) than when the wheels were spun in the correct direction).

[JPG]

I'll try to get some measurements. Do you want the measurement taken with the blade on and tensioned? Incidentally, per the mother ship, cant is not measured this way. Rather, the official method for determining proprer cant involves mounting and tension a 1/4 inch blade. Cant is correct per the mother ship when the back of the blade is riding 1/8th inch from the back of the upper wheel as measured at the 12 O'Clock position.

As for your prior post, I found the meanderings difficult to follow. To begin with, I'm not sure I agree with your observation that the wheels are tapered. If they are, it is very subtle. I also don't see any possibility of the blade walking off the rear side of the upper wheel (even when spun in reverse) as long as the autotrack bearing is there. When I spun the wheels in reverse, the location of the blade on the wheels was unaffected]rear of the blade immediately drifted from right to left resulting in a far more severe angle (albeit in the opposite direction) than when the wheels were spun in the correct direction[/B]).

Makes sense, but not easiest to determine.

Tensioned, same as the 'official' method.

Take my word for it, the wheels have a larger od at the back than at the 'front'. Yes it is slight!!! They are flat across that slanted face.

That resultant angle is only a degree or two.


The same forces that cause the blade to ride against the autotrack bearing are in effect when rotated in reverse, but there is no guide bearing in the front, so it does walk towards the back. I know inconsistent use of front/back. Front of the bandsaw, back of the wheel.


Now i'se cornfuzed? Left/right from what vantage point?

[algale]

Makes sense, but not easiest to determine.

Tensioned, same as the 'official' method.

Take my word for it, the wheels have a larger od at the back than at the 'front'. Yes it is slight!!! They are flat across that slanted face.

That resultant angle is only a degree or two.


The same forces that cause the blade to ride against the autotrack bearing are in effect when rotated in reverse, but there is no guide bearing in the front, so it does walk towards the back. I know inconsistent use of front/back. Front of the bandsaw, back of the wheel.


Now i'se cornfuzed? Left/right from what vantage point?

I was referring from the operator's perspective/point of view. I still don't know what your front/rear of the wheel are referring to. For me everything relates from the operator's perspective/point of view.

[JPG]

I was referring from the operator's perspective/point of view. I still don't know what your front/rear of the wheel are referring to. For me everything relates from the operator's perspective/point of view.

The band saw has a cast metal frame and a plastic cover.

The cover encloses the back path of the blade, and the front path is exposed and is kept in it's normal path by the blade guides.

To my (warped?) mind, the front of the wheel is the plastic cover FACE. The back of the wheel is the FACE closest to the frame. Adding to my propensity to cornfuse, is that I think the EDGE towards the blade guides is the front edge, and the EDGE towards the upper auto-track quide bearing is the back edge.


Yep front/back, left/right, face, edge, wheel, enclosure . . . yields a lot of possible ways to mix it all up.

Surely you understand the need for introductory definitions of future references therein. I did not do that!!!;)

[algale]

After resolving the problem, JPG asked me to take some cant measurements and post.

And I'm getting some non-standard results.

With the Shopsmith 5/8 inch blade mounted and tensioned to the recommended 1/2 inch setting (and I have verified the tension scale has been set/calibrated per Shopsmith instructions), there is I believe both NEGATIVE cant on the upper wheel and the entire upper wheel is deeper in the casing than the lower wheel.

To whit, when I place a straight edge across the upper wheel, and slide it down toward the lower wheel, it just makes contact on the tire of lower wheel at the 12 O'Clock position.

But, if I put a straight edge on the lower wheel it clears the bottom edge of the upper wheel by about roughly 1/4 and clears the top edge of the upper wheel by roughly 1/8.

I must emphasize that there is no tendency either when turned by hand or under power for the blade to come off the wheels and the blade (1) is kissing the groove in the autotrack bearing and (2) is making solid contact on the lower back up bearing.

I must also add that I am getting good results with my cuts. So, what gives??