Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: keithmech on December 03, 2008, 12:36:32 PM
-
I have just upgraded to the newest level *********xx.20.
Everything seems to work fine but when I tried to thread using the wizard
I get un reliable movement on the z axis.Thread starts-goes in fist cut
does not retract on x comes back addes cut on x and cuts fine for random
number of passes then randomly starts to feed in on or out the x durring the cut
obviously with disasterous results.Is this me or the wizard?
-
keithmech,
Is there any chance that your steppers are loosing holding torque, skipping?
I assume you used the threading wizard that is in Turn. While in the wizard, did you click the settings box, which gives you another screen and put an X clearence value in ?
RICH
-
I knew about the clearance setting.I also pulled off the motor and checked the encoder
coupling, long shot as this is only showing up when threading.I seem to have lots of torque
and this will happen when cutting air.I'll try and hand code a thread and see what happens.
-
Has threading worked for you before?
Have a look in config menu and then plugins and enable the Turn Diags one, it may help find out where your problem lies.
Hood
-
I'm seeing the same problem in R3.042.020. Most of the passes are cutting correctly, but randomly the occasional pass will cut at a rapid feed, thereby destroying the pitch or the x axis will move (usually positive) as if cutting a tapered thread.
Peter
-
Have you enabled the plugin so that you can see if it picks up some problems?
Hood
-
I have enabled the plugin. I am experimenting with screwcutting in mach 3. Normal turning has been working fine over the past six months or so, but I have not yet cut a successful thread. I know I have a problem with automatic speed. I have seen others report similar problems in that the spindle speed changes during rapid moves, but the solutions given have not been successful in my case. If I try screwcutting with pwm spindle control, I get the message "scripter compile error in: spindle speed. mis"
I have therefore been doing my screwcutting tests with manual speed control and I have a nice steady feedback from the indexer and no error messages. The only problem is that roughly 2 passes out of 30 either cut at a rapid feed or with a taper on the X axis. The taper is usually positive, so it doesn't affect anything, it just doesn' cut. I tried approx 30 threads this evening, 4 of them were fine. The others were destroyed by the rogue cut at rapid feedrate.
Peter
-
I have enabled the plugin.but no difference.
This is random only on the first pass is the error there all the time
the tool moves the correct pitch but fails to retract in the x axis.
the next pass maybe ok as the next several ,then a random
taper or + or- move of the x axis as the thread is cutting
regular turning works fine.I tried to load the servo motors by
taking a .250" doc on the radius 700 rpm and 10 ipm feed
nice chip but it still worked ok.while threading I cut the doc down to .01
and no good.
-
what should I see on the diagnostic screen for the plugin?
My results seem random from one sample to the next.
-
To be honest I am not sure as I have never had problems with threading so have never used the diagnostics. Best bet is to post the info in the diagnostics and I will ask Art to have a look.
Hood
-
Hi ALL,
I haven't tried threading with V3.042.20 or even loaded it on the my laptop.
So just want to peg a time line and version which was used by me and some info for others to relate.
The last threading I did was with V3.042.018. Steppers,Single index slot, G32 code ( no G76 CANNED CYCLE )
and all worked just great. Before cutting any threads, spindle speed ( manually set and manually varied)
along with the G94 the G95 was checked. Mach's RPM readout was the best ever, varying and recovering to correct DRO indication inclusive of the z axis behaving appropriately. Backlash compensation was used.
Now i only cut about five threads.
The first experiment confirmed to me that the puny lathe was going to do good because spindle slow down was varied to mimic a heavy cut and the z movement responded, tried to deliberately screw the thread up by manipulating the speed at the start, during ( about 40%), and end of the thread. Leaving two finish passes to clean it up. The thread didn't come out to well, but, remember this was just a brutal test.
Now the other ones were "very good" and to dimension. So with that limited testing......life is good.....and left things alone.
The SS stepper setup also seemed to work very well in dry running but without backlash available no actual testing was done.
So until I try out .20, can't say much more.
RICH
-
It's been a two weeks, since fooling with threadiing so don't rmember specifics on the new
PLUGIN CONTROL>Turn Diagnostics. It's actualy a pretty neat tool. ART should give the exact description,
but, it shows that index pulse triggered thus out of G94 and into G95 ( threading in effect ) and the variations
on the speed ( there is a limit on just how much the speed can vary ( don't ask / don't remember / and what i knew may not apply). Need to keep your eyes on the screen and motors in view for a good indicatioin of what its
telling you ( especialy if you can..... safely.......fool around ). You can see a few of the same things if you click the
Diag ( little tab located in the upper left side of the screen) while threading. Don't screw around too much if your running a slow computer.
RICH
-
Hi:
First thing to check is the Spindle RPM, is it stable in its reading? Secondly, try soem G32 moves, Its not clear from the description if this is a G76 or G32 based program,
be nice to see it run in just G32's to see if its a vb error of somekind or not. Also, Id report that spindle.m1s error to Brian, sounds like that macro file is hosed somehow..
Art
-
my spindle rpm is stable within 1 rpm.This is using a G76.
I'll go out and try a G32 and see what happens.
-
Here are some further observations I have made today.
If I use auto speed control (PWM) my speed is unstable 285 to 315 at nominal 300 RPM (varies with feedrate on Z axis)
In this condition I get erratic movement of x axis on almost every pass of the thread and variation in pitch. This is the same on G32 and G76.
If I use manual speed control, the speed is quite stable (298 to 300). In this condition most passes of the thread are correct, but 1 or sometimes 2 passes result in movement of the X axis and or variation in feedrate.
After several more tests I was able to observe that once or twice during the complete thread my spindle DRO speed would jump to 1000 RPM for a second or less, then settle back at 300. When this happened, the following pass would have the problem with movement on the X axis and/or change of feedrate, then continue normally. This also happens in both G32 and G76.
It therefore seems that the problem is speed related. I can understand that with the variation in feedrate, but the strange movement of the X axis was not so apparent.
The OP says he has a very steady spindle speed. I thought I did, but it wasn't until I watched the DRO throughout the entire thread that I noticed the momentary change.
I have played with index debounce, speed averaging and step/direction pulse in motor tuning, cleaned the optical switch on the indexer but these don't appear to affect it. I tried reducing the acceleration of the Z axis and this appears to have helped. It may just be a fluke, but I will try some more tests at different levels.
It wasn't my intention to gatecrash the OP's original message, but I believe from todays tests that there is a clear link between the spindle speed feedback and the strange movements on the X axis.
-
I disagree at the moment on the spindle speed feedback.This problem is only there with treading.
My spindle speed is very stable ,I tried taking very small cuts so no way would the spindle slow down and I get the same
error.I noticed also that when this random x axis takes place it is as if the cycle restarts itself ie,
say 5 cuts fine which should be to finished tread then 5 more cuts tapering in random.
-
I have just cut 6 successful threads, 3 in G32 and 3 in G76. I solved the momentary blip in spindle speed feedback by switching off enhanced pulsing in general configuration.
I still have the same faults in auto speed, but that is a seperate problem. I will try some more tonight to see if the faults re-occur.
-
Keith, have you monitored the spindle in the plugin to see if it registers a missing index pulse, also whether the rest of the info looks good?
It may be that you are seeing a very stable speed in the DRO and that you are not taking a deep enough cut to alter it but however it may be the feedback that has a glitch and thus the threading is playing up.
If you look in the Mach3 folder you will see the macro that is for threading (m1076), make a copy to the desktop for backup then open the original from Machs VB Editor and you will see a line that says Test = False (or something like that) Change the False to True and when you get code from the wizard it will produce G32 rather than G76, this will help you see what the wizard is putting out and whether Mach is deviating from it.
Hood
-
I am now seeing repeatable screwcutting without errors in both G32 and G76.
-
Hi:
Cant see "enhanced pulsing " doing that..
You may want to turn it back on just to be sure, enhanced pulsing usually makes the pulsing
smoother..
Art
-
Hood,
Thanks for posting about the change required to get a G32 code as I completely forgot about it. :)
So attached is a pic of what needs to be changed.
RICH
-
before I had a chance to try Hoods m1076 threading macro swap I
pulled all the electronics from the machine ands set about to rewire .
maybe I have some noise or something .Holy crap what a way to spend an afternoon.
??? :o.More to follow.
-
And we're back.Redid the whole electrical cabinet .I still have a problem though
I tried the m1076 false/true.Still nothing.However , with the turning plugin
I see that my actual rpm and set rpm are constant with in 1. the locked rpm are very differant.
What controls the locked rpm?
-
KEITHMECH,
You are not getting the G32 code output from the threading wizard after changing the m1076 file?
RICH
-
yes I am getting the G32 code output.This makes no difference at all.I tried
changing the encoder disk from 1 slot to 4.I ran the spindle calipration and
my true speed and entered speed are on.I even tried cutting a thread with
.001 infeed per pass .With straight turning this lathe can take .250
on the radius for a doc.It is a hardinge.I'm machining c12l14 the closest thing to butter.
-
Hi:
At the moment, locked RPM WILL be different from actual RPM . This is by design. The program is using the
interrupt system for threading, which means the RPM may be off by a certain % , but pretty much everyone will
see a difference between locked and actual, thats normal.
I dont recomment using 4 inputs on the slot though, just one, and in the INDEX input, dont use the timing input if you
dont have to. If the RPM is steady, then all shoudl be well. In the threading diags plugin, an led will light up if an index slot is missed,
does it light up?
Art
-
Hi Art, no the index light does not light up.I made up a new 1 slot disk 20 mm slot width.
I adjusted the pid loop so that with a 2x4 wedged against the spindle the motor holds a nice steady rpm.
The thread pitch is good . If I stop after a cut or 2 and check the pitch its as good as my pitch gauges.
I wonder if having my pulse board connected to a second pp might be the problem.I find it strange though that the x axis
has a mind of its own.
-
I added a dwell g04 of a second at the start of each cut, just to give the computer
time to think what it is doing and now I have some very nice threads coming off the machine.
I can live with that for now.Kind of looks like its waiting for the number on the half nut to line up ;D
I have a smooth stepper on its way for another machine, if it works as advertised I'll probably
get one for this machine as well.
-
Well I know the SmoothStepper works flawlessly for threading on my lathe but then again so did the parallel port. I have about 20HP at the spindle when threading at 1000RPM so I didnt expect any issues. However a month or so back the spindle sheared a key in the headstock and the spindle had no power at all, it was still rotating due to the tight fit of the gear on the shaft but as it was threading the revs were dropping drastically, down to under 200RPM during cuts and back up to 1000 when the tool was out of the work, I decided just to let the cycle finish and to my amazement the thread was perfect. :)
Hood
-
Kind of ot for this thread but I installed my first smooth stepper.Took about an hour
to mount connect configure and be amazed.Works great on the big mazak.
I'm kind of reluctant to get the speed up much past 400 ipm as the table on the machine is 1500
lbs.Now just have to get the tool changer sorted out.I have a cubloc plc to handle that.
I'll be installing a smooth stepper on to the lathe soon.
-
Sounds like an interesting machine, what model is it? What were the original rapids on it?
Hood
-
The machine is a mazak v5 micro center.(hardly micro at 9600 lbs), as far as original rapids I have no idea.It came from a university with low hours ,1300 ,and after the shipping company had there way, no control.I had originally tried to fit a Fagor
8055 but dumped that once I got on to this Mach thing.Believe me the support for fagor is not even on the same planet as Mach.
The machine looked like crap with a layer of crud composed of old cosmoline and dust ,it took weeks to clean up.The
paint underneath was great as are all the slide ways.The original spindle speed was 6k but I replaced the original
dc motor with an ac /vfd.Now it runns at about 8k.smooth as silk.The tool changer has 24 pockets and uses 5 proximity sensors to read the binary code on the face of the tool drum.I used the servo drives from Viper and the original motors .
This project started out as a trip to hell but I found that it turned into something more than cnc capabilities for my shop.
-
Just done a google to see what it looks like and seems to be a nice sturdy machine :)
From what I have found the rapids seem to have been 12m/min for X and Y and 10m/min for Z so your 400IPM is not far off.
Hood
-
Is there an option to choose whether full form screwcutting tips are used ?
If a sharp pointed tip is used the thread must be cut deeper than necessary just to get the required width. This is a problem cutting threads on thin wall tubes that engineers have designed to be strong enough with a proper thread form.
Is there a way to tell mach 3 to create the bottom radius and extra width to compensate for the lack of tip radius on the tool.
The only other options I can see are to either grind the point off the carbide tip or buy new tool holders and a packet of tips for each pitch I might want to cut.
Thanks,
Glen.
-
You choose the finish X dimension and that controls the thread depth.
Hood
-
That's what I usually do but I have to cut very deep to get the required width. It weakens the tube having a sharp point at the bottom of the thread and being much deeper than should be necessary.
Glen.
-
So you need to use the correct profile tool.
Hood
-
Ok Thanks.
I just thought there might be an option for Mach3 to generate the thread form with a sharp point tool based on the pitch.
Thanks,
Glen.
-
You could maybe mess around with an offset as you would when doing a multistart thread but just make the offset much smaller. For example run the thread with a start Z of 6mm then run again with a start Z of 5.9mm.
Would be much easier just using the proper insert though ;)
Hood
-
You're right, it would be easier using the right insert in the long term.
Thanks for your help. ( nearly 14,000 posts - wow )
-
Actually this is what i was looking for. It was in version 1.84 but I don't know how successful it was as it seems to have been removed in later versions.
-
Think all that will do is calculate the depth required to form the full thread width depending on the tools profile, not offset to widen the thread, but worth a try.
Hood
-
As Hood posted you need to account for the shape of the tool tip and adjust cut depth accordingly when using the wizard.
This is a problem cutting threads on thin wall tubes that engineers have designed to be strong enough with a proper thread form.
Well maybe the engineer needs to be educated on current tool profiles and thus adjust his calcs to take into account stress intensification factors and wall thickness so that tooling costs are minimized and specialty designs are not required. Matter of fact, the thread form should comply with industry standards and most thread inserts provide for the correct shape. Even if you use a rounded tip you still have the same factors ....I guess it depends some on the number of threads threads being done.
Is there a way to tell mach 3 to create the bottom radius and extra width to compensate for the lack of tip radius on the tool.
Actualy there is a way to get the extra width but will have a flat bottom. You better have a good lathe system.
Go to Config>Ports & Pins>Turn Options and in box for Cut Type enter a value of 1
as that defines that the threading will be done with Alternate FLank Cutting and the Wizard will generate the correct code. During threading
it will cut alternately between the front and back flanks of the thread, you will need to set the depth correctly for a sharp point and set up correctly.
You can actualy cut an acme thread in this way with a sharp V tool. Note that it will not radius the thread root though. No Sharp V tool is
exactly a point and will have a radius, may be small, but it will have one.
You may want to have a read of the Threading On the Lathe Manual which can be found in Members Docs.
See section 5 and section 7.1 on configuration settings on Cut Types.
RICH