[rockaukum]

Hello all,
I am stumped with this one. It seems I am missing steps in the X axis. I am running a v-carve application which pockets the background and leaves the letters raised. When the machine goes for its second pass it is off by 3/32" to the left (towards zero) I had a similar issue with the Z axis going too deep and did some reading here on this site and found that changing (lowering) the acceleration and or feed speed may work. It did on the Z, however now I am having issues on the X. I have run the cut several times and each time I have change either the feed speed or the acceleration and have had no luck resolving this. It seems to be consistent each time so I'm hoping one of you great minds are able to assist me with this. I have posted several photos of the results and on the cuts are the speed and acceleration noted for reference. All photos are taken from the end of the table where zero is to the left. The photo of the spindle shows the original zero in pencil and after the run, I hit "go to zero" and that is where the spindle ended up. I also did some simple pockets and had no issues show up. I am running Mach3, g-code is created in Sheetcam. Also, When I shut down the machine and the power was cut to the motors, the gantry jumped. I did not notice which way but it appeared that it was under tension for some reason.... If you need more information, please let me know and I'll do my best to get it to ya.
SO, if anyone had any suggestions to help me out, that would be great!
Thanks very much!
Rockaukum

[awerby]

It could be lost steps, or a mechanical malfunction of some sort. Check the couplings on the axis that's having problems. I usually use a Sharpie to mark across the shaft and coupling, then run it a little and check to see if the line's offset. It could also be a pulley slipping, if you're using them. But if not, reduce the acceleration value by about half and see if that cures the lost step problem.

[rockaukum]

Thanks awerby,
Here is what I did today. After reading many posts with similar issues I checked several things and tried several more...
1. Checked the "Shuttle Acceleration" originally set to .25 so I left it alone
2. Checked the backlash. All set to 0 with speed % set a 20 and not enabled
3. Ran a "Drivertest" and it showed a pulse per second of 25353 / 4 and a pulse rating of "System Excellent"
4.Checked for the pulse width in motor tuning and could not find it.... Read to try to set it at 15?
5. Decreased the acceleration by half. I went all the way down to 10 on the X and Y Ran the pocket operation and came up with the same result.
6. Marked the motor shaft and the gear and ran the program. No movement noted....
7. Ran a different pocket with the acceleration set to 35 for both the Z and Y. The results were the same.
8. Ran the pocket operation with "circles" instead of "Zigzag" and the results were the same.
I took photos of the results.
If this should be posted in a different forum, please advise.
Thanks in advance of any and all help....
Rockaukum

[ger21]

Is your spindle trammed? (perpendicular to the table in all directions?) If it wasn't so drastic, I'd say it's a tram issue. But the spindle would have to be really crooked to get results like that.

Because it only happens when the depth changes, it's probably not the accel, or lost steps. My guess is mechanical. Is the Z axis hitting something when it goes deeper, causing the tool to move to the side?

[rockaukum]

Thanks Gerry,
Yes it is trammed and calibrated (I think form a suggestion from you over a moth ago). Re-trammed when I replaced the router with a vfd and spindle. There is about eight inches of travel on the Z and for these cuts, it is not hitting anything, such as bottoming out or at the end of the lateral travel. I don't hear any audible of the motors missing. What are your thoughts of running the program and stopping it at the half way mark and then hitting the "go to zero" to see if it lines up? I guess that will tell me if it happens during the first half or when it starts over? I'm new to this and therefor it I'm at a total loss. The machine was working good with the occasional over plunge on the Z. Now this?
Thanks again,
rockaukum

[LeeWay]

What kind of machine is it? How is it constructed?
Post a few overall pictures of the machine.

[A_Camera]

Could it be that you are driving the steppers with too low current?

In fact, I had recently seen similar problems with my own machine and realised that I set the DIP switches on my drivers wrong, so the current was too low. The result were fine but occasionally when the Z was pulled up at full speed the stepper stalled during deceleration and when this happened of course, the consequences were dramatic and not amusing at all. It took me a while to realize the cause, I suspected lose wires, broken/faulty driver, screw lose, alignment issue, too high acceleration, too high speed, and a lot of other things, but in the end it was just a simple flipping of a DIP switch what was needed. Strange that my machine worked well for quite a while not experienced any problems at all, and even when I saw this for the first time all the other axes were fine, as well as movement of Z down. Now, after I set the current higher I could increase speed and acceleration back to what I had before on Z, which is 8000mm/min and 700mm/s² as well as increased the X and Y from 8000mm/min to 9000mm/min. Run a half an hour high speed tests several times and it definitely confirmed that it works now again.

Anyway, check your driver configuration and make sure the current to the steppers is what it supposed to be.

[rockaukum]

Thanks guys.
Here are a couple pics of the machine. I purchased it second hand and it was built by a guy who's company name is Lightening CNC. After I purchased it, I tried to contact the builder and found horror story after horror story and had no luck with the contact. I had to reconfigure the setting to "Mill" as it was set up as a plasma on Mach3. It was running good for a while then the router crapped out. So I went with a spindle (2.2kw) and vfd. After that was installed and trammed the machine was working fine as well. Then I decided to connect the shields for the motors and the limits (Spindle had ground to the body and the shield is connected at the spindle). Next I went tot he shop with my son to show him the machine and this is when the issue started to show. I tried to contact Sainsmart (?) to get info on the bob but I was unable to open each link they sent,. I really have no idea how to check the current and even what current is required for the motors. So if you can point me in the right direction, I'll get going on that as well. It has a 24v power supply in the control box. Again, if there is any other information needed, please advise and I'll do my best to provide it.
Thanks much!
Oh and as to the last comment with my last post... When the power was shut off to the machine, the right side of the gantry appeared to move?
Attachment 366564Attachment 366566

[LeeWay]

24 volts will be ridiculously under powered for a machine that size. I have had several similar BOB's go bad on me. Usually just one axis. Swap pin numbers and it would work for awhile and then another would go dead.
Don't know anything about the motors either, but I would suggest investing a little in some better electronics. Those are scraping the bottom of the EBAY barrel.

[ger21]

So it started after you grounded the spindle?
Then my guess is some electrical noise issues may be causing the motors to stall. Try running the same program with the spindle off, above the table, and see if it returns to zero. Then do the same with the spindle on.

[A_Camera]

I don't think you need to worry about the BoB, but I'd worry about the 24V and those drivers. They look like based on the crappy Toshiba chip, the TB6560. Those are nothing to have. There is also no way anyone can help you with the settings unless we know much more, like what sort of motors you have, and how to set those DIP switches on those drivers.

BTW, I have two similar BoBs and they work fine. Of course, anything can be broken, but I really think the BoB is the last problem.

[LeeWay]

I will quantify what I mentioned on the BOBs. They were the C10 versions from Keling.
They were also running Gecko 203V's with a 72 VDC PS. Some on Nema 34 motors. Some with 3 amp. Some with 6 amp.
I was most likely pushing the design limits on the BOB pretty hard.

[rockaukum]

Ger21,
No, it was operating just fine after the instal of the spindle and vfd. I had been reading a lot and came to the realization that ALL cables should have the shields connected and decided to connect up the shields and attach them to the case of the control box. It was after that that the issues started. Prior I was able to do cuts without any issues. I think the longest cut was about 45 minutes with 1/8" depth cuts totaling 1 1/2" deep. Worked just fine and I was extremely happy about that as the price of the wood alone was just under $400. Today I did multiple air cuts (Spindle and vfd off) at various speeds from 100 ipm down to 40 ipm and then went to zero only to find that it was off. It appeared to be the same amount as before however I did not measure the results. When the vfd and spindle are operating, I do get a little "snow" on the monitor but it did not affect the cuts at the time.
LeeWay,
I see what you are saying by way of a machine that size. However, I have only been doing small cuts with the exception of when I surfaced the MDF. The motors are the NEMA 23 as posted in one of the earlier pictures. I am guessing that by the response I will be needing to change out the drivers to something better and one that will be able to handle more volts. My confusion still lies in that the machine was working prior and all of a sudden it decided to mess around (sounds like a marriage gone bad!).
I soldered a wire onto the shield wire for each of the drivers and the limit switches, then connected those wire to ring terminals and drilled a hole into the metal case, scratched off the paint and screwed the rings down. Does that sound correct?
I am thinking my next move will be to disconnect the shields and tape them up, run a cut and see what results are...
Thoughts on that please?
Also, If these drivers are a POS, what is recommended with the corresponding power supply? This is a hobby for me and the main reason for the purchase was to utilize the machine for ski building as that is my passion.
Thanks again,
ra
Another thought is that this is occurring during the rapids? Any way to decrease the rapids and run a test to confirm? When I have been doing cuts that are not multi layer deep, it was operating fine after the shields were connected. I'll do a cut like that tomorrow to confirm that it is still the case.

[ericclinedinst]

If the problem started after tying the shields together how about undoing them and seeing what happens? Seems fairly logical to me.

Also. Shields on steppers are supposed to be connected on ONE END only.

Ger21,
No, it was operating just fine after the instal of the spindle and vfd. I had been reading a lot and came to the realization that ALL cables should have the shields connected and decided to connect up the shields and attach them to the case of the control box. It was after that that the issues started. Prior I was able to do cuts without any issues. I think the longest cut was about 45 minutes with 1/8" depth cuts totaling 1 1/2" deep. Worked just fine and I was extremely happy about that as the price of the wood alone was just under $400. Today I did multiple air cuts (Spindle and vfd off) at various speeds from 100 ipm down to 40 ipm and then went to zero only to find that it was off. It appeared to be the same amount as before however I did not measure the results. When the vfd and spindle are operating, I do get a little "snow" on the monitor but it did not affect the cuts at the time.
LeeWay,
I see what you are saying by way of a machine that size. However, I have only been doing small cuts with the exception of when I surfaced the MDF. The motors are the NEMA 23 as posted in one of the earlier pictures. I am guessing that by the response I will be needing to change out the drivers to something better and one that will be able to handle more volts. My confusion still lies in that the machine was working prior and all of a sudden it decided to mess around (sounds like a marriage gone bad!).
I soldered a wire onto the shield wire for each of the drivers and the limit switches, then connected those wire to ring terminals and drilled a hole into the metal case, scratched off the paint and screwed the rings down. Does that sound correct?
I am thinking my next move will be to disconnect the shields and tape them up, run a cut and see what results are...
Thoughts on that please?
Also, If these drivers are a POS, what is recommended with the corresponding power supply? This is a hobby for me and the main reason for the purchase was to utilize the machine for ski building as that is my passion.
Thanks again,
ra
Another thought is that this is occurring during the rapids? Any way to decrease the rapids and run a test to confirm? When I have been doing cuts that are not multi layer deep, it was operating fine after the shields were connected. I'll do a cut like that tomorrow to confirm that it is still the case.

[A_Camera]

If the problem started after tying the shields together how about undoing them and seeing what happens? Seems fairly logical to me.

...because on this forum "start earth" is a buzzword. Some people believe that it solves all sorts of issues and very few actually believes and even fewer KNOWS that it can also cause problems.

Of course, it should have been the very first thing to do and test. If something worked before a "fix" and not after, then it needs to be "unfixed" as first and most logical test.

Another very serious installation error I see in that picture is the wiring. It seems that the BoB and the drivers are not connected using the screw terminals. That is definitely a potential error source and should be rewired properly. Ribbon cables are not a good idea for this sort of use.

[rockaukum]

Well life is good.... I seem to learn something every day....
Ribbon cables are no good. Please explain why? Just curious. What wire is recommended for this?
Usually going back to the last changed item is a logical process when it is followed by a problem. However in my mind I was making a change to provide a safer operating machine, thus figuring their must be some other issue(s) causing the problem.
Did my description of how the shields were connected sound correct? Also they were only connected at the control box.
Let's hope tomorrow brings about a positive result!
Thank you for the help and suggestions. One day I'll have it figured out....
ra

[ger21]

However in my mind I was making a change to provide a safer operating machine, thus figuring their must be some other issue(s) causing the problem.

As the previous post alluded to, shielding is a sort of black art. You're dealing with something that's invisible, and somewhat unknown. Connecting shielding can easily cause more problems than it solves. Unless a problem gets instantly better, you don't really know if the shielding is doing anything?

[LeeWay]

They always have trouble with the shields on Star Trek.

[rockaukum]

Thanks for sticking with me guys,
This morning I disconnected the Shields on all cables that I had previously attached. I ran the same cut and the results were a bit improved. It appears I am still missing steps on the cut as the second pass is still off. Originally it was off by 1/8" and now it is off by 1/16". Half way there!
I ran an additional test cut with several rectangles, two passes each and they turned out good. This leads me to believe the tramming is okay. I also did one more test cut of a star and had good results with that. These are much smaller cuts which makes me think that maybe it is happening during the rapids as there are more in the larger drawing. Thoughts on that?
It just occurred to me as I was previewing the post... If you look at the pocket cuts inside the "O" and the "R" you will see that the alignment is good. These are the first two actions of the cut process. Rapids???
Also if someone could address the previously stated items....
ribbon cables are bad. Just curious as to why and what wire is a good replacement? shielded (HAH!) or what?
It was mentioned that the drivers were not so good. Suggestions for a replacement along with power supply?
Thanks for the help,
ra
First cuts with shields connected
Attachment 366608
Cut with shields not attached and taped up. (80 ipm)
Attachment 366610
Sample cut (small) and these look okay? shields disconnected (80ipm)
Attachment 366612Attachment 366614

[ger21]

Are your sample cuts done in one or two passes?
What's really strange is that it only seems to be losing position when it moves down for the second pass.

Can you post the g-code that you're using?