Jobs ruined after long run

[BernieNUFC]

Hi

I try to do as much as possible before bothering people by reading/watching video etc but i am out of my depth here now.
Running a 720-Z for 8 months now slowly learning about it and using vectric vcarve pro.
The other week after running a small shield in 3d finish it got to 99% and suddenly did a large L and at a 6mm depth straight through the work.
Yesterday it ruined another piece when after about 1.5 hrs running it suddenly overrun the x axis by 20mm and carried on as tho nowt happened.
Today i started all a fresh and again 99% 40 mins in and it suddenly plunged lower and finished the last run too deep and offset by 5mm and ruined the Pine wood.
Running it at 44mm/sec .25 ball nose in pine.
The only common denominator seems to be 3dfinsh but its all simple work nothing out of the box.
Where could i start to track this fault (but i cant read g code) as i cant waste any more wood as its all for charity and it costs me.
I have only done small sorts of work until these shields and never had any problem, but i am only a hobbyist and do little work with it. I also read here and agree with a previous poster about anything wrong with leadscrews etc and it has to be shipped to Germany and man it costs, a bit of a downer, but cant see why it would be these as no backlash or owt, it just has a moment of its own and goes off on one.
Its on a dedicated PC (win 7) no screensavers or power save options and no internet connection to distract it

17968
Many thanks for any pointers Bernie

[JAZZCNC]

Hi Bernie,

Ok well you need to start by checking the machine from head to toe for any mechanical issues. Mostly likely cause would be binding or failing bearings etc.
Any of these and the motors will struggle and start losing position or worse case stall completely.
It's actually quite common for things like this to happen on long jobs that run at higher feeds where torque is starting to drop off. The combination of heat which causes binding and steppers working in speed range that can be affected by resonance can also be a factor.

If you can't find any mechanical issues and the G-code is correct then it's strong possibilty you are being affected by Mid band resonance. This happens like the name suggests when the motors are working around there mid rpm level which at 2500mtr/min then you could be approaching(depending on screw pitch).
If you are being affected by this then you have limited options. Changing the feedrate to lower or higher speed is an easy test to see if you can move the motors out the frequency range where resonance affects drives/motors. If it is resonance then you'll likely struggle to go higher before motors stall.

Next option and the one which is most likely to work is changing the micro stepping settings on the drives(If they have any). This can shift the resonance to different frequency which the drives can handle better.

Replacing the drives with better drives is the last option and one that will most likley cure the issue if resonance is your problem.

Checking all the wiring both on the machine and in control for lose connections is second thing to try after checking mechanical.

If you post the G-code by zipping it up or in txt file (don't post all lines on here) then I'll take look to see can bad code.

[BernieNUFC]

17972

Hi Jazzcnc
Thats a very comprehensive reply and very thorough thank you, i will start doing the checks you mention, in the mean time i have attached the NC file.
The machine has probably done less than 50 hours since new so i dont think there should be any parts that are ready to fail just yet, well there better not be.....:-)
I am using WinPC-NC and the settings defaulted for the 720 so i am not sure where to start checking micro stepping.
No loose connections as i have just relocated it and made everything a much better and tighter setup.

Thanks again
Bernie

[JAZZCNC]

No loose connections as i have just relocated it and made everything a much better and tighter setup.

This makes me suspicous not re-assured.!! . . . . Did it do this before Relocating.?

Also Just because machine isn't very old doesn't mean it can't hit trouble. Infact it's about the right amount of hours when things could come loose and need tightening after settling down.!

[BernieNUFC]

No it did not do it before the move but equally i did not do these files only the very short work jobs.

I can understand about the bedding in fact last week after the first incident i gave it a good checkover just in case but could not find what might have caused the initial fail

[JAZZCNC]

quick look at the code shows no issues so not that.

[BernieNUFC]

Checked every Allen key bolt, every screw and removed covers to check wiring etc, went through every connector to controller and then to PC, everything is tight now, only one screw holding an earth wire was loose enough to take a nip up.

[JAZZCNC]

Checked every Allen key bolt, every screw and removed covers to check wiring etc, went through every connector to controller and then to PC, everything is tight now, only one screw holding an earth wire was loose enough to take a nip up.

Have you checked for any axes binding. This is mostly likely cause.

After checking this I suggest you try running the same G-code at different feedrates to test if resonance is the cause. If it runs fine at lower feed rates and at higher feedrates then it's most likely Mid band resonance.

[BernieNUFC]

I will try it at 80% to see if it improves then, thank you

[BernieNUFC]

Ran at 80%, motors got to 32.5c (did not seem too hot to touch) and i thought when it finished it had cracked it, What i thought this time that looked like a failure was a cutout which turns out to be a knot problem, so it may have worked, will try different wood but now i hope to adjust the speed at source
fingers crossed

[Blackrat]

ive had a brand new bearing give me nightmares on my Y axis ballscrew, it was on the floating end of the screw

randomly it would start to shudder for a few seconds and then go away, it took me many hours to pin point it, but when i found it ... oh boy it flew out the shop like a rocket on 'roids

[BernieNUFC]

I thought i would post my results in the off chance it may help another new person if they have a similar issue as me with the high-z s model.
It turns out it was me, i was driving it too hard at 45mm/sec and as Jazzcnc thought the bushes were getting too hot and were binding, i have now run at both 20 and 30mm with total success, just it ran longer obviously, the model above mine the t series does not have the same bushes so i imagine the problem may not occur so easily.
Thanks to Jazzcnc and Philip at Protools who walked me through the reasoning.
One happy camper again

[JAZZCNC]

I thought i would post my results in the off chance it may help another new person if they have a similar issue as me with the high-z s model.
It turns out it was me, i was driving it too hard at 45mm/sec and as Jazzcnc thought the bushes were getting too hot and were binding, i have now run at both 20 and 30mm with total success, just it ran longer obviously, the model above mine the t series does not have the same bushes so i imagine the problem may not occur so easily.
Thanks to Jazzcnc and Philip at Protools who walked me through the reasoning.
One happy camper again

Bernie I don't believe the bushes getting too hot is the cause. 45mm/sec (2700mm/min) isn't overly fast and certainly won't or shouldn't be getting any bushes remotely hot unless they are wornout and knackerd.

Don't know what reasoning phillip at Proto may have tried to parm off on you but any machine at the price they charge for those things should easily be capable of cutting at 2700mm/min. If not then send it back because it's unfit for purpose.? By that I mean they sell it for cutting woods and plastics which both need cutting at those kinds of speeds and higher.!

Now provided the machine is mechanicly sound, which it should be with such low millage and that Mid band resonance isn't the cause(Which it very well could be!) Then there are only a few reasons that you would struggle to get speeds you need and none of them very good or what Philip would be keen to admit.!!

Low ballscrew pitch less than 5mm or very poor quality low effieciency lead screws.
Under sized motors or poor spec motors.
Drives running motors with Low voltage. ie 24Vdc
Drives running motors with wrong current settings.
Very poor or low frequency pulse engine.

Now I'd guess the screw pitch is at least 5mm so I'll count that out and even Low spec 5mm pitch lead screw should give 2700mm.

Next and prime suspect is Low voltage driving the motors. Steppers get there speed and to some degree torque from Voltage. So if the drives are being run on low voltage of less than 36vdc then they struggle with torque when the rpm rise. Also if the motors are wound or wired series type then torque drops away very quickly when rpm rise.
Too low voltage and motors using series wired or low spec motors with very high inductance are common cause of low performance.
So why would manufacturer use Low voltage.? Because it means they can use cheaper drives and less chance of problems because they don't run the voltage near the drives Max capabiltys.

Correct Motors using decent drives driven with good supply of Voltage 44vdc or more and connected to 5mm pitch ballscrew should easily achive 2700mm/min cutting speeds on machine this size. Even if using less efficient Acme lead screws should still be no problem.

So what I'm saying is you either have another issue like Mid band resonance, which is what I would like to hope it is. Or the Spec of the machine is very poor and IMO unfit for purpose considering they advertise it for cutting woods plastics etc which it should do at correct feeds n speeds.

[BernieNUFC]

Thank you for that very informative post, one that I shall follow up.

In a like for like world what would be its competitor so I can glean a bit more info about specs etc, if that's possible :-)

cheers
Bernie

[JAZZCNC]

In a like for like world what would be its competitor so I can glean a bit more info about specs etc, if that's possible :-)

Well first you need to tell me the spec of what you have exactly because I'm just working on the little I know about those machines.

But what I can tell you for sure is what minimum feeds I would expect from any machine built using decent electronics and components.

If 5mm pitch ballscrews with 50Vdc drives running on approx 44Vdc and using parallel port then I'd expect at least 3500-4000mm/min cutting speeds and 5000mm/min rapid speeds. If same setup but with 10mm pitch then nearly double those speeds.
With Decent external motion control card then even higher speeds could be run like below will show. The only difference between these figures above and those achieved below was I used Cslabs IP-M motion controller.


To give some Idea of type of feeds that setup should give. Not so long ago I converted well used Denford router which used 5mm pitch and fitted with 50v digital drives running 44Vdc and it could reach well above 10,000mm/min. It wasn't run or tuned to work at those feeds to give very healthy safety margin but hopefully it shows what can be achieved with that kind of setup.

[BernieNUFC]

TBH I don't have anything specific to achieve except light work on wood, mostly 2.5d or some 3d plaques, so I was not looking for high speeds or big cuts. That said now I have got into the hobby a bit more I find that I do want to go faster certainly would like to be at 45mm/sec as I am conscious of the electric bill as well (yes laugh :-) ) but as all my work is for charity fundraising it eats into the money I can give out if I have to start covering the bills, so the High-z out of the box really fits the bill.
Changes I would like......well back to my original speed if possible, a quieter router (currently a trend t4), it works lovely but dam its noisy and it goes through the house, and I need to make a better extraction at the cutting point as the one I brought with the machine does not do the job it really needs a proper skirt.
So all in all I don't need a lot, I will get some more details on current setup and post to see if that offers up any more info.
Thanks

[JAZZCNC]

TBH I don't have anything specific to achieve except light work on wood, mostly 2.5d or some 3d plaques, so I was not looking for high speeds or big cuts. That said now I have got into the hobby a bit more I find that I do want to go faster certainly would like to be at 45mm/sec as I am conscious of the electric bill as well (yes laugh :-) )

This is another common mistake made by new users and what annoys me so much about theses low spec machines that can't reach correct feeds for what they are advertised to do.
The feeds are really dictated by the material and cutter your using not your preference or desire to be in rush or not. Also 3D work (dispite what you see there's really only 2D or 3D) often means high feeds rates or/and fast acceleration because of the small moves with light DOC or tiny step over amounts.

Cutting slower just costs you more money because of excess tool wear and leaves poorer finish. It also as you say cost's more in electricty and time.