Go Down

Topic: CnC Person needs Arduino code Person (Read 4204 times) previous topic - next topic

jumps4

Today I repaired my mess up from yesterday and reinstalled all the optocouplers facing the correct direction.
I ran a test and everything worked good.
I did a short calibration test and everything seems good.
I'm waiting on parts for tomorrow to install switches.
There is still a lot of work to do to get everything mounted in its final place.
Steve

jumps4

#481
Nov 26, 2019, 10:07 pm Last Edit: Nov 26, 2019, 10:10 pm by jumps4
  I have finally got back into the shop today and have started trying to place everything in the enclosure.
I still have to mount the controls for calibration, the power supply and stepper motor controller.








jumps4

#482
Dec 02, 2019, 04:06 pm Last Edit: Dec 02, 2019, 04:15 pm by jumps4
Progress
I have made all the shielded cables and assembled everything onto the machine for real world testing.
The enclosure pic for the power supply and motor controller is blurry, but the time I seen it, it was assembled.
Today I will start testing and I'll make a video showing everything working including calibration.
I still need to make tool holders for collets, drill chucks and boring bars that bolt to the side of the turret.
Steve




















jumps4

#483
Dec 03, 2019, 08:16 pm Last Edit: Dec 03, 2019, 10:55 pm by jumps4
I made a video of testing the calibration of the turret.
It is shaky...

Video of turret calibration


Video of
Automatic tool offset
measurement


Video of
First part made with turret



wildbill

Getting closer to cutting metal I see, calibration looks good.

I think I need to limit the number of messages about calibration though.

jumps4

You are amazing Wildbill
There was not one issue I have seen yet.
 Both parts are identical
Steve

wildbill

Here's ufe. I took the version you had set the constants for and made a few cosmetic changes: moved definitions around to more logical places, better comments and suchlike. In theory, it should be no different to what you have functionally.

Then I made a small change so that the repeated message about tool calibration only appears once. Not important, just makes reading the output a little less painful.

If you do use it, please test it carefully away from the workpiece. I don't want to find that one of my "cosmetic" changes wasn't quite as harmless as I thought.

wildbill

Here's uff.

As discussed, the turret rotates continuously at power up. This is because the Due is looking for a LOW signal from the Acorn to tell it to run the stepper. Until the laptop driving the Acorn boots and connects, that signal is low, so we get perpetual rotation.

Now, the Due waits for a press on the calibration move pin before it will consider moving or calibrating. Hopefully, it doesn't interfere with regular calibration.

jumps4

#488
Dec 04, 2019, 05:37 pm Last Edit: Dec 04, 2019, 07:13 pm by jumps4
Wildbill
 There is a problem with both new versions.
The turret will not move at all in ufe or uff
reloaded ufd and it works.
Steve

EDIT: I found the issue, pin assignments were reversed for step(pulse) and direction.
I will run more tests today now the new version works.
Thanks WildBill

wildbill

Hmm, my copy of ufd has the same pin assignments for the stepper as ufe and uff. I think we must have diverged inadvertently. If you can let me know what they're really set to now (I assume 51 for dir and 53 for step), I'll make and post ufg with the correct settings so we can be sure we have the same thing.

jumps4

They were accidentally reversed then I wired the board and I probably forgot to tell you...
my bad   lol
 All test are going very well and the turret requires a press of the rotate button prior to movement.

jumps4

#491
Dec 04, 2019, 09:20 pm Last Edit: Dec 04, 2019, 09:23 pm by jumps4
 I set stop delay to 5000, cal is off bad from -.012 to +.012
set back to 1000 and
going to test ufd again

wildbill

Something is odd. I see in the text file that tool one showed up at 1000 pulses, taking 4000 steps to achieve. Clean as could be. But tool two at 2000 pulses needed 4004 steps. Tool three at 3000 pulses took 3988 steps.

This isn't new though, I see similar variation in earlier logs where the gauge was showing phenomenal accuracy. It seemed then like the encoder was the source of truth, irrespective of what the step count said. Now even that doesn't seem to be correct. I'm confused.

jumps4

I believe The issue is mechanical.
It is repeatable and consistent and varies to a particular tool number.
1 always high
3 always low
7 always low

tool 1 is always .014 high so if I calibrate it -.014 the result should be zero

jumps4

#494
Dec 04, 2019, 10:49 pm Last Edit: Dec 04, 2019, 10:52 pm by jumps4
Tomorrow I will disassembly the turret and check the alignment of everything.
I may have been reversing too hard and set the alignment off inside for the sprag.
Maybe the sheer pin is bent.
now tool 7 calibrated high, it was calibrating low
I'm confused myself
Steve

Go Up