Issue with SCP 2.1 rel 003

[EdThorne]

Is anyone else having this issue?

I have been testing the new SCP 2.1 Rel 003. I notice that the "Move to 0,0,0" button does not move to 0,0,0. The x and y axes move to their 0 position but Z does one of two things. If it is presently lower than 0 then it moves to +0.75". If it is above 0 then the z-axis just stays at its present height. Pressing the "Move to 0,0,0" button a second time moves the z axis to 0. It therefore becomes a two step process to move to 0,0,0.

When running a "G code" file, it appears that one must be sure that the gantry is actually at position 0,0,0 before starting. Otherwise, the Z-axis first starts at position 0,0,0,010" and every pass is 0.010" high. In other words, cutting to -0.125" would actually cut to -0.115". The offset stays constant and doesn't accumulate. It is just always 0.010" high for each pass. This doesn't change regardless if z is below or above 0 when starting and is not affected by the starting positions of the x and y axes.

So the workaround for now is to make sure that the gantry is actually at 0,0,0 by pressing the "Move to 0,0,0" button twice if necessary before running the G-code. The G-code tracks properly if the gantry is moved to 0,0,0 before stating the run.

I haven't tried the virtual z setup so I don't know if this has the same offset problem.
Ed

[4DThinker]

I actually like the way it works now as before the move would ramp down as it moved to 0,0,0 and depending on where it started and occasionally run through a clamp on the way.

[EdThorne]

Hi 4D,
The cut is off by 0.010". I don't see how this is a good thing. I don't want it to cut to 0.140 when I tell it to cut to 0.150.
Am I missing something?
Ed

[EdThorne]

Attached is a simple tap file that cuts a square. The PDF walks through the steps with Z set to non-zero and Z set to 0 before starting a run. The resulting cuts are different. It may be something that I am doing wrong but I don't see what that may be.
Ed

[Joseph Poirier]

2.1 has a new safety feature.
Prefences... 2nd click to (0,0,0) [x] {default setting} ...
This is where you decide to use 2 clicks, 1 to safeheight over x0 y0, and 1 to z0.
To use just one click to (0,0,0) uncheck the box, and remember you're the one that told it to run down to the board by doing this.

Safeheight is now defined as 1/8-inch over the thickness of the touchplate or the defined safeheight or the current position... whichever is the high point.
this is to prevent crashes into the top of the machine.

The system is supposed to set the starting point to be wherever the current safeheight position is to avoid the tapoff at z0, but the actual cut depth should be the same whether or not you are at z0.0 when you start your run.

However, you may be correct... the system is supposed to move to 0,0,0, but it may be offset slightly to prevent nicking the board... the next update should not have the same issue, nor should the virtual since it sets the starting point in the controller. That said... 2nd run in the controller will be dead on, but 1st one may be off by 1/100th.

Please test our new 1.6 with the same cut... and leave the router anywhere. I think you will find that it cuts the right depth on first run out of the box no matter where you leave the router. But please let me know your results.

[EdThorne]

Hi Joe, Ver 1.6 Rel 002 works fine. It doesn't experience the 0.010" cutting offset error. I'll turn the feature off on 2.1 until it gets fixed. I really like this new feature when it works like in ver 1.6. The gantry no longer moves to 0,0,0 for the first move. It instead moves to the first actual cut position. Lots of us will love this feature. 4D is correct that this will eliminate moves that are generally unwanted. Thank you very much for your help!!!
Ed

[Tab]

Ed, When is the new version being released ?

Tracy

[EdThorne]

Ed, When is the new version being released ?

Tracy

Hi Tab,
I am sorry but I have no idea as I don't work for NWA. I have faith, however, that Joe will fix the issue and that this fix will be included in the next released build ... whenever that may happen.
Regards,
Ed

[KevinO]

Hi Ed,
I confirmed your issue with V2.1 rel 003 with virtual turned off. I also had a 0.010 offset if I didn't start at 0,0,0.

However, I tried to repeat with virtual turned on and got different results. First thing I noticed is that the gantry goes to 0,0,0.010 when "Move to 0,0,0" is clicked, and clicking a second time didn't change the position. Note that this occurred whether or not the preferences box (one click or two) described by Joe was selected. I jogged the Z to 0 thinking this would make sure I was starting at the right position, but unfortunately when I ran a test cut, every pass was 0.010 too deep and the final depth was 0.010 inch too deep (from the screen and confirmed by measurement). The maximum virtual correction in the z axis was only +0.004 so that doesn't explain it. Also, when the cut was finished, it returned to 0,0,0.810 instead of 0,0,0.800. I don't get it.

I probably shouldn't have jogged the z to 0 before cutting. I'm going to try again and just leave the position at 0,0,0.010 before cutting. If that doesn't work I'm going to wait for Joe's update to the software before using virtual again.
Perhaps if you have time, you could repeat your test with virtual turned on. I'd be curious to see if you had a similar result.
Regards,
Kevin

[EdThorne]

Thank you, Kevin. I didn't try Virtual Zero. I'm sorry to hear that it is also a victim of the bug. Hope that Joe is still following this thread and will fix it in the near future. I guess that I will have to revert to 2.05 and live with it for the time being.
Regards,
Ed