The bug is still there!

[myxpykalix]

Trying to round stock turns into tapers.
Here is what i have done. I raised the height of the indexer 2" to eliminate any depth limit issues it may have had with shorter bits.

I centered the tailstock to the divit in the center of the indexer shaft, so its perfect.

I made a template to the center of the height of the divit Zzero'ed from the template.

chucked up a piece of 2" square cherry. Used a centerfinder to mark the center of the cherry for the tail end.

I input this info:

7623


And the measurements of the wood was:
1.996 squared
1.803 rounded at headstock
1.630 rounded at tailstock at 13 inches

I'm running ver 3.6.1 beta 6
-----------------
here is code:
VD,,5
'----------------------------------------------------------------
SA 'Set program to absolute coordinate mode
SO,1,1 'Turn on router
PAUSE 2 'Give router time to reach cutting rpm

' file cuts in 1 rough passes plus a finish pass
ZB
MZ, 1.25
MX,0
M5,0,,0.9375,,360
MB,720
ZB
M5,13,, 0.9375,, 234000
ZB
MB,360
ZB
MZ,1
MX, 0
MZ,0.9375
M5,13,, 0.9375,, 468000
MZ,1.1875
SO,1,0 'Turn off router
END

[bill.young]

John,

1) Has the z-axis lost position? If you return to the X-axis location where you first zeroed the Z-axis and move the Z axis back to 0...MZ,0...is the bottom of the bit still exactly at the center of the chuck, or has it moved up or down?

2) Moving the tailstock up to the headstock doesn't really tell you if your Indexer track is aligned correctly...it has to be lined up with your ShopBot's travel both side-to-side and up-and-down.

To check, move the tool so that the bit is near the chuck end and the Y-axis is centered on the centerline of the chuck...your Y=0 location. Next move the Z-axis down to the center of the chuck...the Z-zero location...and then move the tool in the X-axis until you get to the tailstock end. Is the point of the tailstock's center exactly at the bottom tip of the bit in both Y and Z? If not you'll need to adjust the track until it's exactly centered or you won't get accurate turnings.

Bill

[cmagro]

Jack,

To add to what Bill said...if you move your tailstock a lot you'll find that when you loosen the bolts you can easily jiggle it up and down a good 3/16"...I've had sawdust in the track that if not blown out would have caused the problem you are having. Whenever you move your tailstock you need to make sure the center of the head and tail are still the same.

Christian

[bill.young]

3) When the indexer was cutting and moving down the blank, did the display for the Z-axis in the red panel stay at .937 the whole length of the turning, or did it change as it moved along?

[myxpykalix]

Bill,
I took my template set my Z zero to center of chuck.
Ran the bit down to the tailstock and the bottom of the bit matched up with tip of live center on tailstock.

Ran the part file again and watched the readout to comfirm it never went below .938

Ran it and the heastock end of turning measured 1.78 and at the tailstock end it measured 1.63

I took the part out of chuck, took template and did a MZ 0 and it came to the top of the template and stopped.

My belief is that even if your stock is offcenter the part that is cut will/should stay the same diameter at the top as the bottom.

This taper wasn't as bad but there is still something going on. I don't believe it is anything physical out of alignment. Any clues?

[cmagro]

Jack,

Check on something for me...

Look at the round area where it meets the square on each side and tell me if it circle is centered in the square or not.


7624

And while you are at it check to see if there is a little blond girl sitting in front of your TV while static is on...if there is ask her if "they" are here yet.

Christian

[myxpykalix]

Looking at the center where you have your arrows pointing to, there is a difference maybe the thickness of a line drawn with a pencil. Not enough to make a difference.

[cmagro]

I am really stumped at this point Jack...I'll take another look at all of the info again on my next break and see if something was missed..

I do have something else to try though...I'm going to send you a file with the same parameters but the turning will be done going back and forth instead of rotating..I'd like for you to try that and see if it tapers.

Christian

[myxpykalix]

i'll be checking my email and give it a try. thanks

[cmagro]

Jack,

Put a new blank in your indexer and run this file.




7625 (1.4 k)

Same parameters but I've programmed it for you to use a 1/4" straight or spiral bit so make sure you use that.

One more thing....you need to position the blank with a flat side up and square (or level). I usually just put a small framing square against the bot and square the blank. The reason for this is the program will rotate at 45 degree intervals removing the corner sections first.

Let me know how it comes out..

Christian

[myxpykalix]

I'll sneak out to the shop now and try it, hopefully won't wake the neighbors, if not i'll get to it in the AM. I really appreciate you taking the time to help.

[myxpykalix]

Here is the results of my test. The rounding still is out by about .15
Here is my preperation for and cutting of the file, tell me if you see something i dont.


7626


7627


7628


7629


7630
shows square and parallel to the aluminum track


7631
finished cut at top of stock


7632
finished cut at bottom of stock .14 difference


7633
look to the left for perspective of difference in measurements

[bill.young]

John,

I know I'm sounding like a broken record here, but the only thing I can imagine that explains this is that your track/tailstock is out of line with the movement of the gantry. My suggestion is to doublecheck that the bit stays centered as it moves between the headstock and tailstock.

If the tailstock end of the track is too high, the rounded section will still be centered in the blank but smaller than it should be...if it was too low that end would be a little bigger. If it's off from side-to-side it will still be centered in the blank, but I'm thinking that the tailstock end will be too big.

So my vote is that the tailstock end of your track is a little high in relation to the motion of the bit. It only has to be off by .065" to show that much taper.

Bill

[myxpykalix]

Bill,
i appreciate your answers even if i have "heard that tune before" lol. However I don't think you are correct as the second picture shows it centered on the shaft plus I have the headstock and tailstock aluminum track bolted to a long 2x6. So they are on the same "plane".

When I bolted the 2x6 to the frame i measured the distance in the "gully" and the only other thing that occurs to me is to check again the X movement against the position of the aluminum tracks again.

[myxpykalix]

UPDATE:
Still problems but to eliminate track/tailstock issue I remeasured my setups

7634
setup block to center of shaft

7635
bit zeroed at center of shaft

7636
tailstock measured to template

7637
headstock parallel to tailstock from the top

7638
headstock parallel to tailstock from the side

7639
laser line set to ensure track is parallel

7640
Notice that when the bit is driven down 24" to the tailstock the end of the bit is about 1/4" below tip of tailstock.

Since the wooden template and the tailstock to headstock verify to me that they are on the same plane and are level with each other both in the X and Y by virtue of the pictures this tells me where the problem is, is in the Z axis moving down.
I have tugged and pulled testing the Z axis and its not loose or moveable by hand.
I believe it is a software issue since we had this problem before as Bill stated.

Another reason i believe this is the case is because when i did my 8' tall fluted column I did not have any problems. I have upgraded a few months ago to 3.6.1 beta 6.

[myxpykalix]

also when moving the carriage in the X the readout stayed at zero but obviously it moved down.

I cannot imagine that anything is wore out since i probably don't have 200 hours on this bot.

[jdgrahamwaldorf]

On everything we do with the shopbot we assume that the bit is travelling on a flat plane. It appears that you are zeroing on a block at one end then sliding it to the other end on the table to check the the tailstock that is also attached to the table. Was the bed of the lathe "surfaced" like you would on your regular table?

[cmagro]

I still think you are missing one thing....you've done a lot of centering but with regard to depth of cut the only thing you need to make sure of is that the bottom of the bit remains at the correct height relative to the center.

I may be missing one step and you may have already done this but....

secure a v bit and locate it at the head in the center (in both directions)....manually move the bot along the x to the tail end....is it in the center as far a z and y?.

Again you may have already done this but this is really the most important thing to verify.

Christian

[bill.young]

John H,

If I'm looking at it correctly, your last picture shows that a bit that's centered at the headstock end, when moved to the tailstock end, is lower than the center of the tailstock...that the tailstock is high in relation to the bit and the ShopBot gantry?

The measurement from the track to the headstock will ALWAYS be the same as the measurement from the track to the tailstock...it's the relationship between the ShopBot gantry's motion and the Indexer that's important

Imagine putting a piece of 2" material on your ShopBot before you ever surfaced the table, with one end of the tabletop lower than the other. Measuring from the tabletop to the top of the material (like measuring from the track to the headstock or tailstock) would show that the material was parallel to the table, but when you cut that material one end would be thicker than the other. It's the same with the indexer...the track has to be parallel to the motion of the tool, both side-to-side and up-and-down, by surfacing the material that the indexer is mounted on as John G. suggests above.

Bill

[waynelocke]

I agree with Christian about centering on the headstock and manually moving the gantry. Do Christian's suggestion at different points along the track by moving the tailstock.

To expand on Bill's explanation, the track could vary in the Z direction so that it could be correct at the 8' location and off at the 30".

I would also look at the 2 X 6. Did you surface it? It could twist and turn and not be a consistent dimension. I know that it is bolted to the frame but if it wants to twist or such those forces are going to show up somewhere. Bolting the track to different spots could also be a problem. Some parts of the 2 X 6 could be softer than others or the track could be bolted tighter at one spot than another. A piece of oak or maple would be a better bed or even laminating several pieces of plywood.

[scottcox]

Jack,

That last picture says it all. Your tailstock is higher (in relation to the bit) when you move it away from the indexer head.

I would concentrate on placing shims under your t-tracks so that in any position on the x axis, your head and tail are still parallel to the bit.

I think Wayne's suggestion to use a hardwood or plywood would really help to eliminate future movement, but any material will have to be surfaced in relation to the bit, not just planed flat.

[myxpykalix]

I appreciate having all this help to figure this out because i have outthunk myself on this.

You are correct in the observation that the tailstock to bit relation is off. However if you look at the 3rd and 4th from bottom picture) you will notice that the headstock to tailstock relation is correct in the X and Y.

If the heastock and tailstock is level with each other than the variable would be the Z axis?

Also if the 2x6 was odd thicknesses as suggested to make the bit lower by 1/4" over a 24" span would mean the 8' long 2x6 would have to be at least 1" (just a guess) thicker at the opposite end to cause what you describe and its pretty uniform.

I also checked the rails and they are level. I checked the level of the bed that the indexer is in and its level.

My next test is to run a string line with a string level independant of the table and run the bit down that line. If it stays level then it may be infact in the indexer setup. Again I appreciate all the help i'm getting here because there are "sharper pencils in the drawer" who can think of something i'm overlooking.

[jdgrahamwaldorf]

If you are levelng with a builders level try turning the level 180 degrees and seeing if the bubble is still centered. As a surveyor I always take meaurements in both directions to cancel small errors. Also try measuring down from the table that has been surfaced to see if there is any discrepency.

John

[jim_stadtlander]

Jack,
In your previous post(what the heck is going on)which is related to this problem, you had a picture that seemed to me seemed as if the taper was not uniform going down to the tailstock end. If you put a straight edge agianst the finished "bad turning" does it appear to have a belly or is it flat? It seems to me, that if the headstock/tailstock aren't lined up it still would be a flat straight taper.

If it's a flat straight taper, I would then suspect the alignment as to everyone's ideas above. Something may have gone out of alignment since you last used it. With regards to you thinking it is a software problem : in the direction(x/y?) that you think it is out of alignment, have you moved it to your flat table top and run it back and forth to see if it is out of alignment in the same way? Seems like it would do it there too.

If your tailstock is 1/4" difference down than the headstock, wouldn't that give you a larger diameter at the tailstock by a 1/2" ??

......Jim

[jim_stadtlander]

Oops, sorry just went back and seen that the tailstock is higher than the bit........So the finished turning should then be 1/2" smaller in diameter at the tailstock end, right????

[harold_weber]

Jack, instead of using the block of wood you labeled "Good Jig" in the photo, make another block that goes vertically from the center of the headstock to the bottom of the Y car when the Y car is over the headstock.

Now, with the center point on the tailstock 24 inches away from the headstock, move the Y car down to that location and use the block to check the vertical distance from the center of the tailstock point to the bottom of the Y car. Are the vertical distances the same? If not, you have eliminated the Z axis as bring the problem.......

[myxpykalix]

The one thing i did do recently is raised it up 1 1/2" (thickness of a 2x6) due to a issue of small bits bottoming out the Z before it reached the depth i needed.

I am doing a couple things now. I bought a 1/4" foot long steel rod and chucked it up. I am going to use that as a probe and run it down the side of the 2x6 to make sure they are inline with the X travel.
Then i will do the same for the T track.
Then i am going to run a string line with levels to check the travel in X and Z.

Because if i have an independant level line and I know my rails are level then if the Z goes below the line that tells me its a Z height issue (or disproves it) so i can check other things. Again I appreciate all the thinkers out there.

[cmagro]

Jack,

Anyway you can do some sort of surfacing before installing the track?

By moving the tail up to the head it is clear that the indexer unit itself is perfect height.

So if raising it up 1 1/2" had you put another 2x member atop the existing one surfacing it would make it nice and even.

The point here is that the frame that you put your 2x6 on is most likely the culprit and slightly off with the level of the bit along the x. So whatever you put on it will also be slightly off.

[myxpykalix]

Christian,
In theory that is a good idea however to get a flat surface relative to the indexer bay where it is to be mounted, I would have to surface it down there and my bits won't go that low. I am remounting and redoing everything as I write.

[beacon14]

Just surface the top of the last layer before you mount the indexer - you only need to surface the area that the indexer actually sits on.

[myxpykalix]

What the problem is, is that my Z won't travel low enough to surface the board as it is about 12" below the surface of the table. I have gotten far enough in my redoing of the setup to see that I am level and parallel so i think i'm ok on that part.

[scottcox]

Jack,

I used shims under the t-track to level mine. I just checked relative to the bit every 6 inches or so all the way along the x axis

[Gary Campbell]

Jack...
You may need to reset the indexer on its bed. Using a V bit similar to your pictures, set the head end with the V bit touching the center of your headstock. (better if you can use a point)

Then move your carriage down to the far extreme of the tailstock. Move ONLY the axis that is parallel with your indexer bed. Using the V bit aligned to the point on the tailsock, set the height of the tailstock the same as the head.

Now shim your bed complete from one end to the other. Once that is done, you can use the Vbit to center the indexer in the Y direction. Other than a few micrometer type adjustments, you should be within .010 in both directions.
Gary

[scottcox]

Jack,

Another thought... if you're using wood as your indexer bed, it may be beneficial to add a temporary extention to your z axis so your router will reach the bed in order to surface it properly.

[myxpykalix]

Ok I redid EVERYTHING.
Leveled the 2x6 base to the indexer bay.

7641

Headstock/tailstock parallel and level

7642

Line is still level with center just a different angle. Bit right at line level zeroed there

7643

bit at opposite end of X (8' down) now look at the gap

7644

Here is the bit taken down to the string level with the inset display showing a .300 difference

7645

This shows table is level

7646

Its too late to try a cut but i'll try that tomorrow.

[myxpykalix]

Scott,
I'm not sure what your talking about "temporary extention to your z axis". How would i make that?

[bcammack]

String a level line across the gantry, irrespective of the bits that move in the Z axis. Run the gantry from one side to the other and see if it has a similar deviation to that of the bit tip. I know that there shouldn't be if you surfaced the table, but...

IOW, take it one step back from the tip of the bit to something that ought not to be off, just to insure it isn't. This is sort of like asking, "Is it plugged in?", and usually garners the response, "Of course it's plugged in!! That's the first thing I checked!!", but having worked in a Tech Support role, it's important to establish a baseline of information and build upward from there.

My brother once bought a new laptop because he thought his Sony Vaio was dying. Turns out the power strip he was plugging into at the client site wasn't actually plugged into anything and he never bothered to check...

[myxpykalix]

I have leveled my indexer base of 2x6 but Scott suggested i surface it which would at least take that out of the equation as a factorhowever i don't have a bit long enough and i'm not sure what scott was talking about "temporary extention to your z axis".

Although I have not cut any full sheets of plywood in a long time to have to use the whole table but based on the picture of the .300 difference I would be cutting into the table if this table was that much out of square and pitch.

My next test is to take a 8' 2x4 on the X axis lay it on table, program a .25 deep cut along the full 8' then take my digital calipers and measure the depth.

I may take the indexer out of the bay and put it on table top and test it there (if it will fit).

[waynelocke]

Is the difference apparent when doing an MX or JX move? Is it also apparent when moving the gantry by hand? Move the gantry by hand and command over the table and check the z position.

Just a thought, check the VA command and make sure that the indexer is the B channel and the z axis is correctly selected.

[waynelocke]

Is the difference apparent when doing an MX or JX move? Is it also apparent when moving the gantry by hand? Move the gantry by hand and command over the table and check the z position.

Just a thought, check the VA command and make sure that the indexer is the B channel and the z axis is correctly selected.

Am I missing something? The picture showing the gap with the string at 8' shows the gap above the string. Based on the previous pictures and posts, the gap should have been below the string. No?

[Gary Campbell]

Jack...
I don't want you to take this the wrong way, but the levels that you show are no way capable of calibrating your machine. I use a 4' digital level with a 10' machined straight edge. Another great feature is that the digital can calibrate itself.

1/2 of 1 degree is virtually undetectable with a bubble vial. 1/10 of 1 degree is almost 3/16 (.167) in 8 feet. Due to their low quality, a string level has virtually no use.... anywhere. Most strings will not stretch tight enough to remove all bow. Even the best of levels applied to a bowed string cannot give acceptable results.

In reality, it doesnt matter if your bed or indexer are LEVEL. What is important is that their axis' are parallel. use a vbit to the center points to set the indexer parallel.
Gary

[ed_lang]

Jack,

IF you just cannot get this right, let me know. I will stop by and fix it for you.

No reason to go around and not enjoy the indexer.

Just let me know.

[myxpykalix]

Here is the latest results after redoing everything.
This is measurement at headstock

7647

This is the measurement 12" down to the tailstock

7648

I can live with it being 1/1000th off over 1 ft or less than 10 thousandths over 8 ft.

I kept saying to Bill Young "I know I'm sounding like a broken record here, but the only thing I can imagine that explains this is that your track/tailstock is out of line with the movement of the gantry." but he just didn't believe me.LOL
(sarcasm and humor)

Somewhere when i raised the indexer must have been where the ghost crept in and screwed me up.

I hope that for now it is resolved. I want to thank all of you who helped me figure this out, I would not have gotten this resolved without your help. Especially Christian (for making me files) and Bill Young.
It just reinforces the value of this forum and the "new friends" who have helped me out. You don't know how frustrating this has been.
THANK YOU!!!!

[cmagro]

Jack,

I'm very happy everyone was able to help you out in this matter..I also envy your patience..I would have posted a picture of a broken window and a tapered newel on the ground 50 feet from my shop by now.

On a nice fresh note...for those of us not sick of the indexer yet....

I am working on three projects involving the indexer...all newels.

One is a design like the newel I posted previously but the spirals go all the way through

One is a two-tone design using a walnut core and an alder shell..once carved the walnut will appear through the design (materials might vary - client is not sure).

And finally a newel (new design) that hollows out a section for a light and hollow flutes (or spirals) up to the top.

I will be posting them in the Show and Tell section and make sure to take plenty of pictures.

Christian

[myxpykalix]

a hollow spiral like this?


7649

I have some links i'll send you that might help you some with your designs.

[Gary Campbell]

Jack...
Glad to see you are up and running.
Gary

[myxpykalix]

thanks, and thanks for all your help I appreciate it.