Hello Guest it is November 11, 2024, 08:07:29 AM

Author Topic: Lathe Tool Compensation Is there a better way??  (Read 17797 times)

0 Members and 1 Guest are viewing this topic.

Offline Dan13

*
  •  1,208 1,208
    • DY Engineering
Re: Lathe Tool Compensation Is there a better way??
« Reply #20 on: January 17, 2010, 09:10:06 AM »
I would assume you would use the electric probe as is, without machining it to setup the tool table and then when you're set to machine the part, take a cut and measure it.

Daniel

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
Re: Lathe Tool Compensation Is there a better way??
« Reply #21 on: January 17, 2010, 09:27:47 AM »
Ok so say you do that, you set your master tool off the probe, set all other tools to the master via using the probe then you have to machine the part, measure then offset the master again, sounds like we are again back to a manual method.

Hood

Offline Dan13

*
  •  1,208 1,208
    • DY Engineering
Re: Lathe Tool Compensation Is there a better way??
« Reply #22 on: January 17, 2010, 10:47:10 AM »
Why to set you master tool off the probe? Any reason it could change its position? I was actually more thinking in the way of making a macro to take advantage of the electrical probe and set all the tools automatically with respect to the master tool. You could probably use temporary DROs to store the true position of the master tool before touching off the probe and then touch it off, override the X and Z DROs with say 0,0, and set all the tools with respect to that, then return back the stored true position of the axes - all in a single macro. This way you don't need to reset your master tool each time.

If you do it this way, and considering your machine's repeatability, I think you don't even need to machine the part each time and measure it.

Hope it makes sense...

Daniel  

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
Re: Lathe Tool Compensation Is there a better way??
« Reply #23 on: January 17, 2010, 11:09:58 AM »
You seem to be moving the goal posts every time you reply ;)


Hood
Re: Lathe Tool Compensation Is there a better way??
« Reply #24 on: January 17, 2010, 11:51:45 AM »
OK i have been giving this some thought and there is a solution for most tools.

Permanently attached to the headstock would be two probes one for X and one for Z this would be behind the chuck so as not to interfere with the working environment.

Great idea or what.

Phil
The Good Thing About Mach3, Is It's very Configurable

The Bad Thing About Mach3, Is It's Too Configurable

Offline Dan13

*
  •  1,208 1,208
    • DY Engineering
Re: Lathe Tool Compensation Is there a better way??
« Reply #25 on: January 17, 2010, 01:12:50 PM »
You seem to be moving the goal posts every time you reply ;)


Hood

???

Offline bowber

*
  •  216 216
  • Kirkby Stephen,Cumbria, UK
Re: Lathe Tool Compensation Is there a better way??
« Reply #26 on: January 18, 2010, 03:38:46 PM »
You could make a probe plate from a vertical angle or square section, have this mounted to a clamp that fits to the lathe bed and a stop for the front of the gap.

If it's made with some adjustment you could get it on centre (I think mach lathe allows for this) this way it would be repeatable and accurate and no need to do the dia turn to set the final tool position.

Write a macro to do the touch off etc and you should be able to reset mid job, assuming the stock doesn't interfere.

Steve

Offline tjhj

*
  •  26 26
Re: Lathe Tool Compensation Is there a better way??
« Reply #27 on: January 18, 2010, 09:39:01 PM »
So maybe I am way over thinking this, but why not add the ability to measure the angle of the tool post.
Ok so here is the idea,
Take a Isolated Probe, mount it in a standard tool holder. Rounded of some kind.

Now I normally do embedded systems, so I am thinking take a potentiometer and mount it to the tool post in a manner that as the angle of the tool post is changed it changes the pot. It should not be a problem to achieve 12 bit resolution. And as most single turn pots have a rotation of 260 degrees. So considering that 10% of the ADC resolution will most likely be lost due to not getting a perfect ending on the pot. It should be quite realistic to get a position range of 0-3686 over the 260 degrees. The result is accuracy to 1/14th of a degree. If a 14bit adc was used 1/56th of a degree accuracy could be obtained.

I think the 12 bit solution would provide accurate enough angle offset to accurately position the tool.

So knowing the angle and difference to each tool from the probe in X and Y at 0, the new offset can be found using a macro.

The biggest question is how to get the information into Mach. I was thinking I could have it output pulses based on position, per angle. Maybe as an additional axis encoder? Then the macro could make a new offset based on the encoder position.

Maybe?

Does this even make sense.... Or am I missing the new point.

TJ