[John Wales]

Hello everybody.

I'm new to the community and to CNC. As you can guess I have some problems. The first is how to post messages. (post quick reply?) (go advanced?) I'm new to forums too.

I posted a similar message today but it went to the wrong forum. At least I found the proper place to post or I hope I did.

I'm running Mach 3 with Gecko 201 drives on a home built table. I modified a tiny G code file to cut five identical circles. The first circle does not complete where it started by has a small offset in X and Y. The four additional circles look like I stepped on a Slinkey. I even reloaded and upgraded my Mach 3 with no benefit then used another computer. This is some sort of cumulative error involved here. The digital readouts still give the same coordinates but the positions have actually shifted.

Is there somebody that can tell me what's wrong?

Next problem: Where can I get a small solid state relay that will be triggered by the 5 volts at the 'bob' and control the 120 volt AC that drives a spindle motor?

I've been curious about 2.5 dimensions when dealing with cam. Would somebody with a lot of wind explain what this really is and compare it to 3D?

Thanks for any help and I want to repeat again that I'm glad to be here.

John

[ger21]

This is some sort of cumulative error involved here. The digital readouts still give the same coordinates but the positions have actually shifted.

Is there somebody that can tell me what's wrong?

Most likely, your machine is losing or gaining steps. Try lowering the accel and velocity in motor tuning by 30-50% and see if that helps.

[John Wales]

Hi Gerry

Thanks for the help. I ran the machine today and did so much slower (F numbers). The circles were tighter but I still had 15 mm wide kerf when I should have around 3. the distortion runs diagonally so this tells me it's a problem in both the X and the Y ordinates.

I went over the machine again and tightened up everything but a repeat run showed no change.

John

[ger21]

Can you post the g-code?

[John Wales]

Hi Gerry

Here is the code. You will notice there are a lot of repetitious lines. They were intended to lock the machine into the basic position of 6,0. The feed may look high but one circle takes well over a minute to complete even with the tool producing very little drag to the system.

G0 G49 G40.1 G17 G80 G50 G90
G92.2
G21 (mm)
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G00 X6 Y0
M5 M9
M30

This code was used as well:

G0 G49 G40.1 G17 G80 G50 G90
g92.2
G21 (mm)
G00 Z0.2
G0 X6 Y0
G3 X6 Y0 I-20 J0 F100
G0 X6 Y0
M5 M9
M30

The result here was one 'circle' that did not end where it started.

In my attempts to narrow down the problem, I reduced the velocity dramatically and did the same with the acceleration. The steppers are rated at 150 oz-in. I replaced the X and Y motors with 540 oz-in units and the same results were found.

John

[ger21]

Can you post a picture? Use the "Go advanced" button and click on "Manage Attachments".

Do you have the correct current sense resistors for the motors you're using?

What do you have for pulse widths in the motor tuning screens.

While the machine is cutting the circles, open the Windows task manager and see what the CPU usage is.

What is your acceleration actually set at?

Have you run Drivertest.exe in the mach3 folder? Do that and let me know what the results look like. Or better yet, post a screen capture.

[ger21]

I've been curious about 2.5 dimensions when dealing with cam. Would somebody with a lot of wind explain what this really is and compare it to 3D?

Long winded I'm usually not.

Generally, with 2.5D CAM programs, the majority of cutting is done with the X and Y axis only (2D). The tool will move up and down to the cutting depth, but the Z will not move when cutting (.5D). Some 2.5D programs may let you ramp into the cut cut, but one at cutting depth, the Z usually won't move. Typically, 2.5D CAM programs work from imported 2D drawings.

With 3D CAM programs, you load a 3D model into them, and toolpaths are created that follow the surface of the model. All 3 axis, X, Y, and Z, can move at the same time. Keep in mind, that there may be different machining strategies, and some may plunge to depth and cut using only X and Y (typically for roughing operations), but the capability to move all 3 axis.

[John Wales]

Hi Ger

I've been away and did a bit of thinking. I altered the power supply so it was filtered better. It helped a lot. Another power supply tended to remove all the interference i was having. I can do circles now!!

On another topic. The homebuilt machine had a bridge and a cap only to provide DC to a spindle motor. It screamed when it worked. Another thing it did was make the Z axis slowly move, depending how it was last used. The X and Y motors were effected too.

The power came from the same cord as used to power the computer and P/S for the drives. I used another cord to draw power from the other side of the line (two times 120V AC to make 220V AC). The problem went away.

*****

The old equipment I'm trying to get going has some complicated circuit boards. One says PDF 155. On it is another small board with PDF 161. There is no other identification. A note says 'checked 1987'.

This daughter board may be to provide the pulses for the larger board. I can't get them to work with Mach 3. The connections to the motor are through a connector and wire that I do not have. I followed the pinout for the parallel port connector but the rest is just guessing. I can get it to hum or make noise. Sometimes it shakes.

Any ideas on how to find a schematic?

John Wales

Thanks for all of your help Ger.