[dale99b]

Another day, another issue with my 3040 router...


I can cut perfect squares, and drilling works great. So I thought it was all calibrated and happy until I tried to cut a trapezoid shape (see image). The black lines on the third cut show roughly what shape it should be. The problem is, the X doesn't step at all over those long Y movements, so the sides end up parallel. There should be around 0.5mm X change over the 40mm Y movement.


I can feel the X stepper sort of dithering very slightly, but it doesn't step.


I can cut a square no problem, without losing steps. So it seems like some sort of resolution issue; single X steps are ignored. I've tried changing speeds, acceleration and pulse length but it doesn't make a difference.


I'm using Mach 3 via parallel port on an old laptop and docking station, with all the power-saving turned off.

Any thoughts? Is there anything that can be changed on Mach3 or the stepper drivers to sort this?

http://www.cnczone.com/forums/attach...d=382188&stc=1http://www.cnczone.com/forums/attach...d=382190&stc=1http://www.cnczone.com/forums/attach...d=382192&stc=1

[ger21]

Run drivertest.exe in the Mach3 folder, and see if you get a steady pulse rate, with a smooth, flat line.

Also, check your CPU usage in Task Manager while the machine is running.

[dale99b]

Cool will give that a try tomorrow, thanks.

I've got an old PC with parallel port I can lug out to the garage, so can try that if the laptop signal is poor.

[dale99b]

Driver test looks good I think (image below). I have RMclock running which keeps it stable. Anything else I can check? Parallel port voltage?

http://www.cnczone.com/forums/attach...d=382202&stc=1

[KH0UJ]

Check for X axis backlash

[dale99b]

Check for X axis backlash

Minimal. The X stepper isn't moving when it's told to, so it's a control or motor issue I think. Just not sure how to identify and fix it.

[KH0UJ]

Did you try to move the gantry itself? (X axis)

Can you post the measurements of your sample cut? let me code it my way and post it here for you to try it out on your unit

[dale99b]

Problem solved; steppers needed to be active high, not active low like the chenglish instructions listed.

So for continuous stepping it was fine, but single steps were skipped

[Sterkte801]

I am having identical issues with my Chinese 6040 machine.
When moves are made at higher feedrates, the machine is very accurate. When it tries to move in smaller steps(Below 0.050 changes in gcode), the motors do not move. They make noise as if they are trying to move but do not move.
It is like Mach3 is telling the motor to turn, but the motor does not turn.
I was going to do some tests to see what resolution this occurs at, but it seems line any move smaller than 0.05 it happens.
I can say this, when I first boot up Mach3, the mill motors will not do step moves if the feedrate is less than 50mm/sec. I have to go in and manually do an F150 command in the MDI, then can make step moves.
This has to be related to the issue we are seeing.
Another thing to note, when I set my spindle speed, I have to set the speed opposite of what I want. What I mean is speed of the spindle is max of 24000 RPM. If I want it to run at 24000 rpm, I have to tell Gcode to run it at 1000 rpm. If I want it at 10000 rpm, I have to tell it 12000 rpm. This goes along with what might be happening, if the spindle is backwards it is possible the configuration on the parallel interface board, or motor controller is also backwards.
It seems to likely be related to that TX14207 board.

Does anyone know if it is possible to limit the resolution that CAM software will create toolpaths for? I would like to try limits in the software to see if that is causing issue.

One other post on a search, the guy found if he changed the switch on the board from positive to negative changing it solved his problem, was wondering if that could fix it.
I have checked all of the switches, they are as recommended by the guys who sold me the machine. This does not necessarily mean they are correct now.

Have also read many posts regarding noise on the controller. Have done several things like adding grounds, adding a cap across the controller power input to limit input noise, changed where machine was so it had less possible input noise, tried changing, accel, and velocity settings from lower to really lower, one at a time, and both at a time. Also tried changing to a higher level. None of these made a difference.
Tried running with spindle turned off, still no better.

Any ideas would be appreciated.
Cliff

[dale99b]

Have you tried changing the pins to active high in the Mach 3 settings? That's what fixed it for me. The instructions were wrong. Or change to active low if they're currently high.

[stephann]

Problem solved; steppers needed to be active high, not active low like the chenglish instructions listed.

So for continuous stepping it was fine, but single steps were skipped

Dale, you have, with your simple solution, solved a problem that has bugged me for two weeks now! I was about to go and buy a new controller and drivers for my 6040 CNC.

All I had to do is just search every possible thread in existence to find this comment!! So, just flip the check marks to X's under Ports-and-Pins->Motors for each axis, and Voila!!! Works like a charm.


I'm back to making chips :-D

[Thomas91]

Did you mean that you need to click on these checkboxes?