Hello,

Working to calibrate the machine after fixing the x axis pully alignment ran the Circle Test3.cnc to test skew. I am not running fan fold foam, as I am ordering this weekend, using 3/32" chipboard I am noticing skew. The circles are elliptical, but almost coming back to the same spot. I think this is a good thing. It could be the material.

In all of this when running the Z Axis Calibration with CNC USB the Page Up and Page down numeric pad controls reverse. I don't know why. Has anyone had this happen before? To fix the problem I reload the CNCUSB Settings provided with the CDs and I've tried the process many times and achived the same results. I've made sure that I specific that the bit moves -1", which is down. It is usually about a 1/16" off and I put in the measured distance, see the ratio, hit okay, and when I page up, the bit moves down.

I've updated to the newest software version and also updated the firmware. Any thoughts? Am I doing something wrong?

Thanks,
Rob

 

Rob,

I know when I first started up my machine the Z axis was reversed. If I remember correctly I just went into the settings in the CNC USB software and told it to reverse. I never had a problem with it after that.

I am not at my computer right now, so I can't tell you the exact menu choice that I did. Let me know if you can't find it, and I'll look for it when I get home.

SlowBipe

 

Mark,

Can you confirm this? I thought someone on the forum said there is a bug in the calibration. That it doesn't save the reverse settings. So they return to the unchecked condition after the calibrations and you need to manually go back into there and check the box again. If this is true, then we should let Andrej (sp?) know.

-Kwok

 

I have written Andrej about this bug (reverse checkbox is cleared when you calibrate its axis) and others and have not received any response from him as yet.

The elliptical or oblong circles are caused by backlash which appears to be inherit in the two roller design as the drive rollers reverse direction.

Backlash is disabled or non functional in the USB CNC software so I have had to temporarily revert back to Mach 3 to get my machine working properly. Hopefully, the next beta release will have these issues corrected.

 

Thanks everyone for the info. I'll try it out and verify that this is the fix.

 

I always wire motors so that there is no need to use "reverse" checkbox. If motor rotates in wrong direction I swap two wires.

When calibration asks you how much did machine actualy move you must enter negative number, if direction is wrong.

For example, if current position is 0, you want to move 4 inches and your machine goes to -4 then you should enter -4 and reverse will be turned on. I'm not sure and I need to test if currently active reverse setting is used or ignored.

 

No one is questioning the use of negative numbers to calibrate the axis.

If the wires were reversed we would not need a reverse checkbox, but since we have one, clearing it after you have finished calibrating the axis is dangerous and could get someone hurt. When we entered the calibration routine the checkbox was checked. We expect that it will remain that way after the calibration is complete, which is not the case.

 

I disagree. Axis calibration tool must change axis reverse setting if required - that is if negative number was entered.

I'll test this and fix today if I found a problem.
[pre]
//I suspect this is code I use:
Axis.Reverse = IsNegative
//but it should be like this:
if (IsNegative) then Axis.Reverse != Axis.Reverse[/pre]

My RC transmiter has end point and travel adjustments but I always try to configure control horns mechanicaly first. Just because some things can be done with software does not mean that they shouldn't be done mechanicaly first. This also applies to backlash.

 

The reverse axis checkbox allows the wires to be reversed yet process the gcode as if it were not.

Regardless of what it does or does not do, if I set the checkbox to indicate I have reversed wiring, I want it to stay that way. Just because I calibrated the axis does not mean I changed my wiring at the same time!

Calibration has nothing to do with it. The wiring did not change, so the checkbox should not either.

The checkbox was set before I entered the calibration phase. When I calibrate the machine, the checkbox is cleared, so now when I execute a G1 Z 1.5 thinking the gantry is going to go up, it plunges the bit into my hand (if it happened to be under the gantry).

Mach 3 does not clear the reverse checkboxes when you calibrate with it, and neither should your software.

 

If calibration shows that your machine moves in wrong direction it should set "reverse" to correct value.

I don't think it is a good practice to have your hand under your spindle. Especially if your machine in not configured and you don't even know which direction will it go.

 

What is it you don't understand? The hand under the spindle was just to make a point that you could get hurt because the software made a change you did not expect!

My wiring is reversed. I WANT it reversed. I KNOW that it is reversed. I CHECKED a checkbox to tell the software that I know it is reversed!

When GCode executes a command to tell the Z to go negative, I know that it will go up.

Now when I tell it to move .5 inches and it moves .6 inches, I need to calibrate it, I know enough to use minus values etc. and the calibration process is not in question here.

The fact is that the checkbox was CHECKED before I entered the calibration routine and it is UNCHECKED after the calibration routine completes.

When I go to cut material and I do not realize that the software UNCHECKED something without telling me about it, I am expecting my axis to move in the same direction as it did before I calibrated the machine.

The Z axis now goes down when I expected it to go up like I told it by checking the checkbox in the first place.

The CHECKBOX is being cleared on any axis I calibrate even though the checkbox was checked.

This is wrong!

Mach 3 does not clear your reverse axis setting just because you calibrate the axis.

 

You know, Flash, I opted not to get the USB board because I suspected that just what happens above will happen. I surmised it form going through the site before I bought the printer. I realized that Mach 3 was done professional and that this software was not going to be there any time soon. l'm sorry to see that I was right.

 

I was correct. Latest beta version 2.10.1101.2401 that I just published has this fixed.

 

I too did not get the USB board and I am using a smoothstepper with Mach3. I do think the the CNCUSB software is on a professional level, and I am suprised at how quick Kroko addresses issues people find. Those two things are the main reasons I am considering getting the CNCUSB board. So, thanks for the hard work Kroko.

I do understand logicaly that if you calibrate the machine by moving -5 inches and then tell the machine that it moved +5 inches that it should think to reverse the axis, but I think it should also give some warning, like a pop up box to let you know that it reversed it. That way you can eliminate some confusion, and certainly help keep things safer. :02cents:

 

Thank you Kroko. I emailed a list of problems on the 15th regarding this and other checkboxes which are altered without the user expressly requesting it. Will you address those issues now? You can PM or email me if you would rather answer this privately.

It really makes no difference to me as I have switched back to using Mach 3 for now, but I am sure that those who have no choice would really appreciate it.

To be fair to Andrej, he addresses most problems almost instantly, and there are a lot of features I really like about the Planet-CNC board which for what we are using it for, should be just fine once all of the kinks are worked out.

Mach 3 has been around for a long time and has had time to mature. I consider it the gold standard and is what should be strived for when we look at other alternatives.

If and when it becomes a fully workable solution for me, then I am sure my wife would love to have her computer back and I can once again go back to a USB solution.

As 3DMON mentioned, backlash is probably the biggest issue that needs to be resolved at the moment.

 

Calibration will not reverse axis direction unless negative number is entered. When calibration results are displayed there is notification that axis will be reversed.

I'm working on backlash. I already have experimental version but it needs to be tested. I need to add some backlash to one of my machines first.

Thank you all who acknowlege my work.