Hello Guest it is September 23, 2023, 09:24:29 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - mrprecise44

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 »
General Mach Discussion / Re: Rotate Mill head
« on: October 24, 2014, 01:45:52 AM »
Hi QA:

You don't say if you can rotate the mill head, to enable the spindle to be parallel to the X axis, however there is a lot to read before you will get something to work. In the terminology of G-code, the three planes are as follows:
G17; X,Y plane, with the Z axis perpendicular to the plane.
G18; X,Z plane, with the Y axis perpendicular to the plane.
G19; Y,Z plane, with the X axis perpendicular to the plane. This is the active plane you are describing.

If you will look in the G-code definitions, starting with the descriptions of G02, G03, they discuss the relative offsets to be used for each of the above planes. For the G19, the offsets are J and K.

You can play with some lines of code, and watch the tool path screen. The G-codes in the Mach3 program have very few examples to go by. I do not have the book, but the best source for G-code description are the books by Peter Smid. His CNC Programming Handbook in up to the 3rd. edition.

Look on Amazon, they have an entire page for the books on CNC by Peter Smid.

Changing the plane of activity from the normal X,Y is pretty obscure for this forum. I had to do a machine job using G18, to enlarge a mold cavity, and it took a lot of playing around to get it working. I ended up doing the job, but there was a bug in the Mach3 display, and instead of showing a half-circle cavity, it showed a full circle. At first I thought I had written the code wrong, but doing a dry run, with no tool, the motion tool path on the screen was actually correct, and did not move into the upper half of the circle shown on the screen. In my case, the path was the center of a ball mill, and milled a perfect half-circle cavity.

Smids books are not cheap, but from all I have read, they are the best.


General Mach Discussion / Re: Setting "steps per" on rotary (A) axis
« on: October 10, 2014, 01:18:48 AM »
Good suggestion, Hood.
What is the "North Britain" about. I heard the referendum did not get the votes, but that is all I know.

General Mach Discussion / Re: Setting "steps per" on rotary (A) axis
« on: October 09, 2014, 01:25:42 PM »
Hi Mike:

Glad it's working OK now. True enough, a book with all the little "gotchas" would be handy.

Regarding the 119.970 DRO reading is really of no consequence. Mach3 does the math, and shows the numbers in the DRO to the fourth decimal place.
What is shown is .03 degree. In reality, the difference between your readout, and 120.000 degrees is not measurable with standard machinist equipment. On a 1" diameter shaft, the rotational distance on the surface is equal to .000262". You can be off more than that just trying to get the centerline of a machine spindle aligned with the rotary table rotational axis

Mach3 cannot move the rotary table any closer to 120.000, as that is the result of the basic motor resolution, the gear ratio in effect, and whatever micro-steps the amps is set to move.
However, visually, it is an annoyance not to see a nice 120.000, since that is what you programmed. If we could modify the DRO, to 2 or 3 places rounded off, that would be more pleasing. I have seen some comments on the forum about doing this.

I have a 4th. axis driven by analog servo, with encoder feedback, and still get the annoying number effect.

Now you get to play with the rollover issues, going the wrong way, why can't I .......


General Mach Discussion / Re: Problem converting from LinuxCNC to Mach3.
« on: October 08, 2014, 09:52:09 PM »
hi ckovacs:

Glad to hear your problem was solved.

General Mach Discussion / Re: Problem converting from LinuxCNC to Mach3.
« on: October 08, 2014, 01:33:49 AM »
Hi ckovacs:

You might want to give the New Fangled Wizards a try for generating code for simple things like a slot. The wizard takes the tool diameter input in the set-up file, and the tool number for the tool table height offset, and generates simple code, using all G40, no cutter comp codes. Things like slots, circles, hole patterns are super fast, and quick. It will show a tool path preview, and then post the code directly into your tool file screen, ready to go. The Wizards run without exiting Mach3.

To be off by .040" is a huge error. It could be a goofy cutter comp getting into the mix, or an erroneous tool table wear setting.

If you write a simple program by hand,  to cut a square, about .100 deep, like 1.000 X 1.000, with no cutter comp, and measure with calipers, it should cut a 1.250 X 1.250 square.

M6 T1 (.250 cutter)
G43 H1
G0 X0.0 Y0.0 Z.1 M3
G1 Z-.100 F4
X1 Y0.0
G0 Z.1 M5

This will eliminate the machine as a problem. The machine program is telling Mach3 where to drive the cutter, and that is what it is doing. It will cut the shape "exactly wrong" every time, without fail.


General Mach Discussion / Re: Jog keys
« on: October 08, 2014, 12:47:16 AM »
Hi Anniepoo:

Welcome to the Mach3 Forum.

Go to the "Config" menu at the top of the page. Scroll down to "System Hotkeys". The table "Jog Hotkeys" is where you set the key for an axis. If you click on the Z++ button, and then hit the key you want for the Z axis "UP, that key code will appear in the box. Click on the Z-- button, and hit the key you want for the Z axis "down."

Most people use the PgUp and PgDn keys for the Z axis.

After setting the Z keys, and clicking the OK box,  scroll down to the bottom of the config menu, and click on "Save Settings."


General Mach Discussion / Re: Setting "steps per" on rotary (A) axis
« on: October 07, 2014, 01:29:33 AM »
Hi Mike:
There a couple of settings you need to enable.

The steps per degree should be a simple calculation based upon the ratio of the rotary table, the reduction ratio from motor to rotary table input shaft (if any), and the steps per revolution of the motor, which is usually 200 ppr for stepping motors. Also, if you are using micro-steps, this must be accounted for.

You did not say what the rotary table ratio was, and if there is any motor reduction between motor and rotary input pinion gear.

For the rotary table to act like a normal axis, go to the "config" menu, and scroll down to the heading "Toolpath."
The upper right box is labeled "Rotations." Check the axis of rotation, usually the X axis; check the A-Rotations Enabled, and check the "Use Radius for Feedrate."
Click OK, and scroll down to the bottom of the config menu, and click on "Save Settings."

Now open the "Settings" menu. Check the appropriate boxes for Axis you will be using. The X,Y, Z, and A should be black, and the B,C lit green.
The heading in the upper right corner is for "Rotation Radius." Since you checked the "Use Radius for Feedrate" in the Toolpath menu, here is where the part Radius is entered to make the axis's move in sync. Enter some general value, like 2.00 if you are using "Inches". If you are doing a real job, you should enter the actual part radius.

As before, go to the config menu and "Save your new settings.

Under "Config", open the "General Config..." page. On the left side of page is a box "Angular Properties."  Check "A Axis is Angular."  Click OK, and go to bottom of menu and "Save Settings."

Now close Mach3, and restart. Your new settings should all be enabled, and the rotary axis will act like another axis.

You must make sure the correct motor steps for 1 degree of movement. Usually, you can set the A axis Velocity and acceleration quite a lot higher than the X,Y, or Z axis.

Do a test with the MDI line. Set X to a move of something like 5", and A to about 40 degrees; G1 X5 A40 F25, and see how the two axis move. It should work in sync just like an X and Y motion.

If a command to move 40 degrees does not come out "Exact", you have made an error in the steps per degree calculation.

Also, this is all assuming you have the motor settings correct for steps per, velocity, and acceleration. This is all trial and error, and listening to the motor sound.

Hope this helps.


dspMC/IP Motion Controller / Re: Dspmc Encoders not in sync with mach 3
« on: October 05, 2014, 09:30:37 PM »
Hi Gfact:

I don't know anything about the EMC encoders. For the 5v supply to the encoders, you should setup a 5v supply, and not take the power from the I/O board. When I setup my machine, I kept getting errors on one axis. Upon checking the voltage at the encoder, I was reading about 3.5v and not the full 5v. I changed over to a separate 5v supply for the encoders, and input the 5v on the I/O board where the encoder wires connect. This cured the problem. It was a couple of years ago, so I can't give any details, but the low voltage was a problem as I remember.

The AMT-102 encoders with the differential line driver made a big difference in the machine. I have had zero problems in homing, and accuracy since.  Also, my original machine wiring was very high quality, with shielded cable everywhere, and individual wires in the loom with foil. Also, each pair of all the signal wires, i.e. A+/A-, were twisted pairs with separate foil wrapping on all the pairs.

When connecting the RJ45 connector to the loom wires, I make sure to use the twisted pairs of the RJ45 wires match the twisted pairs for all the signal wires.  You have to do a continuity check from the I/O board, all the way back to the RJ45 connector to make sure there are no mis-matched wires.

From what I have read about grounding, it is not a good idea to share power supply grounds because of ground loops. This can occur when there is a very tiny voltage difference in grounds, causing a current, which then creates a ground loop. Since the encoders are only 5v., the grounds are more vulnerable than the 24v signal wires that the dspmc uses. The dspmc uses a 24v supply, and the encoders can have a separate 5v supply. Also, make sure the AC wires are properly connected, i.e., Hot, Neutral, chassis Ground green wire.

Do not connect the DC negative to the chassis ground, as that can introduce noise into the system.

From your description of DRO's not moving, crosshairs on the run screen etc., it sounds like there is a major malfunction in the program behavoir. You might try re-loading the program, and also make sure there are not any old programs running in the background. The forum has various accounts of some programs interfering with Mach3.

I would suggest getting a new hard drive, and start a clean install from the ground up. Considering the cost of machine, tooling, and time spent overall in setting up a finished machine, the hard drive is a small expense. You can re-install XP from a disc, but Windows 7 will have to be a new program. Newegg has Win7 on sale for 85.00 this week. This way, you know exactly what the computer environment is.

That is all I can think of at this time.


Hi Tbc:

Using a probe and/or touch plate introduce require some electronic knowledge, since most probes function by a contact "opening", and a touch plate functions by pulling the probe signal "low" or to ground. On Dave Bagby's web site, he goes into this problem in detail, and has schematics.

On the CNC4PC site, Arturo has a schematic for the touch probe.

Another problem, aside from differing active HI/active LOW states, is the LED within the probe. When the probe contact opens, there is a tiny residual current from the LED, causing the circuit to appear still "closed". Another factor is the differences in breakout boards, and the signal being something less than 5v. when it should be a "HI", or never quite 0v. when it should be a "LOW".

Probing is definitely not a simple "plug and play" operation.


Hi Tbc:

What you want to do has been implemented in several screensets, using the Mach3 probe
function by talented Mach3 authors over the years.

On the Mach3 Forum, under Mach Screens/Finished Screens/Mach Blue Probing by Big-Tex.
This is a working tool probe screenset.

The MSM Screenset by Dave Bagby. This screenset has a touch probe, touch plate,
and automatic tool offset functions. This is perhaps the most sophisticated screen set
available, and it is licensed.

The third example is a prolific writer/machinist, Daniel E. Kemp, who goes by the
name "Hoss". He post all his work on on the CNC Zone.
Here is the link to his tool probe script;



Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 »