Hi all. It's been a while but this is probably the best place to post. I have a CNT motion systems router, about 12 years old so not brand new by any means (and moved to this location 2.5 years ago), and It's been having an issue with getting out of position without the pc-based control knowing about it (an inherent weakness when sending servo drives step and direction signals it would seem).
The problem has been going on for over a year and is quite intermittent. I have looked over the axis in question mechanically (dual motor R&P up the table or Y on this table) and recently swapped out drivers and what I thought was the offending side's motor. However, since all of those steps seemed to help but none of those steps in fact did end the problem, I have been diving into proper trouble shooting for the last few days.
Long story short, it seems that there is a power issue in some way that is affecting the step and direction signal wires, and basically adding in phantom signals that the servo drive happily obeys. I've further narrowed it down to the spindle's (HSD) cooling fan of all things. The fan had also been acting funny, to the point I had replaced it, and generally was working the way it's supposed to (ie spinning) ....but not always. It would seem that the "not always" is the likely cause of the motion problem, as unplugging the fan seems to completely rid me of the issue.
I am still not sure if it's a grounding issue of either the fan to spindle enclosure or somewhere else along the path to ground (this is either through the frame or connected to spindle ground if the spindle body is itself grounded by wire) or a short in the power line for the spindles fan. I guess I'd probably rule out the short because I think the power lines are all fused and no fuse has blown nor breaker tripped.
Unfortunately, the tech contact at CNT is away at the moment and the local automation electronics guru that comes recommended might take a while to come see me, so I'm A) looking for any advice and B) posting something that might help someone in the future as there are lots of HSD spindles out there.
I think my options to move forward are:
1) run with a separately powered fan, like an 115V computer type cooling fan for the moment.
2) rerun the power cable to the cabinet so that it bypasses a few connections that could be at fault
3) stick a chunk of sheet metal between the offending power wire and the data cables and ground that to the electrical enclosure.
Any other ideas?
From contributor je
it may be a static issue with your dust collector or just a general grounding issue.
From contributor Ma
Is the fan cable shielded? Does it physically run within close proximity of the cable assemblies in question? You may be getting some "noise" from the fan cable intermittently. At any rate, shielding the fan cable is a pretty good low cost attempt that will at least rule out the possibility.
Hope this helps.
Mark T.
From contributor Jo
I would make sure you separate the power cables from the signal cables. They can be next to each other as long as they run parallel. As soon as one needs to make a turn. It needs to cross the other cable at 90 degrees. This will avoid noise problems. It is also good to use shielded cables for the motor powers, etc. Also for any signal cables.
I would also check all the grounds and make sure they all have good connections. I would check them with a OHM meter. Also make sure you have a seperate and GOOD earth ground! Yes, it is connected to the ground form the electrical company and their system. But it is not one I trust. I would either install a separate ground rod or ground it to the building/pipe that goes into the earth. You would be surprised that this can fix 79-90% of any strange/weird electrical problems machines have.
Johann Huck
The Machine Warehouse, LLC.
704-839-6615
johann@themachinewarehouse.com
From contributor Gr
Well,
I was able to speak to Paul, the tech guy at CNT today and he suggested the fault could be along the cable somewhere, not just at the fan itself. So, decided to go the route of the 115v fan plugged into a extension cord for the moment. The airflow seems relatively similar to the 220v fan it comes with.
1) it gets the power to the fan without using the wire in question and allows the power for the fan down that wire to be stopped at the cabinet.
2) I haven't ruled out some issue with the fan itself.
3) I don't have to ruin the real fan's wiring harness to temporary hook up an alternate wire.
So far so good. If the fix is for real (as I've tried a number of things that seemed good for a day or two) then I'll get some new cables from CNT.
I think the next CNC equipment of any type I buy I will take the route recommended when buying Chinese equipment. Extra spare parts for everything! It may only make sense 10 years down the road if all is well, but it still counts. On this machine, 10 years later, the HiWin bearings are discontinued, the Teknic servo motors are discontinued (drop in replacement coming at some date perhaps) and the servo drives are "legacy".
Fortunately, CNT itself has had no issues and I've been able to pick up parts along the way as I've needed them and found out they were discontinued, so this is just general advice.....get spare parts with your machine.
On the grounding, I have not run a better ground wire (yet) but I have the wire and clamp to do so. From what I read, it must be connected to the building earth, not a local ground rod....fortunately my unit has the building ground about 15 feet from the machine. I haven't done it yet simply because my troubleshooting pointed to a more direct problem.
Article I referred to linked below. Very good as there is quite a bit of varying opinion on grounding!
From contributor Gr
And thanks everyone for replying!
From contributor ka
i dont think its"noise"
each cable from the incoder have 2 input ad each one of them have opsite signle each time
and the controler must get in each signal the pulse and the negative
i see many problem but never was noise
or the conectors or the cable and also was one time with old biesse that it was the electrick card in the controler that input the signal from the encoder
From contributor B.
Greg,
Just curious....what model CNT Motion do you have?
I have had my CNT 900 in two locations and have grounded it to a separate ground rod in both locations.....not to the ground rod for the building wiring.
My first reaction was cables. Trust Paul at CNT.......good guy and knows what he's talking about. He is pretty much the top guy at CNT....both for support and building the machines.
BH Davis
From contributor Gr
Hi guys,
Katz, the issue of noise is on the step and direction signals that go out to the servo drives from the PC control, the drive-encoder loop seemed to be fine (although that took me way too too realize). This, I realize now, is an basic weakness of any controller that sends out step and direction data without reading back the encoder data itself. The motor encoder loop is fine, but it's been getting false signals to move.
I do think that it seems odd that the break on a single phase 220V power line in one cable affected the step and direction signals in another cable....I would think that if the cable or cables were grounded properly (which I gather is the metal shielding to ground on the panel/controller side only) it should have protected the data signals. But, I haven't checked if the cable is grounded or not. I also haven't checked if the step and direction are headed to the machine at 24V and then stepped down to 5V.....I gather that is the best approach as the higher voltage is less sensitive to these EMI issues. Nor am I very knowledgeable about how EMI effects would happen.
However, spindle is staying cool, problem seems solved for the moment and parts are getting produced. After 1.5 years of "WTF is wrong" I'm happy, not least of which is because the problem was getting worse as I was troubleshooting.
Paul at CNT was great but most of the trouble shooting was done by myself. Not finished yet either as I need to convince myself the cable is at fault before buying a new cable from CNT......reason I say that is when the issue was happening I was getting servo humming when I flipped a heavy sheet over on the table, and that behavior seems like maybe a connection fault rather than along the wire. I guess a multi meter and checking continuity under various cable flex scenarios is in my future. Visually wires are all in good shape but I realize that probably means nothing.
From contributor Jo
Greg,
I am not sure when your first posted this problem. I, too, have a used CNT Motion Series 1000 router. Paul made it very clear to me how important it is to properly ground the machine. I installed a separate cable and copper grounding spike. I did not 'piggy back' onto my existing shop's ground.
Joe T.
From contributor Gr
Just a quick update to this. The 110v fan piggybacked ontop of the spindle is working fine and I'm slowwwly gaining back confidence in my machine to hold position for longer runs.
I am planning on running new, modular, data cables from PC cabinet to the drives with a new cable carrier on the side opposite the power lines, modular meaning shielded Ethernet cables. Ironically, machine has been too busy to make a lot of progress on the work. Once I have the new cables run and all connections mapped out, might try out a different control too.
I have a thread on CNC zone for this if anyone later wants to see what progress has been made. cheers!
