CNC CONVERSION - 6" ATLAS LATHE

[RICH]

THREADING CHECK – LONGER THREADING
Mach3 version 3.042.27 and using a SmoothStepper.

For checking of how longer threading is functioning in Mach ( or my lathe ? ) a microscope was attached to the carriage for viewing and measurements.
The tube was finished down to a .7500” diameter, then the surface was colored blue with a permanent marker. A very pointed sharp tool ( no radius ) was used to scribe the lines on the shaft as it was cut.
The threading was as follows:
PITCH:  0.050”  and 29 degree angle ( basically there is no step change in the Gcode for Z as it just cutting radial)
PASSES: 60 @ 0.0001”
TOTAL DEPTH OF CUT: 0.003”
RPM: 402 / 20 IPM
Picture #57, 58 show the setup and scribed material after the threading cycle.

Now it is quite obvious that the pitch is changing as the threading is progressively getting longer.
I guess that has not been fixed yet in MACH! On the starting end of the thread there are five cuts in a space
0.005”, then in the middle there are 5 cuts in 0.011”, and at the end of the 2” thread there are 5 cuts
in 0.019”. This is shown in picture #60 and then enlarged to show the cutting grooves at the end of the threading as shown on picture #61.

Each groove measured 0.003” deep ( I have .0003” backlash in the X ) but the widening / changing of pitch is not do to the Z mechanically ( I have 0.0006 to 0.0008” backlash in the Z ). The rpm is not changing.
The steps per unit a correct.

In the last post a ¾-13 x ¾” long thread was cut and there were no lead problems. If this thread was
cut was 2” long only ¾” or  maybe 1” would have been usable. I didn’t get any X & Z combination moves,
but, as shown in the pictures, somewhere in the threading cycle there is still something causing a change in the Z such that one bigger cut is “always” done somewhere in  the cycle. On a punny lathe the thread may be ruined but on bigger machine you would just have a wider cut and may not even notice that quirk!

More threading tests need to be done but did notice that the affect is not as pronounced as in a smaller
thread. Maybe Mach4 will have have it fixed, but until then, there are still problems with threading on the lathe.
 
Till next time…….keep the threading short,
RICH

[WoodyCam]

Hi Rich,

Correct me if I am wrong, but I thought Art had spent quite some time looking into this and found that some PC's have timing issues in the way Mach3 spits out Z movement for threading?

What happened with all that work? Did version .29 implement a different means of generating timing that overcame this issue? It may be worth trying version .29 just in case. Change log doesn't list it.

On my ORAC I have a really old, but seemingly good NEC 1Ghz PC and using the parallel port, single index and version .23 (I think) and I get no variation in pitch that I can measure, certainly up to 100mm length. Lines up with a rule perfectly the whole length and functions as a thread. Lucky perhaps. Deep cuts though are a no-no on that version, otherwise the X axis goes crazy (but never the Z for me). I've not tried vn 29 yet as I don't seem to have the timing issues.

Woody.

[RICH]

The work kind of stopped as there was only Chip and I trying some things out that Art was testing.
It was difficult testing to do. I just posted in Problems Threading on the Lathe thread, so you may want to follow along in that thread. The combination Z X move that your seeing has been around a long time, it is a bear to really
see / define the actual movements when it happens. Haven't tried lastest versions and don't really know if there are any threading changes in them.
RICH

[BarryB]

Does this sort of problem affect all things done with Mach 3, or just threading?  It looks like a major problem to me...

[RICH]

BarryB,
Just the threading from what i know of.
RICH

[RICH]

Figured i would post this as it's a nice option for a conversion. Yes they come in an
assortment of colors. Even pink if that's your fancy, but the colored ones cost more. 
RICH

[Tweakie.CNC]

Very neat. It is often the simplest of ideas that are the best.

Tweakie.

[RICH]

Hi All,
The lathe is working great, but a sore spot was my computer.  I have never recomended a specific computer in any of my postings. I finaly broke down and had one put together. The company said it would work with me and put a solid computer together for running Mach. If it didn't satisfy the driver test, then components would be changed unitil it worked as I wanted. So off we went for a 1/2 hour lunch, returned and it was put together!
Good price, unbeatable service, with an absolute guarentee you could't beat!
Like, if the mother board was the problem  they would change it out, processor, and any component, with the cost only changing if the component cost more!

I need to get back to the company ( they have been around a long time and i have purchased pc's from them before)
and confirm that all is well. Then i will modify this post with their info.

The PC is solid. Here are some basics:
- Asus motherboard - M2N68-AM SE2
- 2 GIG Ram
- AMD  64X2 Dual Core Processor 5200+ 2.7 Ghz
- Asus graphics Card - EN9600GSO
- Rosewill PCI pp card - RC302

The pc will run my steppers at the same speeds as the SS. The PPS in the driver test will only change by 1 at 25khz and maybe 4 at 100 khz. Single line  with no bumps and you can play with the mouse and it dosen't screw it up.

This is the pc i am using for the threading tests if you have been following that thread.

RICH

[RICH]

Got side tracked for a while and back to making and testing out some of the assessories for the lathe. Here are some pictures of the drilling / milling  attachment that will be used on the lathe.

FIGURE 1&2 shows  the drill in a holder and the adapter plate. Setup is quick as no shimming is required. The base can be attached to the carrage on the X, Z axis or even on an angle. Machining can be done via a coded program or manualy by using the MPG's. The drill and not shown matching collet model were modified to remove axial play. Figure 3 shows additonal air driven grinders ( made by Dotco and Ingersoll Rand ) some of which  have an adjustment for speed. So rpm range is up to 60,000 rpm. The same holder is used for all of them just that an adapter is reguired for use of the air driven ones in the holder.

Figure 4,5  shows how the drill is alligned. A piece with a precision reamed hole is mounted in the chuck or collet. Using a extended eye relief microscope,  MPG's are used to allign the ground pin into the hole. The slightest movement of the pin can be detected. You can see .001" easily. The axis is then moved into the reamed hole some .5" deep.

RICH

[RICH]

Been experimenting  drilling holes. Manualy using the MPG's,  G83 canned cycle and some macro's. Good allignment, just the right pecking, and quality

drills are mandatory for accurate drilling and hole size. You will break a small drill in a heart beat if things are not right.
Feeds( F.01 TO F1) can be increased once a starter hole is accomplished.
A 0.020" drill is shown against a 0.001" feeler gauge in Figure #1.

Figure #2  shows ( 12 equaly spaced 0.020" holes into a .375"  diameter piece. The manual indexing band on the chuck ( Fig 3 ) was used and each division

is 1 degree. I have a veneer that can be used to bring it down ot 5 minutes of an angle.But in the future ...never ends.... a stepper will be used to rotate the

spindle. Figure 4 shows two mounting options i am considering. A removable bracket along with a spindle adapter will provide quick setup and removal.
 
At 200 steps / unit and 10us, resolution would be 10.8 minutes, or if i use the stepper with a precison 10:1 gear box ( no backlash in it ) then resolution

goes to 1 minute.
1 MINUTE = 0.0003" @ 1" distance
So if the you can hold the drilling location to 0.001" @ say .25", that would be equal to approx
13 minutes. To eliminate rotation inaccuracy the geared unit should be used.

RICH