[CPK_Jim]

Hi all,

I have a g320x drive that is going into fault condition occasionally for some reason. It does not seem to matter if the machine is at idle or being used. It will drop out just sittn there or while running. Seems to be super random and lately its been doing it like 6 times a day its dangerous and annoying.

Anyone have any ideas?

All 3 axis on my machine run the same motors and drives. I'm running at 24v. I have swapped drives, motor, cables, power supplies. Yet the problem remains the same. I even messed with the tuning on the drives. Ive unhooked the step direction inputs, still mysteriously faults.

Could this be some sort of noise interference?

Ive been running this machine for 3 yrs with no problems at all. Recently i move the electrical panel from being mounted to the machine to being mounted on the wall. At this time i rearranged the layout of the components and swapped out the breakout board for one from cnc4pc. When I did this rearranging i forgot to mark the drives as to which axis the came from and cant remember if i had a hard time tuning them originally or not. I also swapped the spindle motor for a 3 phase hooked to a vfd, but ruled this out by not powering up the vfd.

Im out of ideas,

Jim

[erikjgreen]

My first suggestion would be to re-verify all your new and modified wiring. Double check that every wire goes to/comes from where you think it does, and also check voltages/states where possible.

While you're doing that, look at what generates noise and where. VFD drives are electrically very noisy, and can generate spikes and phantom signals in nearby unshielded wiring that can cause encoder feedback and drive command signals to go wacky, especially on unshielded single ended connections.

If you look at your VFD's install manual, it probably recommends a line reactor on the input side and possibly a load reactor on output, plus an RFI/EMI filter somewhere in the middle.

Lots of people put the VFD in a dedicated metal enclosure as a faraday shield, and put a fan in there for cooling.

Erik

[will gilmore]

Do the drives all fault at once? Have all three drives faulted?

[CPK_Jim]

I have checked all connections and wiring multiple times.

I ruled out interference from the vfd by not powering it up. the 320x still faults in the same manner.

I have uncoupled the drives so that they dont all fault when 1 goes to fault. The drive im having the trouble with is for the Z axis. But I have had each one of the others fault in the same manner a couple times but not at the same time as the Z drive. and no where near as frequent.

I am running a separate power supply for each axis. I thought maybe the supply was going bad so i hooked the z drive to the same power supply as one of the other axis so i had 2 drive on the one supply the z faulted and the other did not. So im ruling out insufficient power

One thing i did think about today, is grounding. Before with the control panel being attached to the machine the machine was grounded because the control panel is grounded. Now that I have moved the panel, the machine itself is not grounded. So I grounded the machine to the electrical conduit on the wall using a piece of wire. So far no faults since doing so and I have been running it for about 6 hrs since then. I wont say its fixed though, not until 3 days goes by without it faulting.

I'll keep ya posted as to what happens but im doubting that this fixed it. So again I need ideas, if anyone has any others.

Jim

[CPK_Jim]

Well just as I suspected the problem remains the same. I did get a good 10hrs of running it without any trouble, so maybe im on the right track. Or just lucky for a day....

This is driving me nuts,

Jim

[awerby]

on the axis that's faulting? If you've got a fairly long run (say more than 5 ft.) then a differential might be called for. We had all sorts of intermittent problems with our Z axis until we broke down and spent $100 on the differential kit from US Digital, which comes with a sender. a receiver, and a shielded cable. Suddenly it started working perfectly!

On the other hand, this does sound like a grounding issue, especially since it worked okay before you changed your wiring. The fact that it's working better, but not perfectly, might mean that the new ground is good but not great. You might try one of those copper grounding stakes that you drive into the actual ground, rather than relying on the conduit ground.

Andrew Werby
ComputerSculpture.com — Home Page for Discount Hardware & Software

[CPK_Jim]

the wires are about 4-5ft. I tried a short wire about 2 ft and still had the same problem.

Im thinking it may be the ground, because its better now then it was. And yes the ground I used is not that great, its just a piece of wire with an aligator clip at each end. So i am going to try a better grounding method probably 2 good ground cables from the machine to ground. And yes i will be installing a good ground rod into the ground. I was thinking about this and realized that there is no ground rod for the building that i am in so it has to ground through the wiring (350ft worth) to the ground rod at the house. I'm still running on my temporary power due to the fact that the power company wants $8,000 just to get the power to my building, and that dont include any permits or anything on my end.

But I will keep the differential idea in mind if i still have issues.

Ill keep ya posted as to what i figure out, but keep those ideas coming.

Jim

[bill south]

Jim;
I would suggest that you revert back to the old BOB and see if the problem goes away. I've received boards from CNC4PC that did have bad channels and you may have one with an intermittant component.
Just a suggestion!
Which break out board did you install? I've had to remove a C11T because I never could get it to funcation as designed. But, this is just my experience!
Bill

[CPK_Jim]

well the problem still exists, even after puting in a ground rod and good ground cable. So it was just a fluke that it worked ok for a couple days.

Bill
Im about ready to put everything back the way it was. I guess the old saying if it aint broke dont fix it holds real true here. I installed the C35 i thought it make for neater wiring, but im on the fence as to weather i like it or not. And not that i think about it there is a bad output pin on it. My old breakout board was as basic as they get so id rather not go back to that but i may consider buying another of a different type and try that, since i need another for my lathe project anyways. any suggestions wich is a good one?

Jim

[LUCKY13]

It sounds like you have swapped everything around except for the wiring, meaning after you made the changeover. So if you have used the same wiring on the Z axis each time you changed a part ( like swapped drives, motor ect,ect) then try different wiring harness for the Z axis.


Also try different ports on the BOB, like run the Z axis where you had the X or Y hooked to the BOB (both to and from the BOB).



And just because the problem showed itself when you made a change doesnt mean something else didnt happen so stay open minded about it.


What about the connections, or ports them self between the BOB and the computor?


If you do go as far as changing BOB's again I would suggest the PMDX products. CNC4PC does seem to have a board here and there that leave the shop with a problem. WIth that said they are good about backing there products. Also I dont know which product doesnt work good with others but it seems I read here and there about X BOB didnt work well with X setup in the CNC4PC line.


Problems like this can be hard to track down sometimes. Just dont let it get you flustered and try to work systematically and keep notes about what you try.


And like I said earlier be open minded. A machanical problem could even just happened to start at the same time. Have you checked the pully and belts on the Z axis? Could the wire for the encoder have a intermiting break in it? Does the Axis try to bind at any one point? There is a ton of little things that could be causing the problem so give everything thought. Wires along being routed differently may show a open or shorted curcuit just by being routed in a different dirrection. Might have been good before but with different stress from different routing could show the problem up.




Good luck, Jess

[CPK_Jim]

I have swapped everything including wiring and even tried going back to the old B.O.B. the problem remains the same.

Its not mechanical because i removed the belt from the motor so there was not load at all on the motor and yet it still mysteriously faults.

I am running old encoders that I had to put a pull up resistor on the encoder channels so that the drive senses them. I checked them out and happend to notice that i dont see any kind of filter capacitor, should there be one?

Jim

[morestuff]

If your encoders are the HEDS type, then Gecko does
recommend adding a capacitor across the power pins.

Look on Geckodrive.com under Support>ApplicationNotes>ServoDrives

There is an application note showing how to tuck the
cap under the pins.

Regards,
Steve Stallings
www.PMDX.com

[CPK_Jim]

Thanks, a ton. I think the problem is that I left out the capacitors. It seems i built the machine right before that support document came out. So I have added a cap to the z axis so far so good. And im off to the shack to get 2 more caps for the other to axis.

I think maybe why it worked ok for the 3 yrs is that with the electrical panel being mounted on the machine the encoder wires where super short. When i move the panel and lengthened the wires the problem surfaced? IDK just a guess.


Jim

[CPK_Jim]

Ok so,

The end result of my problem ends up being that i needed capacitors on my encoders. All is well now no faults at all since adding them.

The 320x drives are awesome pieces of equipment i would recommend them to anyone. I will be ordering at least 2 more real soon for my lathe project!! For me servos are the way to go if there is any problem at all you know it!! I don't like the uncertainty of steppers for a machine like a mill or lathe. I run steppers on my plasma, and will probably order a 540 for that too.....

Thanks a ton to all that looked.

Jim