Loosing steps in my X.

[DanO]

Running Mach3 with a Gecko 540 turning Kerk lead screws. Using Bobcad 23 for the programming. Ø.031 cutter stepping out and down in .015 increments. Went back and cut some simple diamonds without an issue. I’m certain my motor coupler is tight. Looking at the program my XY position prior to the Z move is identical on every repeat. I did notice that when I changed my cutter size and recomputed the tool path it changes the error. It's really not as extreme as the pic since it's kicked at an angle.

I still think it’s gotta be mechanical but everything is tight? I’m going to rotate my geometry 90º tonight and see what happens.

 

[scdiveteam]

Hi Dano,

Mr. Hoppe and I have seen this and our problem was that BobCad can glitch the G-Code sometimes when you re post and save it after changing a verified vectorize part. We recomputed the geometry for the tool path and re posted and saved and then tested in wood. It was fine.

I think our problem came as we had translated the geometry to scale it down a % without re testing in wood. In the future if we change anything after we have a tested and proven the G-Code with our compensation for fit, we will always re test so it does not happen on a cue again. Once the G-Code is tested it should be good for the long haul.

I was thinking some thing was wrong with the machine, controller or steppers but Troy knew intuitively to start at Bobcad and work in that direction. It's nice to have a computer professional in charge when you have problems that need trouble shooting.

I hope this helps. Cool Skull!:shocked2:

Rick

 

[DanO]

I sure hope it's a glitch. I draw in Autocad then bring it into Bobcad. It's like it looses it's X home position. I'll bring it in fresh and start a new file. I'm also going to rotate 90 and see what it does. If the fresh file doesn't work I'll pass it to you guys to be amused.

As far as the skull goes, I drew the eyes and nose symetrical so i could inlay them with stone if wanted. First off though I'm playing with making the eyes part of the pocket. Since that part of it looks like it will work I think I'll change them to look more like eyes than stones.

 

[Mr Hoppe]

If you have verified that your X Y coordinates are the same for every pass, then it doesn't sound like a problem with your G-code or your cad/cam program. I would be looking at Mach3, drivers, and mechanical components.

Mach3 has an Lpt port sensitivity setting that can cause or help missing steps at times. IIRC, it increases/decreaces the signal strength that is sent to your controller I'll look up that setting and let you know where to find it, but that's the next thing that I would check.
Has this machine ever worked properly or is it a new setup? If it's not new, ask yourself what has changed since it last worked properly. Run a program that you know has worked in the past and check the results.

You can also use the MDI part of Mach3 to move x a certain number of inches, measure to validate accuracy, then repeat the test to see if you are dropping steps.

 

[Mr Hoppe]

Dan,
The parameter that I was referring to is found under Config > Motor Tuning. Then look at step pulse on the bottom of the resulting screen. I believe the signal strength values are 1 to 5. Tweak and test until movement is smooth.

Did you do any testing on this last night?

 

[Mc2]

If steps are getting missed

Dano, I would think if X where missing steps then the bottom of your skull would be too large as well. If you decide that X is missing steps, it may not be all mechanical. It could be loosing the steps as the X rapids back to its starting point. To rule this out, reduce the velocity setting for the X axis in Mach 3 motor tuning. Cut it in half. The machine is going to screw up in a rapid move over the part. Not likely going to happen when you are cutting wood as the bit will just break. If you think it maybe the controller, lets us know. I will make some suggestions on checking each axis out.

Good luck.
Jim.

 

[GBCues]

Has anyone ever asked Mr. Wilson about opening up a CNC section?

In some ways, it's a little off-topic, but in so many other ways, what cuemakers do with CNC is unique.

CNCZone.com is nearly impossible to navigate, it's gotten so large.

Just wondering.

Gary

 

[Mr Hoppe]

IMO we don't have enough CNC related questions here to warrent it's own section, but if we did, that would be a nice section to have.

 

[mortuarymike-nv]

cnc

hi dano
one good way to test to see if it is mechanical.
cut a circle and then cut a dome if it is mechanical it will show up there.

mike

 

[Guerra Cues]

IMO we don't have enough CNC related questions here to warrent it's own section, but if we did, that would be a nice section to have.

Yep, that's a great idea.


Sent from my iPhone using Tapatalk

 

[scdiveteam]

Running Mach3 with a Gecko 540 turning Kerk lead screws. Using Bobcad 23 for the programming. Ø.031 cutter stepping out and down in .015 increments. Went back and cut some simple diamonds without an issue. I’m certain my motor coupler is tight. Looking at the program my XY position prior to the Z move is identical on every repeat. I did notice that when I changed my cutter size and recomputed the tool path it changes the error. It's really not as extreme as the pic since it's kicked at an angle.

I still think it’s gotta be mechanical but everything is tight? I’m going to rotate my geometry 90º tonight and see what happens.

Dan,

If you went back and cut some simple diamonds and that was not an issue, I would think your controller, motors and machine is OK.

Does all of your other Tested G-Codes repeat ok?

Rick

 

[DanO]

With some invaluable advice from a few select people on here, I built my machine last winter. THK ball slides, Ø3/4 Kerk screws with zero backlash nuts.

http://www.timocues.com/CNC/Inlay machine1.wmv

I haven’t really got out of the box with it till now, just simple diamonds and darts. I did make an attempt at Arnot’s shark but mine looks more like a porpoise. You can call whatever you want if you buy it.

When I did cut the two “porpoises”, one tail had a slight gap. Might have been the start of my problem?

Should have checked this thread yesterday but here is what I’ve tried last night. I jogged about 18” down my X and ran the same program. Same results but maybe not as severe of an X shift. And yea Jim the bottom of the skull is also shifted like the top. I let the machine sit a few hours, then rotated the skull 90º in Autocad. Sorry I’ve been on Autocad for 20+ years so why fight it. I brought it into Bobcad and regenerated the exact same program settings. IT CUT PERFECT. Last I brought in my original Autocad skull and remade the program fresh. IT SCREWED UP AGAIN. What does this mean. I’ll tell you what this means. I have know freakin idea what this means but I can’t turn my cue 90º. :smile:

The rapid move in mach3 makes sense. The actual cutting section looks to be repeating perfectly, it is just being shifted in the X. I’ll slow that down tonight and see what I get. Also all of the geometry is radius’s by the way. No elipses anywhere.

One last thing, Alton mentioned it may be heat related. My Gecko control box has two small fans but there might be something there also. I’m noticing the second test cut seems to be worse than the first.

THANKS GUYS!

 

[Guerra Cues]

With some invaluable advice from a few select people on here, I built my machine last winter. THK ball slides, Ø3/4 Kerk screws with zero backlash nuts.

http://www.timocues.com/CNC/Inlay machine1.wmv

I haven’t really got out of the box with it till now, just simple diamonds and darts. I did make an attempt at Arnot’s shark but mine looks more like a porpoise. You can call whatever you want if you buy it.

When I did cut the two “porpoises”, one tail had a slight gap. Might have been the start of my problem?

Should have checked this thread yesterday but here is what I’ve tried last night. I jogged about 18” down my X and ran the same program. Same results but maybe not as severe of an X shift. And yea Jim the bottom of the skull is also shifted like the top. I let the machine sit a few hours, then rotated the skull 90º in Autocad. Sorry I’ve been on Autocad for 20+ years so why fight it. I brought it into Bobcad and regenerated the exact same program settings. IT CUT PERFECT. Last I brought in my original Autocad skull and remade the program fresh. IT SCREWED UP AGAIN. What does this mean. I’ll tell you what this means. I have know freakin idea what this means but I can’t turn my cue 90º. :smile:

The rapid move in mach3 makes sense. The actual cutting section looks to be repeating perfectly, it is just being shifted in the X. I’ll slow that down tonight and see what I get. Also all of the geometry is radius’s by the way. No elipses anywhere.

One last thing, Alton mentioned it may be heat related. My Gecko control box has two small fans but there might be something there also. I’m noticing the second test cut seems to be worse than the first.

THANKS GUYS!

One thing you might want to try is using connectors with voltage regulators. That should take care of your temperature issues. What stepper motors are you using?


Sent from my iPhone using Tapatalk

 

[DanO]

I bought the 270oz bipolar motors from deepgroove1.com. I had issues right out of the gate with the cables so i bought the high performance motor cables from cncrouterparts.com.

 

[Guerra Cues]

I bought the 270oz bipolar motors from deepgroove1.com. I had issues right out of the gate with the cables so i bought the high performance motor cables from cncrouterparts.com.

I see. I got my motors from Keling and even the control box. Got some nema 23 387oz.
Another good source for cnc stuff is probotix. They have several different mounts for different routers.
If money allows I will try getting an accurate and quiet spindle. Won't be cheap for sure


Sent from my iPhone using Tapatalk

 

[conetip]

With some invaluable advice from a few select people on here, I built my machine last winter. THK ball slides, Ø3/4 Kerk screws with zero backlash nuts.

http://www.timocues.com/CNC/Inlay machine1.wmv

I jogged about 18” down my X and ran the same program. Same results but maybe not as severe of an X shift. And yea Jim the bottom of the skull is also shifted like the top. I let the machine sit a few hours, then rotated the skull 90º in Autocad. Sorry I’ve been on Autocad for 20+ years so why fight it. I brought it into Bobcad and regenerated the exact same program settings. IT CUT PERFECT. Last I brought in my original Autocad skull and remade the program fresh. IT SCREWED UP AGAIN. What does this mean. I’ll tell you what this means. I have know freakin idea what this means but I can’t turn my cue 90º. :smile:



We had an issue at work on Tuesday. The posted code was not correct.
When the put it through a back plot program, saw that the cutter was on a path not of the shape of the item.
It was reposted, checked and then was correct.No idea as to how the error occurred, not done it since or before.
You may be better to draw in Bobcad and see if there is any difference in the code.
On one of the machines, if we have the feedrate too high, ie quicker than the machine is capable of accelerating, it does not make round holes or the shape is not correct.The holes can be up to .15mm out of round.
Now that we know this, the parts are programed so that the machine keeps up with the path in the feedrate.
So if programed at 1200 mm/min but the actual is flutuating in x,y to 800,450 the part will be wrong. When actual programmed is 1200 and x,y is 1150,1100 the part will be correct within what can be expected in milling.The smaller the detail , the slower we have to program, so a 4mm circular hole cut with a 3 mm cutter, can only cut at 100mm/m feedrate.
At this speed, the hole when indicated is within .01mm TIR
To get better than this requires a better machine with less backlash in the system. Even though it has 0.001 mm increments, the backlash cannot be set any finer than .002-.003mm.
Mass and rigidity of the machine come into play when you are looking at precision work.
Another test is to see if you get the same results climb milling as to conventional milling. Quite often there is a difference.If you know what works best for your situation,it becomes something to work with.
There are so many things that can be causing the problem apart from the mechanical backlash.
There is a whole range of interference that can also be contributing to the problems as well.
Nice video clip.Thanks

 

[GBCues]

DanO

It sounds like you know what you're doing, but just checking - did you match up the resistor across pins 1&5 to the impedance of your motors?

If not, that could cause some abberant behavior.

Gary

 

[Mc2]

One thing you might want to try is using connectors with voltage regulators. That should take care of your temperature issues. What stepper motors are you using?


Sent from my iPhone using Tapatalk

Please refer to this thread. Paragraph 3 please!

http://forums.azbilliards.com/showthread.php?t=167988

Now then, Dano, if your cam software is causing this issue then you would be able to see it in the tool path display of Mach 3. If you zoom in on your Mach3 tool path taking great care not to rotate it, you will see if the lines are not on top of each other. If it is your cam software it maybe possible that your geometer looks good however it maybe skewed or rotated around the Y axis. If you where to look at it from a front view and not over the top, is it parallel in the X axis direction? In other words does one side of the geometer vary in the Z axis direction? Is it possible to forward me your g-code for Mach 3? I would be happy to look at it.

Jim.

 

[Guerra Cues]

Please refer to this thread. Paragraph 3 please!

http://forums.azbilliards.com/showthread.php?t=167988

Jim,
You not only gave me red rep, but you like to be the Ask The Cuemaker AZB cop now. Got to love it You might know a little bit about CNC but that does not give you the right to put down other people. You need to learn on how to be humble.
I bet you don't act like that on CNC Zone because you would have been banned by now.
People are respected there and you can get an answer to a question without having somebody like you trying to be smarter than a 5th grader.
I am learning as well since I am building my own CNC and came across several issues, that if somebody is having the same problem I can relate to and offer my suggestions.
Also, I did not break any rules of the forum. Who did I criticized?
Just do me a favor, do not quote me any more and stay out of my posts and or threads.
My pm box is not full, so if you have any questions or need to address me please do it in private.

 

[RBC]

Dano

I really think you first need to determine if it's a code problem, or a machine problem.

The easiest way to do that is to look at the tool path like Jim has suggested. You can do that in either bobcad, or Mach3. Just make sure that you look at it from different points of view to make sure you can see it good. If it's in the code, you will see those steps in the tool paths. If they tool paths all look normal, then you have a machine problem.

As for a machine problem, I would first check my speed and acceleration settings. Just drop them down quite a bit, and see what happens. My guess is that's where your problem is.

I too would be happy to look at your code, or the bobcad file for you if you like. Just send it to rbunnell@obcues.com

We've all had things like this that make us want to shoot something, or somebody. But in the end we usually just look at ourselves in the mirror, and feel like an idiot because it was pretty simple. Even the toughest problems seem simple once we have them figured out.

Good Luck!

Royce Bunnell
www.obcues.com