Strange...

[steve_g]

Strange...
Yesterday I had something happen that has never happened to me before with my PRT Alpha...
I was V-carving in the center of my table when the bot headed for home between toolpaths... This alone irritates me as I have fixed my PP several times to eliminate this “feature”, but in the process of trying new versions of control software and reverting to old, I have lost my repair again. This time, however, I had inadvertently left a large wood sanding block in its path (shame on me). The Bot and the block engaged in mortal combat for a brief moment with a great deal of jumping and banging. Before I could intervene, the bot ejected the block and merrily went on its way. I let it continue as its an Alpha and if steps were lost, it would stop, right? Nooooo! it was an inch off!
My question...
The Alphas are supposed to correct missed steps or stop. But, are they protected from added steps?
SG

[bleeth]

Steve: Check your PM.

[dlcw]

Haven't had that happen. But I did have something happen the other day that cost me a broken bit and a wrecked piece of plywood.

Thank you Shopbot. :mad:

First time I ever did this (and the last). I had just started a job and clicked on the start spindle dialog and accidentally clicked twice. Well it turns out that when you click on that start spindle dialog box, the cursor jumps really fast to the Load Job File (which turns into a Click to pause button or something like that), the job paused. I selected resume, was prompted to start the spindle (I clicked OK) and off the job went. Buried the NON spinning bit into the plywood and ripped it up some before the bit finally broke.

Apparently the start spindle dialog doesn't actually start the spindle on a resume (at least on my machine).

I've stayed on the last release of 3.6 as I can't afford to be a beta tester of new control software in a production shop where I depend on the CNC for my livelihood.

Sorry Shopbot, I can't afford to run experimental software to help "make it better". It would be like asking you guys to experiment with your production machines or CAD design computers (any of them, not just CNC) with software that may or may not work to help "make it better". You have a business to run, just like many of out here do.

[ssflyer]

My understanding is that the Alpha driver can take up to 1/2 second before it faults - so yeah, they can lose steps, too - and take you where ya don't want to go!

[myxpykalix]

I have had that happen where something got caught in the way anbd it defaulted (I have a PRT Alpha) and even if i didn't think i lost steps i think i would want to reset everything just in case.
That is why i write down my locations before i start. Scott Worden has a good thread on how to refind in case of lost position, see above....:eek:

[jTr]

Don,
I did just upgrade to new version, and had a remarkably similar issue yesterday:
http://www.talkshopbot.com/forum/showthread.php?t=19666

Upon pushing spindle start, it did fire up, but software paused 2 seconds as usual prior to resuming. The moment it began moving, it kicked off output 1 to the spindle, shutting it off, so the bit idled down as it was being dragged across a completed piece, so same results you experienced.

Contacting support right now - will post insights if I get any solutions.
For the life of me, I cannot think of any circumstance on a resume or complete FG re-start where the long ramping motion would be considered useful. Even on a 3-D carving, it could easily ramp through waster edge border and/or hold downs. It would be really neat if it simply drove directly to re-start point of cut and plunged straight down from safe Z height. This ramp feature has got to go...

jeff

[jTr]

Don,

Support was very intrigued that my machine shut off output 1 on resuming. He had no prior knowledge of this being a known issue, and made a clear note to visit/share with the rest or the tech team, as this is a big one if it persists. Here is what was concluded from that conversation:

- Based on your experience and mine on new vs. old versions:
There is no "bug" or solution to the ramping move, other than to avoid stopping at the end of a part cut. In other words, try to wait until it's in the middle of a straight line cut, not when it's about to move to a new start point. The ramp motion I referred to is considered simply a 3-D move to new cut point and is not an alteration or new feature of the revised software.

- My incident may be considered a "perfect storm": maybe a double mouse click and or a coincidental static charge that caused the software to skip a step, either forward or backward in regards to triggering the spindle off.

Choice: Call it a fluke, take the steps above to avoid further issues and continue using newest version, or dump it and step back to an earlier release.

One thing I have learned about these machines, as odd as it may sound to some, there seems to be a bit of kinetic link between operator and the "operated". If I'm impatient and do not wait for software to complete it's task 100% (plus a millisecond to breath), it can get clogged and act unpredictably. I'll chalk it up to that, exercise a few more moments of patience, proceed and report.

BTW - I upgraded due to the fact that I was experiencing a steady increase in lost communications, regardless of prior efforts that seemed to have conquered the issue with additional grounding. I realized the latest spat of comm errors was directly related to my increasing of move speeds. Upgraded to take advantage of larger data buffering features of new software, as prompted by other contributors here. So far, no comm with the new version.

jeff