LinkBack URL
About LinkBacks
Did you build your own machine? If so did you ever stop to think that maybe its your design or machining/welding abilities that suck?Originally Posted by aaronbee
| | |
Bunch of tip eat'n, metal hack'n, error trip'n worthless pieces of.... baahhhh! Save yourself. Don't do it. Run away! You don't want one. You think you do, but you don't.
That's it. No question. Just sharing some wisdom.
Did you build your own machine? If so did you ever stop to think that maybe its your design or machining/welding abilities that suck?
Actually, they don't suck, they blow. That's how they eject the plasma to cut the metal. Yah yah. har har.. ;-)
Yeah. I don't like my plasma cutters (plural), I have 4 of them, two of which are on cnc machines. If I did not have them I would spend about 6 to 8 times the labor laying out and cutting parts...then the grinding afterwards. Instead I fire up the plasma....cut through any conductive metal.....with accuracy and ease. Here's a few pictures of what they can do:
Jim
1st pick.....a variety of stainless, carbon steel and aluminum. 2nd....3/16" aluminum zipline parts, 3rd......some small belt tighteners, 4th...metal art, drawn and cut in less than an hour, last 3/8" steel parts cutting while I'm doing other things in my shop. Hate these machines! (well I probably would if I had the wrong brand, or obsolete technology!)
No. I didn't build it. It was made by a company using components from CandCNC. The truth is, it's never worked reliably. I have an extremely ambiguous relationship with it. I can't hate it entirely because, overall, it makes me money. But other days I wish it would run right just long enough to cut itself into pieces so I could sweep it out the door.
Some days it runs fine enough. Other days the arc shorts out; the control box trips red every 5th part; the THC wont track the metal; it's supposed to lift to a safe height of an inch, but gets to a half inch, makes a grunting noise and moves on thinking it's at an inch high when it isn't (as in DROs read an inch, but it's at a half or quarter in reality); goes to find z home switch and error trips when it finds it (at which point the control box trips red, but mach3 doesn't get the message and keeps on running the program like nothing happened which means you've totally lost your place now and that piece of steel is toast).
I've grounded the thing. I've read volumes. Its on its own breaker. The components are all recommended distances from each other. I've read tons. I've made calls. I've sent emails.
It's just annoying. And I was super pissed with it last night. That's all.
Sure does sound to me like the z stepper lost position cuz something bound up. Or because, for some reason, the z stepper is at the rat's edge on torque, which might be because the drive electronics for it is not QUITE at the right current for the stepper.
As for the tripping out not being seen, that sounds like you have some kind of connection issue between the control electronics and the computer. That could be any of a number of things, bad connection, IO port on the computer having that one bit not functioning properly , etc etc. Has it EVER detected a tripped situation correctly?
Really sorry to hear you are having those issues. Problems are usually easy to find, INTERMITTENT problems generally SUCK to find. I can see why you'd be frustrated with it.
Also, are you using servos? or steppers?
(as in DROs read an inch, but it's at a half or quarter in reality); goes to find z home switch and error trips when it finds it (at which point the control box trips red, but mach3 doesn't get the message and keeps on running the program like nothing happened which means you've totally lost your place now and that piece of steel is toast).
I've grounded the thing. I've read volumes. Its on its own breaker. The components are all recommended distances from each other. I've read tons. I've made calls. I've sent emails.
It's just annoying. And I was super pissed with it last night. That's all.
It is frustrating. I eventually got to a place where I just accepted it and decided to deal with it. All my questions and emails to the people who built it or supplied the parts kept coming back to the same place, "there's nothing wrong with the machine; you're just screwing it up somehow". Which, honestly, I would be perfectly fine with it being all my fault if we could just identify my screw ups so I could change them. That has yet to happen, though.
Since you seem willing to give it a crack, I'll tell ya what I know.
It uses steppers. 4 of them. 2 for Y, 1 for X and 1 for Z. There's 1 home switch on the floating head.
Sometimes Mach3 will recognize errors and stop the program. For instance, if the torch registers an error (grounds out, loses air pressure, fails to fire, etc.) it will stop and wait for that to be cleared. There are cases where the control box error trips red and Mach3 will stop the program, but other cases where it won't.
By far, the most common problem is associated with Z drive. The grunting noise I mention is the same sound it would make if you bottomed-out the machine screw drive. It's the sound of a motor coming to a dead stop and overloading. But when it does it, the machine screw is not bottomed out. It's not gummed up. Not that I can see. It moves free enough the rest of the time.
The most common occurrence is this; the torch will go to cycle through finding the Z home switch, an error will occur which trips the control box red, Mach3 does not get that message, the program continues to run and you can watch the Z DRO count off for eternity. As far as Mach3 is concerned, it will be at -34.0000 inches and counting forever if I let it go. All it knows is its waiting on a signal from a switch that's never gonna get tripped because the machine isn't running anymore.
That's the most common.
The other one is where Z makes the same grunting noise, but doesn't error trip the control box. For instance, say Z should lift 1 inch at the end of a cut. It goes to do that, but may only make it a 1/4 inch before the drive overloads (or whatever its doing) and stops even though Mach3 doesn't get the message. Mach3 now believes it's 1 full inch above the material when it is actually only 1/4. It moves to the next cut where Mach3 lowers the torch to what it believes to be 0.1800" above the material, but is in fact -0.5700" below the material - which, given the inherent behavior of the floating head, means the torch is just sitting on the material. Of course it tries to fire and either grounds out the torch (if I'm lucky) or drags the tip across the metal and ruins it ($5 bucks gone). Of course, the other half the time, when it only lifts a 1/4 inch and thinks it's at an inch, it can hang on the metal and cause a crash, but that's the rare occasion.
It doesn't always seem to be associated with that main error scenario, but I could be wrong. Sometimes it quits in the middle of a cut and not while trying to find the Z home switch, but since the THC is always driving the Z motor it could all be centered around Z and just just doesn't seem that way to me when it happens.
The cases where Mach3 gets the message that the control box tripped red, Mach3 will stop, the torch will continue to fire and burn a hole in one spot, but at least the DROs and program stops so your don't lose your numbers and position.
The torch is a Cutmaster A40. The THC box has MP3000 on it. The computer, control box, and THC are plugged into a wall socket. The torch is on it's own 60 amp breaker that comes out of the main box to the back of the shop. The table has an 8' copper coated ground rod driven through the concrete and into the ground about 4" away from the table leg. These are the things I know.
If it's the motor, the home switch, some way I have the wires crossed... I don't know.
Lastly, thanks for offering some kind of help. I wasn't expecting any. I was just pissed and ranting on the internets, but I'll take what I can get. ;-)
Oh, and one other thing, if it matters, of all the steeper motors, Z is always the hottest one. I can put my hand on them to feel the temperature, and Z will always be noticeably hotter than the rest. Might not mean anything, but just an observation I've made.
Well, so far it SOUNDS like it' s a machine problem and not something procedural , but we'll get to that if nothing really obvious comes up.
Mkay, so that all makes sense. There is some kind of sensor output from the plasma cutter that goes tooooo.. the breakout box, Im assuming, or the control system, I'm assuming and that gets passed onto the PC when the control system sees the error signal from the plasma cutter. SO, *ANY* time the plasma cutter detects an error of some kind, it seems to stop the mach3 program? If so, that helps to eliminate it as a problem source.
Okay, so, I'm going to be from missouri and take on a show me attitude, not to be insulting but to make sure you have cleared various things as being some kind of problem. (After all, I *AM* from missouri.
It sounds like the motor is stalling. Motors stall for any of a number of reasons,
1) The mechanism they are driving gets stuck.
You can check that at the instant it grunts by stopping the machine. The estop or just remove power right when that happens and try to move the z axis mechanism by turning the stepper manually and see if it is binding.
2) the control electronics some how fails.
In your case, it would be intermittent and would be something you'd have to put an oscilloscope onto and watch the signals in real time to see if the drive to the stepper is not working right or not.
3) The stepper is undersized for the thing it's driving.
This doesn't sound likely in your case since you appear to have gotten a packaged setup, so everything is likely sized correctly. AND it would do this a LOT more often than what you are seeing. (How often *IS* it doing this, anyways?)
4) The power supply to the stepper gets interrupted somehow.
Bad connections? Power supply overheating? some kind of shortcircuit that's intermittent happening? ALl not likely in this case, but worth mentioning.
5) the connection to the stepper is intermittent.
The connectors could have wires that are not quite making a good contact and the movement of the gantry or z axis is causing it to work loose, moving things around might be bringing it back, but osmething like that would get worse over time. So I would also think this is not the case.
6) The control electronics could be overheating.
THIS might be a possibility. What kind of controllers are you using?
7) The control electronics could be faulting on the final drive output (ie, teh mosfets not quite fully turning on or not properly driving.)
8) The control electronics might not be supplying sufficient drive current, or might be supplying too much and over currenting the stepper. Or supplying excessive voltage to the stepper.
THIS is interesting... It's as if there's an error situation that the controller is seeing, but is not actually getting back to mach 3.
And man, it REALLY sounds like your z axis is the entire source of the problem.
An experiment you might do is this..
Just stand the head in one place, don't even move it in the x and y directions, and just start cycling the head up and down. Make teh z axis work, have it go through a touch and lift operation over and over and over and see if it stalls the stepper motor out.
Also, does it do this right away? Or after it has been cutting for a while?
Does the system realiably *NOT* see the error when the z axis does not hit the home switch properly?
Oh, and what kind of z axis set up is it? Can you take pictures of it and post them? I want to see if there is some way the moving part that the z axis stepper controls can bind. Or to suggest things to check on it when it does bind. Also, this might be really stupid of me to ask, but do you consider yourself mechanically inclined so that you'd be able to look at something and go "OH! PFff, no WONDER! That 'll bind in NO time!" (I want to help, but I don't want to be condescending or insulting!!)
Right, in this case I'm not surprised. See, the control system can only tell if something bad happened to steppers if they are looking for a home switch to be hit and it doesn't happen. The bad part about steppers is that the control system expects a stepper to step when it's told to, if it doesn't, the control system has *NO* fn' clue that the stepper stalled out. So this particular scenario doesn't surprise me at all, and again, points to the z axis as the source of all the problems.
Whenit does this and quits in the middle of a cut, is it because the system is errored out? In other words, has something told the control system that an error occured? If so, it would likely be the THC telling the control system that the tip is malpositioned. Does the THC have an error indicator of it's own?
Ah, looks like a CANDCNC mp3000 tip height controller. Did you get the control system from CandCNC as well?
!!!! OHhhhhhh... Yeah, that DOES matter, actually. *HOW* hot does it get? Does it , like, burn your finger? Steppers can get up to 100C, you know, boiling point of water, but they really really really don't like that. (I might be misremembering here, but I've been doing a lot of reading lately as I'm about to start my own cnc plasma cutter build) They sometimes get as hot as 85C, which is still pretty damn hot. And is it a round bodied stepper? or a square bodied stepper? Do you know what make/model of stepper it is? And what is the controller that is controlling it? Ie, the stepper driver for it, what kind is it?
Also, I know I'm asking a ****on of questions, but it's all aimed at the same thing, trying to find what section of the system to focus on and to gain an understanding of what you have going on there to offer suggestions on solving it.
Where in Oklahoma City are you located? I'm in Tulsa.
Last edited by tulsaturbo; 05-19-2011 at 05:35 AM.
My Blog
http://www.needfulthings.net/tulsaturbo/myblog/index.php
You mention several times not wanting to offend me by asking simple and obvious questions; and I assure you, I take no offense. I do not have an overly sensitive ego. I've missed the obvious more than once in my life and I don't feel threatened by people who might know more about a subject than I do.
That said, let's carry on.
Excellent suggestion. I am so gonna try this. If nothing else it will give me a controlled way take a video of it with my phone while it screws off.
Hmm... when and where it will start doing it is seemingly random. On a given sheet of metal (4'X8') with, say, 30 parts laid out, the Z motor might make that grunting noise and lose its height position 5 to 10 times. Most of the time it won't error out or trip the control box red. It probably trips the control box once every other sheet. But sometimes it can do it 5 times a sheet. Depends on its mood that day and how badly it really wants to piss me off. I don't see that it waits to do it until its been running for awhile.Also, does it do this right away? Or after it has been cutting for a while?
N0001 G00 Z0.3750Does the system realiably *NOT* see the error when the z axis does not hit the home switch properly?
N0002 X1.9048 Y0.5976
N0003 Z0.1800
N0004 G28.1 Z0.12
N0005 G92 Z0.0
N0006 G00 Z0.4900
N0007 G92 Z0.0
N0008 G00 Z0.1800
N0009 M03
The control box error trips red sometime between line N0002 and N0003. It gets to position X1.9048 Y0.5976. The program keeps running past line N0003 because Mach3 has line N0004 highlighted and the Z DRO is counting off into negative infinity waiting on a signal from the switch on the floating head that it's never going to find because the machine has stopped running.
This is the most common situation. When this one happens, I can usually stop the program, reset the control box, manually bump the torch down until it triggers the z home switch, and then go ahead and hit run again and it will take off.
Now it won't always do it there. Sometimes it might error out on line N0005 or greater, in which case Mach3 will keep running the program until it finds another "G28.1 Z0.12". If it does this, your position is totally lost and that piece of steel is screwed. Rip up another $10 bucks and throw it away.
Here's a video I took of it about a year ago (yeah, it's been doing it for a long time).Oh, and what kind of z axis set up is it? Can you take pictures of it and post them?
If you watch that you can see that the torch stopped, the control box light is red, and Mach3 is still running the program like nothing happened.
I took the engine/tranny/t-case out of a '78 Ford F150 and put it in a '97 Jeep Wrangler once... and made it work. I like to think I'm ok with a wrench. ;-)do you consider yourself mechanically inclined so that you'd be able to look at something and go "OH! PFff, no WONDER! That 'll bind in NO time!"
The only feedback I get is, (a) the machine ain't moving, and (b) the red light on the control box is on instead of the green one.Whenit does this and quits in the middle of a cut, is it because the system is errored out? In other words, has something told the control system that an error occured?
Yes, ultimately it all came from them. I did not build it or put it together. The people I bought it from got their parts from CandCNC.Ah, looks like a CANDCNC mp3000 tip height controller. Did you get the control system from CandCNC as well?
Lastly, the Z motor never gets scorching hot or anything, it's just noticeably hotter than the others. and that may not mean a damn thing. It probably works harder during a cut than the other 3. I cut a lot of light gauge metal (22 gauge, 18 gauge) which means the material warps easy under any heat, especially when I slow it down around corners (like 100 ipm); meaning the THC is always running the z motor no matter what X and Y might be doing. In other words, Z is always running where X and Y may take turns and have a greater chance to cool down.
Thanks for the questions and the suggestion of isolating the z motor. I'll be trying that later today.
I'm pretty new to this stuff, but one issue I had with my Z axis (that was easily corrected) sounds similar to ONE of your problems.
The THC sends signals to the Z stepper fast enough that it doesn't have time to apply an acceleration curve to the motor like it would if you're cycling it up/down--so you may well be able to move it up and down all day without losing position, but as soon as you run a program it'll start losing steps.
In my case, if I had a long cut going on I could watch the Z DRO and it would consistently increase throughout the cut, so at the end of the cut the computer thinks the Z is much higher than it really is. Then it would drag the torch over to the next pierce, or if it got above the material it would move to the next pierce and rapid down smashing into the steel. If it's in its rapid-down move, it won't sense the microswitch--it needs to be in it's slow "find the home" move before it'll recognize the signal.
In my case, it was a simple fix of slowing down the speed of the Z, as well as slowing down the THC Rate. It was stalling and losing steps on the "up" moves, but not the "down".
I can't comment on the control box tripping red though, that hasn't happened to me.
Excellent suggestion. I am so gonna try this. If nothing else it will give me a controlled way take a video of it with my phone while it screws off.
Yeah, It really really really sounds like there is something binding up on the z axis. That g code that's throwing the error is a set reference point command when it's a g28.1 and is a rapid traverse (return to) when it's a g28 , so I'm not sure if a rapid is part of the set reference point on that command, if so, it means the z axis is trying to do a rapid and might be stalling at that time, one of the things you MIGHT try, also, is to get into the parameters for the z axis and lower the rapid traverse rate for it. When steppers run faster, their torque drops off dramatically.
Hmm... when and where it will start doing it is seemingly random. On a given sheet of metal (4'X8') with, say, 30 parts laid out, the Z motor might make that grunting noise and lose its height position 5 to 10 times. Most of the time it won't error out or trip the control box red. It probably trips the control box once every other sheet. But sometimes it can do it 5 times a sheet. Depends on its mood that day and how badly it really wants to piss me off. I don't see that it waits to do it until its been running for awhile.
Yeah, that's just WAY too often, that grunt is the motor stalling out. The easiest sleaziest thing you can do is try and slow it down if it's a matter of moving too fast, but it doesn't LOOK like it's moving all THAT fast, so there might likely be something else going on with it, like the drive current being either too high or too low. The CandCNC boxes use their own drive electronics. I don't THINK he uses Gecko boards in there, at any rate, getting the users manual for that controller might be a good idea to see if there is a way to set the current. Then it'll be a matter of finding out what the right current is and then verifying that the control box is set up with the right drive current. But so far, nothing you have told me points to anything other than the z axis being the problem, once it fails, the cascade of events afterwards seems to be only a result of the z axis forkin' up on yah.
N0001 G00 Z0.3750
N0002 X1.9048 Y0.5976
N0003 Z0.1800
N0004 G28.1 Z0.12 <-set reference point, but dunno if this also causes a rapid traverse. And a rapid on an axis that is near sstalling will likely stall it
N0005 G92 Z0.0
N0006 G00 Z0.4900
N0007 G92 Z0.0
N0008 G00 Z0.1800
N0009 M03
The control box error trips red sometime between line N0002 and N0003. It gets to position X1.9048 Y0.5976. The program keeps running past line N0003 because Mach3 has line N0004 highlighted and the Z DRO is counting off into negative infinity waiting on a signal from the switch on the floating head that it's never going to find because the machine has stopped running.
Right, which means it tried to move the z axis and then it tried to set that as a reference point. So, the position setting command, z0.blah blah probably just ran, the control does not know that the motor did not get to where it was supposed to or stalled, then it tries to set the reference point. Someone else more knowlegable in gcodes would have to step in here, but if at THAT point it is looking for the home switch and doesn't find it after a 'reasonable' amount of time, the control hardware would know, so I'm thinking it HAS to be looking for the home switch right there at n0004. Though I'm confused why it would be a 'set reference' unless set reference means, find home location, and then SAY that it's .12 inches off of the work peice, which would be the reference. i'm guessing that that's what that means (cuz I'm too damn lazy to go look it up on line, aint I a *****?
So the end result is, mach is driving the stepper (or trying to) and indicates it's wandering off to taiwan while the physical system just hangs. So it's either got a stalled stepper, or a bound slide mechanism. Oh, another thing, i can not tell for sure from teh video, but it looks like you have a slide rail system there. I have had a few of those, over time, bind up on me because of dirt and grit getting into the circulating ball mechanism. It might be worth a shot to take that whole head apart and see if they move freely by moving them with your fingers and check. Also looking for nuts or bolts that are not properly secured and perhaps floating behind the mounting plate which might occasionally bind you up would be a good thing too. It *HAS* to be something that's stopping the mechanism from moving, so far I've only focused on the complex parts, the stepper, drive electronics yadda, but it could be something stupidly simple as a bolt that worked loose and is dragging behind the plate that sometimes cocks and binds the mechanism, but as soon as you change direction, it wallows back away from the plate and lets things move again. Most especially since the amount of travel there is very short.
This is the most common situation. When this one happens, I can usually stop the program, reset the control box, manually bump the torch down until it triggers the z home switch, and then go ahead and hit run again and it will take off.
Now it won't always do it there. Sometimes it might error out on line N0005 or greater, in which case Mach3 will keep running the program until it finds another "G28.1 Z0.12". If it does this, your position is totally lost and that piece of steel is screwed. Rip up another $10 bucks and throw it away.
Here's a video I took of it about a year ago (yeah, it's been doing it for a long time).
If you watch that you can see that the torch stopped, the control box light is red, and Mach3 is still running the program like nothing happened.
I took the engine/tranny/t-case out of a '78 Ford F150 and put it in a '97 Jeep Wrangler once... and made it work. I like to think I'm ok with a wrench. ;-)
Okay, that qualifies.
The only feedback I get is, (a) the machine ain't moving, and (b) the red light on the control box is on instead of the green one.
Yes, ultimately it all came from them. I did not build it or put it together. The people I bought it from got their parts from CandCNC.
Who was the manufacturer of the gantry itself? Or, I should say, the plasma cutting table?
Lastly, the Z motor never gets scorching hot or anything, it's just noticeably hotter than the others. and that may not mean a damn thing. It probably works harder during a cut than the other 3. I cut a lot of light gauge metal (22 gauge, 18 gauge) which means the material warps easy under any heat, especially when I slow it down around corners (like 100 ipm); meaning the THC is always running the z motor no matter what X and Y might be doing. In other words, Z is always running where X and Y may take turns and have a greater chance to cool down.
Okay, not what I was thinking then, but, stalling a stepper would likely heat it up a bit. Don't know for sure, electronics isn't my major focus and there are nuances that I just don't know and have to research. If it was HOT HOT like OW **** 911! 911! then yeah, I'd say that was pointing to control electronics or stepper itself as the problem. But from what I'm hearing, I'm leaning towards a binding issue of some kind there in the z axis.
Thanks for the questions and the suggestion of isolating the z motor. I'll be trying that later today.
Posting Permissions
| | |
