Re: Z-axis overshooting sometimes
I wouldn’t randomly add cabinets because cabinets do little t address conducted EMI, they might not impact radiated EMI as much as you would think.
Years ago, (decades? Yikes) I was trying to debug a custom lathe with an old Bandit come controller. We would loose position in the submicrons every cycle of the machine. After realizing that this happened a in conjunction with spindle usage we decided to look into the third party spindle controllers eventually we traced the problem to a relay in the packaged controller. The controller is as a bit of a distance away from the bandit card rack and the rest of the machine control os and the I/O was isolated. Some how the spike from that relay switching each cycle screwed up position maintenance in the Bandit. A snubbing diode (a few cents per drive) ended up solving the problem after the organization suffered for weeks.
Lesson to be learned here 1. Solve noise issues at the generator. 2. Finding the cause can sometimes be extremely difficult. 3. Hand coding G-Code is very valuable in CNC maintenance as we eventually where able to minimize the G-Code until the problem went away and then created a little routine to test our fix.
Quote:
Originally Posted by
drxlcarfreak
Good idea.... When I put mine in I didn't read about it anywhere and tried to mimic other cases that I had seen that apparently didn't have issues. I am actually beginning to wonder if my old nema23 steppers were not an issue and were powerful enough, but because of the added drives in the enclosure I was getting random EMI issues with lost steps...
Re: Z-axis overshooting sometimes
So I messed with it last night. I tried doing one item at a time to see what had an impact. I tried the EMI filter first and that immediately cleared up the spindle requiring a debounce to not trip the limits. When doing this I did notice that the non shielded output of the VFD wiring ran right between the non-shielded limit switch wiring and servo wiring. So I wrapped all of the limit switches with emi shielding and copper tape, and grounded the partially shielded servo motors. I also replaced the VFD wiring with the 16/4 shielded wire from automation direct in the case and the entire length making sure that both sides of the shield were grounded to the case and the spindle. It looks like I must have grounded a wire on one of the limit switches because it is showing permanently switched, so I need to address that, but I am hopeful that this will lick my issues.
I do have a couple more questions about grounding though. I never grounded my case or the cnc itself. I am not sure the reason, but I am guessing that was what I had read in a couple threads most likely. The closest I had is a pin on the BOB on the CNC chassis and another on a touch plate. The only ground from actual panel ground in the case went directly to the VFD which is in a plastic case so there wasn't a chance for the case to be grounded thorough the VFD. Now that I added the EMI filter with it's metal baseplate I noticed that the chassis does now have continuity to ground. Would it be beneficial to ground both sides of the EMI Filter, the VFD and the earth ground all to one spot, or is this sufficient? The 60v, 12v and 5v DC rail grounds now all seem to be grounded, would I want to add them to a star ground as well, or leave them as is? Would it be beneficial to ground the CNC Chassis to the case, or would this add a potential ground loop issue?
3 Attachment(s)
Re: Z-axis overshooting sometimes
Quote:
Originally Posted by
drxlcarfreak
So I messed with it last night. I tried doing one item at a time to see what had an impact. I tried the EMI filter first and that immediately cleared up the spindle requiring a debounce to not trip the limits. When doing this I did notice that the non shielded output of the VFD wiring ran right between the non-shielded limit switch wiring and servo wiring. So I wrapped all of the limit switches with emi shielding and copper tape, and grounded the partially shielded servo motors. I also replaced the VFD wiring with the 16/4 shielded wire from automation direct in the case and the entire length making sure that both sides of the shield were grounded to the case and the spindle. It looks like I must have grounded a wire on one of the limit switches because it is showing permanently switched, so I need to address that, but I am hopeful that this will lick my issues.
I do have a couple more questions about grounding though. I never grounded my case or the cnc itself. I am not sure the reason, but I am guessing that was what I had read in a couple threads most likely. The closest I had is a pin on the BOB on the CNC chassis and another on a touch plate. The only ground from actual panel ground in the case went directly to the VFD which is in a plastic case so there wasn't a chance for the case to be grounded thorough the VFD. Now that I added the EMI filter with it's metal baseplate I noticed that the chassis does now have continuity to ground. Would it be beneficial to ground both sides of the EMI Filter, the VFD and the earth ground all to one spot, or is this sufficient? The 60v, 12v and 5v DC rail grounds now all seem to be grounded, would I want to add them to a star ground as well, or leave them as is? Would it be beneficial to ground the CNC Chassis to the case, or would this add a potential ground loop issue?
It sounds like you are almost there
You don't want anything to do with the Bob and Mains Power Grounding
As the Filter is the supply to the VFD Drive the Ground should go to the VFD Drive then the VFD Drive Ground to the Spindle Ground
The main input Power Ground to the Filter and to the Star Point Ground, for all your other Ground connections Power Supply Etc
All your equipment Grounds should all go to ( 1 ) Point ( Star Ground ) so you add a Bus or a Stud and mount all Grounds to that in to your Cabinet or a Ground Plane Plate
I don't like wrapping cables leaving the foil exposed, as that wrapping can carry EMI and spread it every where
Any of the methods attached will work for Star Point Grounding
Re: Z-axis overshooting sometimes
Quote:
Originally Posted by
mactec54
It sounds like you are almost there
You don't want anything to do with the Bob and Mains Power Grounding
As the Filter is the supply to the VFD Drive the Ground should go to the VFD Drive then the VFD Drive Ground to the Spindle Ground
The main input Power Ground to the Filter and to the Star Point Ground, for all your other Ground connections Power Supply Etc
All your equipment Grounds should all go to ( 1 ) Point ( Star Ground ) so you add a Bus or a Stud and mount all Grounds to that in to your Cabinet or a Ground Plane Plate
I don't like wrapping cables leaving the foil exposed, as that wrapping can carry EMI and spread it every where
Any of the methods attached will work for Star Point Grounding
Fair enough. Sounds like one more ground to go then!
Yeah, I wasn't a fan of it being exposed so I wrapped most of it in electrical tape just so it wouldn't accidentally short into something. Is there something better to use?
Re: Z-axis overshooting sometimes
Quote:
Originally Posted by
drxlcarfreak
Fair enough. Sounds like one more ground to go then!
Yeah, I wasn't a fan of it being exposed so I wrapped most of it in electrical tape just so it wouldn't accidentally short into something. Is there something better to use?
Other than better cable to start with, yes shorting is the main problem when using exposed foil or braid