Latency, Jitters & Lost Steps

[rustyolddo]

Also, go to the downloads section here on the board, look for a file called "Guinness", download it to test with. I posted a link to what it's supposed to look like.

[ger21]

I thought from the explanation in the docs that changing 'I' would change the Radius of the circle, but this also changed the path from 180 degrees to 360 degrees.

I and J specify the center point of the arc. Moving the center will change the radius, but the start and end points stay the same.

[Imagineering]

Rustyolddo, Thanks for that info, I'll try it out tomorrow with 4 decimals.

Ger21, Thank you also, I now have a better understanding of the G0 command.

I thought that I'd solved the problem of 'jitters' but apparently not. The machine is behaving perfectly except for the occasional -Z move. It'll be happily tooling along with a fixed Z-Axis height and then for no apparent reason the Z-Axis develops the jitters and starts slowly moving in the -Z direction. I've had a play, and it's not apparent to me if it's caused by RFI from either of the X or Y Axes. No other electric/electronic devices were running at the time, so that is already ruled out.

My electronics are now in bits all over the bench, as I am now fitting shielding braid over every wire from the Power Transistors to the Axis Sockets in case RFI is the culprit. All my cables are shielded from the Sockets to the Steppers, apart from the last 30mm at the Z-Axis Stepper. I'll do all the earth shielding work, throw it back together and see what happens.
I've also fitted up a 4N25 opto driving a relay for Spindle Control while I've got it apart.

More later . . .
Thanks guys.

[ger21]

Rustyolddo, Thanks for that info, I'll try it out tomorrow with 4 decimals.

I believe by changing to 4 decimals, you'll still get g-code with straight line segments, but they will appear much smoother. I don't think you'll get arcs (G2, G3), but you should get better results.

[rustyolddo]

I believe by changing to 4 decimals, you'll still get g-code with straight line segments, but they will appear much smoother. I don't think you'll get arcs (G2, G3), but you should get better results.

I was running text and had a similar problem with the jaggies. Adding decimal places and it cleaned it up. I'm engraving text that is .060" in height.

[Imagineering]

AAAArrgh, I've just spent the best part of the day putting earthed shielding braid over all my cables that run between the Logic Board & the Power Transistors & Resistors, only to find that it did it again. I did a perfect run with a text of 'ESO3' to test straight lines & curved text together, then I tried it again on another board only to have it do the 'jitter -Z' thing again. As I have a Dremel bolted to the Frame about 100mm from the Z-Axis Stepper, I'm thinking towards RFI from the Motor Brushes. I've marked the Z-Axis and I'm now running the test program through five times continuously, without the Spindle switched on, so as to confirm or negate my theory.

The good news is that Rustyolddo was right about the 4 decimal places, I engraved text of 30mm height with not a jaggy to be seen. Thanks guys.

[Imagineering]

Yippee, I've run through the ESO3 text program five times without the Spindle turned on, and the Z-Axis returned precisely to the original start position after 1 1/4 hours of manouevering. I'll move the Dremel Motor to the back of the Frame tomorrow & try again . . .

[rustyolddo]

Do you have a o'scope?

Whose controller are you using?

The only way that I can see motor noise causing jitters is if it's getting into your dataline from the computer, Step & Dir.

Are you running a laptop or a desktop style computer? Laptops are known to have low voltages on the Para port that can cause problems.

Need more info on your setup.

[ger21]

Try plugging the spindle into a different circuit. Not a different outlet on the same circuit, but a different circuit.




EDIT: I should have said a different circuit from the computer.

[Imagineering]

Try plugging the spindle into a different circuit. Not a different outlet on the same circuit, but a different circuit.

EDIT: I should have said a different circuit from the computer.

I've got enough flexi drive cable to move the Motor Unit to the back of the machine frame, so I'll do that for starters. If that doesn't work then I'm not sure how to go about your suggestion, as I only have one electrical feed into the Workshop.

[ger21]

Can you try running an extension cord to run the computer with?