Tormach losing Z position

[geomagnet]

Hello,

I've owned my PCNC for a few months and I'm starting to get the hang of it...but it's now behaving unreliable.

I have been prototyping impellers out of aluminum, but started to have accuracy issues so I switched to PVC to save $ and time. Now the accuracy
is even worse but appears to be isolated to the "Z-axis" motion. If I zero out
it's Z coordinate and run my g-code in air (or through plastic) then re-measure my true position, the Z coordinate is off between .08" and .2". Not to mention the part is scrapped.

I've called tormach and they sent me a new driver card, but that made no difference. I slowed the feed down to 6ipm and was able to hear the tone of
the stepper as it was travelling actually "hiccup". Thinking a motor was bad, I switched the leads between the "z" and "a" axis so that the "a" motor responded to the "z" signals and it too began to make the hiccup sound.

I've went through my computer and turned off EVERYTHING that is not need to run the machine. I re-installed Mach2 and checked all the settings. Everything is as it should be...however I continue to have problems. Especially if I leave the room, I'm guaranteed a tool crash.

FYI- I'm using Rhino4.0 and MadCam4.0 to generate the toolpath. As mention Mach2 is the controller.

Does anyone have any suggestions as to how I can get this machine working accurately?

[Ron111]

My 2 cents,

Things that you now know:

1. It's not the driver card (you replaced it)
2. It's not a Z-axis mechanical / motor drive problem (you reproduced it on the A-axis)

Is the problem intermittant or repeatable in nature.

Does it seem to be repeatable? at least to some extent? at the same point in your machining sequence?

See if there is an incompatabilily issue between the CAD and Mach2 or 3 (is it using unsupported G-Codes that Mach3 may not support?)

See if the problem can be isolated to certain line numbers of code?

My 2 cents,

Ron

an extra cents worth:

I final thought, with Mach 2 you can assign a different axis for any axis, for example you can re assign the step and direction pins for the x axis with the z axis step and direction and cut air with the current program. If you reproduce the same anomoly the you have isolated the problem to your software. (This way you have included any wiring issues (ie, bad connections, solder joints, corrosion) that can cause intermittent problems.

A couple a years ago, I had a simular problem in which for no appearent reason, my machine would take a deeper plunge cut that programmed. It turned out that me being a rookie, in bob cad when I set the Z axis plung speed I typed in I left off the "F" in "F6" and it would cause the Z-axis to go deeper.

One last thought, If you look at the tool path screen of mach3 can you see where you code looks like the path is wrong or does your path look perfect.
I have found a few problems by looking at the tool path here prior to running the program.

[Hack]

Not sure but with the type of part that you are making, it seems the file might be quite big. If so, the machine may be running quite a while as well. Is there a chance that heat build up might be causing it?

Just throwing it out there

Dan

[geomagnet]

The problem is repeatable but variable, I finally got tormach to run the g-code and they had the same problem but worse. Its certainly a software issue and hopefully monday their experts will figure out what is wrong. All I know is this is kicking my ass.


Thanks for your help.

James

[geomagnet]

Originally Posted by Hack Not sure but with the type of part that you are making, it seems the file might be quite big. If so, the machine may be running quite a while as well. Is there a chance that heat build up might be causing it? Just throwing it out there Dan

The parts are impellers, the code is quite large ~60,000 lines. From my understanding that's not too big. Furthermore, I've chopped out a portion of code 2000 lines and it still fails.

It's gotta be something with the post. I'll have the developers take a look at it and get it right.

Thanks
James

[Don Clement]

1. It's not the driver card (you replaced it)

Replacing the driver card does not necessarily mean the driver card is still not the problem. It could be another bad driver card or perhaps a problem with the DC power to the driver card.

[geomagnet]

Originally Posted by Don Clement Replacing the driver card does not necessarily mean the driver card is still not the problem. It could be another bad driver card or perhaps a problem with the DC power to the driver card.

How would I test for a bad power supply or other driver card gone bad?

[keithorr]

Interested in the outcome. Please update when the issue is resolved.

[NinerSevenTango]

The drive coupling on my Z axis began slipping the other day, and I snapped a drill off when it went from one hole to the next without retracting. I re-tightened the heck out of it, hopefully it never happens again. There's no key in there.

It sounds like you narrowed it down to the post or the software, though.

--97T--

[geomagnet]

Hello,

Yes, tormach took care of it fairly diligently. It required the Mach3 upgrade. The file should be available at the tormach site - http://www.tormach.com/document_library/PCNC3Rel1.4.exe

What a difference, the machine makes smooth turns. Even my 4th axis seems to "coast" to a stop.

Of course my new problem is thermal lock or a virus that causes the computer to lock-up around 3pm everyday. Waiting on a new computer to arrive as I can't find the virus or hardware problem.

Not sure if the Mach3 is the final release or a beta for Tormach, but it's wizards are very buggy and incomplete. Don't delete Mach2 in case you need an old wizard to crank out some g-code that works.

[levelzero]

Those are some nice impellers

Any chance those are compressor wheels for a turbo? How long do they take to cut?

[geomagnet]

I'm hoping they will become parts for a turbojet.
They take about two hours each and are somewhat
rough. I'm new at this so I really don't know how to
get a nice smooth finish. I assume you run additional
finish passes with finer increments then hand sand and
polish. If you have a 4 axis machine you can make them
by repeating the tool path for one blade and rotating.

As far as the turbojet, I've been wanting to build one
for years. I've built them before using automotive turbochargers
with impressive results. The problem has alway been with heat
and the bearings. But now there are space-age materials and
ceramics which can handle the temperature and high speeds
(so they say). So far it's been one technical glitch after another.
I'm sure when I have something to brag about I'll post it on this forum.

Thanks for asking.

James