PDA

View Full Version : X axis issue with the Triumph



Kev Williams
05-24-2016, 4:15 PM
Could use some thoughts from the experts :)

For some time now the Triumph has had a glitch, to do with the X stepper motor going a bit crazy at times when vector cutting curves or circles.

I've been assuming the X stepper is probably to blame.

However, Yolanda and engineers at Triumph have been telling me they don't believe the stepper is to blame. The had me verify tensions on all belts, the tension or lack thereof of the rail bearings and rollers, all the basics. Everything is in order. Here's a video I made just for the Triumph crew to watch--


https://www.youtube.com/watch?v=mODbcMcuIfU

Here's some facts and anomolies about this problem:

-speed doesn't seem to matter much, I can slow the speed down to half what it's doing in the video and it will still do this- but, not as often or quite as 'violent'.

-it's ONLY the X axis that does it, the Y axis always works perfectly- I spent much time watching the Y stepper and holding onto the gantry while running, it never misses a beat...

-the main 'anomoly' that puzzles me, is that in the video, the table location where it started seems to be the worst area for this to happen. It DOES happen elsewhere, but in that general vicinity- about 8" from left and about 18" from the top- seems to be where the worst of it happens. I've had all the belts off to check head and gantry motion, smooth as I could ask for, no bumps or 'catches'...

- sometimes it will do it almost always, sometimes I can't ever get it to do it...

- I've noticed that from a cold start, it hardly ever does it-- actually I can't remember it doing it from a cold start during my tests.

- The machine rasters just fine. And vectors straight lines just fine, no matter the angle. It's only on curves, but rarely on quarter-circles or less.

So the kicker is, I just installed a new stepper that Triumph reluctantly sold me yesterday. While it's much quieter when running side to side than the old one, the problem remains...

Which brings me to these things:

338045 --- not sure of what exactly they're called, but I do know they control the stepper motors so I'll call them the "XY Controllers". It's fun watching the green 'in use' LED's, and I've watched them closely in some of my testing. I've noticed that the smallest stop of either motor is noticeable by a quick flicker... Like when doing a series of circles that are connected, when moving from one circle to the next- only .004" inch- the Y LED will 'blip', as the change takes place in the 3 o'clock position. And when watching the LED's when the machine is running a circle, the lights fade out and back in very smoothly as each axis goes thru it's motion. But when the X axis starts mis-stepping, a little or violently, the X LED fades out and in as smoothly as if the axis was moving normally.

SO--- My troubleshooting has brought me to these conclusions...

New stepper didn't help-
I assume these XY controllers are fed by the main controller, and since the LED's show smooth, fluid 'motion', this indicates the data they're being fed is correct--
I've checked both controller's connections, tightened the lug screws, etc, no change, better or worse--
Therefore, since the X controller input seems to be correct, but the output is anything but--
Main controller NO
Stepper motor NO
X stepper controller -- YES..?

Before throwing money at it, I'm considering reversing the wiring on the 2 controllers to see if the Y stepper starts screwing up...

Interestingly, Triumph told me a customer in Mexico is experiencing the same issue, and they're having a hard time troubleshooting it...

Thanks in advance for the replies! :)

Dave Sheldrake
05-24-2016, 7:55 PM
That sounds like belt skip Kev, good idea to swap round the drivers though and see if it replicates on the other axis, if it does it's the driver, if not then I'd suggest belt

Keith Winter
05-24-2016, 7:56 PM
Kev I see the video, and the mess up, but it might help to know what is the pattern supposed to look like? Does it loose it's place on the bed when it happens? Are you sure the belt is supposed to be that loose?

Dave Sheldrake
05-24-2016, 7:57 PM
Kev I see the video, and the mess up, but it might help to know what is the pattern supposed to look like? Does it loose it's place on the bed when it happens? Are you sure the belt is supposed to be that loose?

That X belt does look a bit on the loose side

Kev Williams
05-25-2016, 12:15 AM
Triumph thought it was belt slip, too but it's not. Tighter belt makes it worse actually. I should've video'd the stepper, when the bad happens you can watch it trying to go both directions at once.

Also, above I said 'it rasters fine', but that's not entirely accurate-- in the video at the first I mentioned I loosened the belt, and that it helped. What was helped was a vibration that's been going on for some time now, most apparent between 450 and 500 mm/sec. Loosening the belt (couple of days ago) removed nearly all of that vibration. I've done a crapload of engraving since the vibration started, and it's never affected the engraving. But usually I've been running it at either 400 or 600mm/second to avoid the vibration. The vibration issue preceded the funky vectoring, and was the main reason I figured the stepper was bad. And back in January 2015 I made a video (link below) of the stepper doing stupid stuff while totally divorced from the gantry....

https://www.youtube.com/watch?v=9jkOy2xbfrc

(https://www.youtube.com/watch?v=9jkOy2xbfrc)Keith, as for the 'pattern', what's cutting is the holes in the graphic below. The holes are typical 30mm notched holes, although the notches are rounded. And yes, it loses it's place, very badly!
338067

Anyway, I'm not sure if these stepper motors need to be 'loaded' to run right, but mine sure didn't with no load! -But- at times when changing speeds on the fly while rastering, I could get the stepper to totally lock up, I have a video of that too, but I'll spare ya... ;)

And, even though the machine still messes up with the new stepper, it DOES run noticeably quieter when rastering and slewing in the dead zones than it did before...

Kev Williams
05-25-2016, 11:10 AM
OK, so correct me if I'm wrong here-- above pics are of the XY drivers,

This is the controller--
338095


and this is the low voltage power supply--
338096

--and unavailable for photo in the back of the machine is the main power supply..

Haven't done voltage tests on these parts yet but will get to that today hopefully. FWIW I did test the main power supply voltages when this all started, all okay...

Question on the XY drivers, do the plugs just pull out or are they locked somehow? I'd rather know first than find out the hard way! ;) Looks like a 5 minute changeover to swap the drivers...

Bill Jobe
05-25-2016, 5:51 PM
Probably no help here, but years ago I ran CNC machines. The problem you describe was referred to as a resolver problem.

That was many years ago so my post might give a few members a chuckle.

Kev Williams
05-25-2016, 6:36 PM
Well, as my bad luck and Murphy's law would have it-- My LS900 just started doing the same thing. Ruined half a sheet of 1/8" Rowmark as it went crazy cutting holes willy nillly... Pretty sure the controller is history...

Really getting frustrating. Since February I've lost TWO controllers for the IS7000 (mine and the first borrowed one), FOUR of my seven EP modules died with 3 days (in April), the Triumph and now the LS900 has lost it's mind...

I'm not far behind...

Rich Harman
05-25-2016, 6:39 PM
My stepper drivers have settings for both current and steps/revolution, do yours? If so, have you verified that they are set up correctly? It sounds like your driver is set to a low current, just enough to work most of the time, but not all the time.

My X axis drive was set to a high current level. The motor ran very hot. I lowered it and now the stepper never gets past warm and it has never lost a step.

Rich Harman
05-25-2016, 7:08 PM
Another thought, if it tends to occur in one area of the table more than the other then that points to a wiring problem. As the gantry moves up and down the table the wires are flexed in the cable carrier. You may have a short or broken wire.

Keith Winter
05-25-2016, 9:04 PM
With this new info I'm wondering if you are having power issues. Perhaps you have reached the limit of a home circuit or you have something causing irregular power?


Well, as my bad luck and Murphy's law would have it-- My LS900 just started doing the same thing. Ruined half a sheet of 1/8" Rowmark as it went crazy cutting holes willy nillly... Pretty sure the controller is history...

Really getting frustrating. Since February I've lost TWO controllers for the IS7000 (mine and the first borrowed one), FOUR of my seven EP modules died with 3 days (in April), the Triumph and now the LS900 has lost it's mind...

I'm not far behind...

Kev Williams
05-25-2016, 10:57 PM
Turns out my LS900 was (hopefully) nothing more than a Murphy's law prank. My NH rep came right over, and do you think the machine would act up? Ran perfectly, for over an hour...
It has NEVER acted anything like this before, and it did it twice (I'll post pics later) -- the kicker? It messed up a piece of 1/8" Rowmark that my customer paid overnight Fedex to get here today so I could make his parts in time for them to ship a machine out today... Just because I promised work out and was in a hurry... http://www.engraver1.com/gifs/pullinghair.gif

But I did manage to stop the bleeding in time that I still have enough material left to finish the job, and the parts will be done tomorrow courtesy of the GCC, but not today. At this point I'm supposing it's a connection issue, and later tonight I'll be cleaning up plug connections inside the machine... just like I had to do with the GCC a few months back.

Rich, I've thought about a wiring issue because of the proximity thing, it's one reason I want to swap the drivers, if the issue changes to the Y stepper, it's likely the driver--but if the issue doesn't change then it may very well be a jacked up wire, as it makes perfect sense...

Keith, I should have plenty of power, I have 200 amp service here and it's divided up pretty nicely, the last time a breaker tripped was about 8 years ago :)
We have 'good' power in this area too...

Here's a pic I sent to Triumph about 2 months ago, doing pure circles, which are about 1.75" diameter. You can see that the frequency of the jaggely's (? lol) is much too tight to be a belt-slip issue, this is the result of a stepper motor that's being fed some really bad instructions methinks... ;)
338139

Keith Winter
05-26-2016, 8:41 AM
If it's not the belt, and not the driver, and not the wiring as Rich said. One more long shot possibility...

I was having a similar issue with a laser with it jumping around minus the irregular circles you are getting. The jumping around part came down to the Trotec wasn't interpreting the illustrator file's artboard correctly. However they did have us go through multiple months of hoops not willing to admit their software/firmware could be the flaw. Among those hoops and questions one question that came up multiple times was power. I'm not sure what is in your area, but has any machine shop, heavy equipment shop, or welder shop been added that you know of? Apparently Welder's put off some real nasty effects to the power grid and cause electrical interference to sensitive equipment within a few blocks according to Trotec and also my Electrician. We ended up having the electrician put in a new line at Trotec's behest (which didn't fix the problem) and also having the Electrian measure the incoming power twice at Trotec's request (didn't find an issue). In my case it was a Trotec software/firmware issue that we have learned to work around, and I wouldn't recommend you waste the money I did running a new line and paying an electrician to measure the line. However I wonder, do you any new welders or heavy equipment shops in a nearby commercial district that you know of?

Kev Williams
05-26-2016, 1:14 PM
I LOVE BRAINSTORMING! :D (found the problem!)

First to answer Keith- the main(?) that feeds our house serves 11 homes, learned this fact during an outage awhile back. I'm the only home business of these 11 homes, and within a 3 block radius is mostly houses, I'm not aware of any other home businesses. 2 or so blocks to our south is a business 'mall', the closest is a body shop, and it's across a major street so it's likely not affecting anything. I'm the big power user around here, as my 'neighbor comparison' will attest ;)

http://www.engraver1.com/erase2/ourpower.jpg

as for the stepper problem:
After Rich brought up possible a wire issue, I was about to start pulling wires on the drivers, and thought I'd try the easy way first...

First thing I did was stack a bunch of those notched circles horizontally across the table in the 'bad' Y position, and ran them. And virtually every one of them engraved whacko!

So I reset the machine, slewed it down to the bad area, then started slewing left to right. All was well until I grabbed the wire carrier and wiggled it.. The stepper went bonkers!
And if I moved the gantry up or down about 4 inches, I could NOT make the stepper act up. Bring it back, and I could mess it up at will...

Pretty much nails down the cause! Triumph agrees... And this morning I tracked down the LS900 problem to an LPT cable switchbox.

So on the plus side, no high$$$ problems, just more niggly time-eating maintenance and repairs. I hope! http://www.engraver1.com/gifs/fingerscrossed.gif

Keith Winter
05-26-2016, 5:23 PM
Glad u found it! Good call Rich on the wiring

Rich Harman
05-26-2016, 5:37 PM
Glad u found it! Good call Rich on the wiring

Yay for me, I'm gonna get out my red sharpie and mark that on the calendar!

Kev Williams
05-26-2016, 6:33 PM
So- my new stepper came with a plug & wires but they were only half long enough. So I added a couple of feet of wire to each one, got it all ready to install... But when I get ready to unscrew the old wires from the driver, I notice red/white/blue/yellow/black/brown wires. The new plug- and the old plug- have no brown wire, they have a green wire... So I try and track down the source of the brown wire, thinking the brown and green traded places somewhere for some reason. Only other wires were the 2 limit switch wires.

So I started unraveling the wire covering, and finally found the brown and green wires spliced together. So that explained that. Seems their plug & wires were too short too, Triumph did the exact same thing I did!

And then, there it was, 'the problem'... the black wire was essentially broke in half! And with the gantry in the 'right place', the splices were in right in the tight bend of the wire carrier...
338175

So rather than muckle with all new wiring, I had my BIL hold the wires while I soldered 'em back together...
338176

ain't real pretty but they're soldered!
338177

and I didn't even forget to put the heatshrink on first! ;)
338178


I wrapped them back up and before I stuffed them back into the wire carrier, I carefully put a slight bend in all the solder joints.

I've ran several test passes all over the table and it's working perfectly, nary a hiccup! Maybe now I can actually get some money-paying WORK done!!! http://www.engraver1.com/gifs/party.gif

Ian Stewart-Koster
09-19-2016, 8:00 AM
Thanks for the analysis, Kev- sounds like you got it right - just that one spot on the bed giving the cable the amount of flex it needed to go haywire (now we know where the term Haywire comes from!)

When our m-300 machine played up, it was entirely across the bed, with the Y axis - there was no local sweet-spot nor danger area.
Meaning if it was a cable or cable/spade/plug connection, it was somewhere elsewhere, and not on a moving part or carriage.