Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: remartinez on March 01, 2015, 02:05:51 PM
-
Hi guys, Need help!
Been searching on the forum for a couple of weeks on threading on the lathe to solve my issue but I have had no success. My issue is that the timing on threads are off on each pass. Question is how do I correct this issue. My thinking is that Mach 3 is not picking up the timing correctly.
Hardware:
C35 board,
C3 board for indexing. board set on active high with Mach3 also set on active high for the pin.
100 mm disc with one 3mm slot
VFD (Teco N3) controlled by a C6 board.
Mach 3 version: R3.043.066
RPM reading on Mach 3 is correct. (meaning that rpm might be off by 1-2 rpm at most)
Debounce set at 0
X and Z axis movement and backlash are perfect.
Any advice will be helpful.
Thank you,
Raul M
-
Have a read of Threading On The Lathe which can be found in Members Docs.
http://www.machsupport.com/forum/index.php/topic,13017.msg85313.html#msg85313
May need to set your rpm manualy as that can interfere with the threading cycle.
Not sure about version .066 infulence,as i have not used it, and others have gone back to earlier version
due to problems with it.
Make sure you have 3- 5X pitch of movement before actual threading to allow for axis acceleration
I'll let others reply about .066 who have tried or used it.
RICH
-
Rich,
Thank you for your response. I have read the thread and manual but still can figure it out.
What version of Mach 3 works for threading?
-
I use .057 and have threaded successfully on easy stuff. I think Mach is OK but my machine lacks the spindle power & stability needed for doing steel over 16mm dia. I'd kill for 1 - 2 rpm variation under load.;) I have a 3/4 hp Baldor DC motor that's belt reduced about 4:1 and it's still too puny.:( Maybe one of these days I'll stumble across a 1.5kw servo drive real cheap. That'd fix it.
Fortunately I don't really HAVE to do CNC threading much since I have another lathe that'll do the jobs I HAVE to do.
-
Dickeybird,
Have you ever tried doing the other methods of threading? ie; Alternate Flank Cutting
It helps when lacking power.
(see page 44 in the manual) In fact you can do acme threading using alternate flank with a sharp V cutter by adjusting the depth. ;)
RICH
-
Wow 3/4hp at 4:1 and you are having problems with spindle slowdown ???
Interesting, (;-) TP
-
I think it's not so much a lack of torque slowdown under load but more of a variation around the set speed which shows up in the Mach spindle speed DRO as a small but constant up & down reading at all loads. (No I don't have "Closed loop spindle control" checked.) I also have a fairly slow Z-axis motor/drive.
I think it's the KB DC speed control that causes it or my indexing disc slot width. My disc/slot is very close to the OP's and is supposed to be OK. I was going to try a wider slot or putz around with the KB but stopped trying to iron out the problem when the time spent/finished part ratio got too high to continue.
When things calm down a little around here I'm going to give it another shot or maybe just thread aluminum or acetal. ::)
-
Milton what is your max rpm?
Hood
-
Dicky,
Thank you for your response. I will re-install the older version of Mach3 and do some testing over the next couple of days and I'll get back to you with my results.
Thanks again,
Raul
-
Milton what is your max rpm?
1600 in direct drive, 400 in belt-reduction mode.
-
Ok was just curious when you mentioned having issues with 16mm dia and threading and 4:1 reduction. I would have thought there would have been plenty torque for even the standard 2mm pitch on the 16mm dia.
Is your Z axis fast enough for it, just thinking if it is not then that is maybe where your problem is coming in.
Hood
-
Ok was just curious when you mentioned having issues with 16mm dia and threading and 4:1 reduction. I would have thought there would have been plenty torque for even the standard 2mm pitch on the 16mm dia.
Sorry, I better set the record straight in case someone reads this thread down the road and gets the wrong impression about Mach’s ability to thread based on my feeble attempts.
My memory is apparently knackered because I found out last night that I was wrong when I said I was having trouble with ¾ hp @ 4:1 reduction doing 16mm threads. I dredged up an old post and found my problem was with 22mm x 1.5 (ER16 size - I guess that’s where the 16 came from) and that problem actually occurred before I had built the belt reducer.
After getting the belt reducer working, I later failed at cutting 50mm x 1.5 (ER40) threads in steel…a totally different ball game! Even then I don’t think it was so much a lack of torque issue but more to do with my errors and lack of knowledge. The completed thread wasn’t too bad looking but was oversized. When I tried to go back & “run from here” with a deeper cut, something went horribly wrong and I had to e-stop. It was all over then as I never could pick the thread back up and had to finish the part on the manual lathe.
I need to gather up some scrap stock (along with my courage) and spend a day or 2 cutting practice threads until it either works consistently for me or I can clearly determine that my machine will not thread properly until further hardware improvements are made.
Until then, I'll keep me big mouth shut about threading.
-
Ha ha no probs Milton.
Hood
-
How to Pick Up A Thread is covered in the Threading on the Lathe write up.
There are some cacluclators that will caluculate the HP required to do a thread. Results should be used as a general guideline.
Manualy calculating requires some assumptions and is not for the novice to do.
RICH
-
There really should not be a problem picking up a thread if it has not been removed from the chuck unless you have E-Stopped and the axes is slightly out of position. Having accurate homing would cure that.
Hood
-
There really should not be a problem picking up a thread if it has not been removed from the chuck unless you have E-Stopped and the axes is slightly out of position. Having accurate homing would cure that.
That's really encouraging. The ORAC had accurate homing on the X-axis then and now has it on the Z so it should be able to recover from a faux pas. That's assuming I learn to use run from here properly.
Also, does Dolphin generate good threading code for Mach?
-
I think in the standard form it produces code that is ok, I most likely altered mine to work out the depth etc without having to enter all in the CAM.
Hood
-
Have a read of Threading On The Lathe which can be found in Members Docs.
http://www.machsupport.com/forum/index.php/topic,13017.msg85313.html#msg85313
May need to set your rpm manualy as that can interfere with the threading cycle.
Not sure about version .066 infulence,as i have not used it, and others have gone back to earlier version
due to problems with it.
Make sure you have 3- 5X pitch of movement before actual threading to allow for axis acceleration
I'll let others reply about .066 who have tried or used it.
RICH
I have spent the pass 6 hours messing around with this issue and I have had no success. Let me explain the things that I have done.
First: I uninstalled and installed Mach3 version 0.57 (as per Dicky)
Second: I checked to see if a Z axis is accurate and it is (including with backlash).
Third: I went to manual speed on the VFD.
Fourth: I increased the slot on the disk to 5mm
I have tried the scribing test and sometimes its accurate and other times its not.
Sometimes I notice that the true RPM readout goes out of whack for a split second. Maybe this can be an issue but haven't figured out how to fix it.
The couple of times that I noticed that it was cutting accurate I let it complete the full thread. After measuring the pitch on the thread it was off by about 0.01 inch. Spindle speed was accurate. it might of lost 1 -2 rpm at most.
Extremely frustrated. Any ideas on what else to try?
Thanks,
Raul
-
A fluctuation in rpm of say 1 ie; flickers between say 99 to 100 should be ok. Idealy the rpm should be constant since the movement during threading is a
cut at a constant feedrate. If the feedrate is not constant the thread will vary in pitch. If it varies too much then the following pass is modified to adjust for the varying feed rate. That's why you set vfd manualy to an rpm so the VFD is not fighting with the correction the program provides.
If you were to scribe a line, allowing for some time for accelration, say 0.5" ( 5 pitch diameters) and then contact with the material using a G01 Z 1.5 F... the scribed line should be constant. You would need to use a microscope to measuse the variance over a few inches. Scribe line is only as accurate as
the ball screw / screw moving the axis.
If as you say, the rpm goes out of whack, say it goes to zero or drops say from 100 to 20, the threading cycle will try to fix the the threading by adjusting the next cycle. I have tested it for extreme drops in rpm, almost a complete stop, and the threading recovered but the thread pitch was not in spec.
In general the more stable the rpm the better the cut screw thread pitch will be. Should you consistanty get something , say less you can always change the pitch in the wizard and it will be used, and the thread will be good. But that is just tricking the program pathing and is self defeating.
This is all in the write-up.
I think you have an index problem, or it can be the vfd's ability to maintain rpm.
I can't speak for version 57....I use 38 and never had a problem.
If you are measuring the rpm, check to see what the +- specs are on your measuring instrument.
-------------------------
A test that can be done to check the acceleration / indexing is to mount a thin disc, the bigger the better, say 2" diameter, and do a thread.
You should see just one tick /scribe mark in the exact same location on the disc outside diameter. Under maginification the variance of the tick marks shold be almost zero / no distance between the marks. In fact if you try to do multiple start threading you will quickly find out just what your lathe can do.
RICH
-
Hey Rich, are you using the PP or an external motion controller?
-
DICKEYBIRD,
Still using the PP.
RICH
-
Thanks Rich.
Hey Raul did you make any progress? Coincidentally I had a little threading job to do in aluminum yesterday that came out great. Even at 300 rpm in direct drive (too lazy to switch to 4:1 belt reduction) everything stayed in sync.
Couple things I changed, dunno if it was part of my problem. I checked "Spindle speed averaging" in the spindle tab and set my debounce to 2. The rpm DRO was WAY more stable after that. I ran the thread several times with good results.
I don't know if the spindle rpm is read by Mach (when threading) via the rpm DRO or directly off the index pulse input but my setup appears to work better now. May have been the alignment of the stars & planets...or maybe the tinfoil hat I wore yesterday. ;D
Oh almost forgot, I generated the code in Dolphin which shouldn't have made any difference but was easier (for me) than the Simple Threading Wizard. I also chose 14 roughing and 4 finish passes. In retrospect I should have doubled the roughing passes to reduce the load a bit as the last few roughing cuts were peeling off a healthy sized chip (for my little lathe.) It all came out OK so I'm a happy camper today. We'll see how things go down the road when I have to do "real" threads in steel.
-
rpm is read by Mach (when threading) via the rpm DRO or directly off the index pulse
Directly off the timing of an index pulse, before, during , and at the end of the threading the DRO just shows results of the COMPLEX cycle.
Spindle speed averaging is the average of the upper and lower timing ( if i recall...) so helps but is not a cure all for a varing or way off
pc kernal / pulse frequency, spindle speed variation, etc. Not sure, but, the rpm used in the cycle is to 4 decimal places ( there were at least 20 different versions
of thread testing routines and just don't recall specifics.)
FWIW,
RICH
-
Spindle speed averaging is the average of the upper and lower timing ( if i recall...) so helps but is not a cure all for a varing or way off
pc kernal / pulse frequency, spindle speed variation, etc. Not sure, but, the rpm used in the cycle is to 4 decimal places ( there were at least 20 different versions
of thread testing routines and just don't recall specifics.)
Ahh, OK, thanks!
-
Thanks Rich.
Hey Raul did you make any progress? Coincidentally I had a little threading job to do in aluminum yesterday that came out great. Even at 300 rpm in direct drive (too lazy to switch to 4:1 belt reduction) everything stayed in sync.
Couple things I changed, dunno if it was part of my problem. I checked "Spindle speed averaging" in the spindle tab and set my debounce to 2. The rpm DRO was WAY more stable after that. I ran the thread several times with good results.
I don't know if the spindle rpm is read by Mach (when threading) via the rpm DRO or directly off the index pulse input but my setup appears to work better now. May have been the alignment of the stars & planets...or maybe the tinfoil hat I wore yesterday. ;D
Oh almost forgot, I generated the code in Dolphin which shouldn't have made any difference but was easier (for me) than the Simple Threading Wizard. I also chose 14 roughing and 4 finish passes. In retrospect I should have doubled the roughing passes to reduce the load a bit as the last few roughing cuts were peeling off a healthy sized chip (for my little lathe.) It all came out OK so I'm a happy camper today. We'll see how things go down the road when I have to do "real" threads in steel.
Sorry guys for not responding to the post. Work has swallowed me in entirely and just spit me out. I got around to playing around with my lathe again. One of the problems I had was that the axis rapid was moving to slow so I made the adjustment there.
The problem that I am still having is that once the rpm reading goes out of whack for a split second, the threading stops working. I have played around with the debounce but I have had no success. I am using the C3 card from CNC4PC.
Does anyone have any idea how I can fix this issue?
Thanks,
Raul
-
Hi, it can be a problem in the Index, it may be entering your signal noise index, if you use an inductive sensor, check that the distance from the sensor to the disk is correct.
In configuration check, Use Spindle feedback in Sync. Mode and Spindle Speed Averaging is actives.
regards
Renato
-
hi
i have a denford orac lathe & do a lot of threading on this machine doing a thread of 1mm pitch this is done at 600 rpm both for internal & external threads , a 1/4" bsp thread 19 tpi is done at 250 rpm
i am using the parrallel port to run my machine and using mach version 062 which i have found very stable over the last 12 months
i think the pick up slot on your spindle disc is not wide enough mine is at least 1/2" wide & i get a very stable speed with very little variation also very stable in the threading cycle . when threading the cycle is always started at 15 mm away from the job to give it time to pick up the pitch & settle itself before it starts cutting
robert