cnc conundrum

pescadoman

pescadoman

Randy
Silver Member
My table saw machine....well the stepper movements, have me stumped(easy to do).

Here is what I know.

1. My bar is flat within .005

2. I can cut a dowel within a couple thou. end to end.

3. To drop down to .650 on the small end I have to input .580

4. The .875 on the other end is right on the money

If Mach is getting to the .875 then why not the .650?
 
Randy

What kind of saw machine?

Is the axis movement you're talking about move in a straight line or an arc?

If it move in an arc, then you would have to work it out with "screw mapping". It's under the function config's drop down. I have a 4 blade saw machine that moves in an arc, but it's a pretty big one and the deviation is small so we don't worry about it. We only use it for roughing passes anyway.

If your mechanism moves in a straight line, then I'd put an indicator on it and check the travel against a commanded move. You may need to adjust the steps per unit on that axis. Mach3 does have an "Axis Calibration". It's on the Settings screen just above the reset button. I haven't used it, but I believe it has you move the axis a certain distance, and it counts the pulses it took and writes those into your motor tuning settings.

I hope that helps!


Royce
 
Might as well post your G-codes .

I've been there , done that . I had to write the numbers wrong .
Thanks to MC2 who told me to use the centers as Y0.
 
JoeyInCali said:
Might as well post your G-codes .

I've been there , done that . I had to write the numbers wrong .
Thanks to MC2 who told me to use the centers as Y0.
Click to expand...



Man, I would hope that the Y axis zero is at the lathe center! That's the only way the numbers can really mean anything.

We also cut the y steps per unit in half so that when we command it to move 1", it actually only move .5". That way our DRO's always represent the diameter and not the radius.

And yes, MC2 was the one who first suggested that to me.


Royce
 
Last edited:
pescadoman said:
yes it is....
Click to expand...

Randy,

A couple more thoughts have come to mind.

If it's the arc movement causing the issue, the diameters will be consistent even though they are off. It wouldn't matter where you were along the X axis, a commanded Y position would always be off the same direction by the same amount.

But, if the steps per unit number were not correct, it could show similar results. The difference would be in measuring the movement. If a dial indicator shows .100" movement when you command a .100" movement, then your motor tuning should be good. But this must be measured at the screw and nut, not out on the arc movement where it could be skewed. If that shows any error, then you need to correct your steps per unit.

If all that is good, then you're dealing with the arc movement, and here's what I would do. Pick a diameter that is right in the middle of the range that I would want to use, and and adjust the DRO's there. So, if you wanted to use .750", I would cut a test piece using the jog functions and get the piece to the .750" size. Then reset your DRO to .750", and reference that axis at that number. (I'm assuming you aren't using home switches)

That way you should be accurate right in the middle of the range you're looking for. The variance might be small enough that it won't matter to you, or you could bump the program numbers a little to get it where you want it.


I'm curious to see how this works out.


Royce
 
RBC said:
[...]

That way you should be accurate right in the middle of the range you're looking for. The variance might be small enough that it won't matter to you, or you could bump the program numbers a little to get it where you want it.


I'm curious to see how this works out.


Royce
Click to expand...

There may be more to this problem than you realize at first blush. I'm not familiar with the Bludworth tablesaw machine, but I assume it's essentially similar to any modeled after the original Kersenbrock concept. If so, that relied on a tapered bar lying on top of a roller bearing ("cam follower"), with the profile of the taper bar determining the shape of the shaft (or butt) and the height of the cam follower determining relative diameters of the tapered piece.

Typically that worked fine, due to the taper bar's carefully determined shape compensating for the arc of the cradle swing. However, when running a straight bar, and using the cam follower's CNC-controlled height to produce the taper, odd things can happen. Primarily, the straight bar will ride on one edge when the cam follower is higher off the saw table, ride flat as the c.f. is lowered and eventually ride on the opposite edge when the c.f. is nearer the bottom of its travel.

If this is the case there are basically two remedies. 1) commit a good amount of trial an error to developing an algorithm that compensates for the problem, or 2) cut a radius on the weight-bearing edge of the taper bar, such that it forms a perfect (or near perfect) semi-circular edge matching the thickness of the bar stock.

When I had to solve this problem - converting one of Joe Gold's machines to CNC - I chose solution #2. Worked quite well, only requiring a tiny bit of compensation for the arc swing of the cradle. If I had to do it again that is how I would go.

TW
 
One of the two centers may be higher than the other to the actual bar your Z axis is adjusting. Also possible you have backlash in the ball screw that could cause your numbers to wind up off when the machine turns around and goes the other way.
 
I won't be able to get to this until next week. I'd like to thank all the knowledgeable people who have taken their time to try and help me...
 
Randy,

I also was in your position with the Blud Machine and did not like the situation as you have described also. One thing I do remember was that my motor tuning on Mach Mill had to be ramped up on the X. That big alum sled offers a ton of resistance due to the friction on the saw table and groove. As I remember the numbers were too slow. That was 8 years ago and I don't remember all of the details and am not an expert with CNC. Only an operator.

After a lot of trial and error getting my machine to work,my taper was still not exactly comming out as plotted on the Cad layout. So I retrofitted to a taper bar set up. I have been delighted ever since.

On my shaft tapering saw machine I converted my Bludworth machine to a taper bar system. On my retrofit I got rid of that big alum billet sled with the uninterupted lead screw and went to a half nut system with a E Stop where my cut is conventional and the machine only cuts in one direction.

I purchased 4 adjustable controllers and 4 DC motors for two machines. The motors were properly sized and torqued for the job. These motors can run all day and barley get luke warm. My butt machine is custom made (not a Blud) also built this way with my butt contour taper bar.

I have three proprietary taper bars for the shaft machine with pivot points on the tip end and can create taper differentials with a fixed mic adjuster to create the tip size I need for each taper contour holding my joint size. These bars were built on a big CNC Mill.

So now I can walk up to my machine drop in a shaft, turn a dial and watch a dial indicator to control my Z elevation cut. This process takes me less than 20 seconds and I can walk away from the machine. 6 minutes later the machine shuts off. I don't have to start up a computer, load a g code and touch off to the piece with a keyboard comand.

I prefer taper bars to using the CNC for tapering because it produces a repeatable precision taper and it is quick and easy to use. It works good for me because I am not interested in making any shaft contours but my own in my shop. This is not a good option for those who produce a ton of different tapers and I understand that.

In my view, sometimes the simplest method is the easiest and most reliable.

I am not trying to talk you into a retro fit by any means but if anyone out there is building a taper bar machine, give me a call and I can turn you on to the DC Motor and controller company and specific model numbers. The motors that I have took me over a week to locate them and the company gave me great engineering support to match the motor parameters to the job at hand. This was a bigger pain in the butt than the retrofit.

Good Luck Randy,

JMO,

Rick
 
Last edited:
Get rid of the arc and stick with compound linear angles. :thumbup:
Switch to Smooth Stepper like RBC's toys.
 
Sorry about your problem but I love threads like this. Everyone trying to help, alternatives etc. I will be much interested in what out come will be.
 
You must log in or register to reply here.
Back
Top