Stalling with 3.6.1

[harryball]

I've encountered an interesting problem. I tried to surface my table with the same file I've been using for more than a year and it stalled at line 702. My communication efficiency is 83% and I've never had a problem with the 3.5.x software. I tried a modified version of the same file where I can enter the depth rather than the default of .02, it stalled in EXACTLY the same place but on line 711. I tried several times, stalls everytime.

I reinstalled 3.5.20, loaded firmware etc... the files run just fine. Table surfaces without any issues.

This is the results on the table. This represents 5 stalls in exactly the same place.

3877


This is the result on the screen. Unexpected Error - Comm.

3878

The software closes and the spindle keeps running. Only when I open the software can I regain control, stop the spindle etc...

ideas?




/RB

[ken_rychlik]

RB, I would take my machine to that location where it stalled and do a wiggle test to try and loose connections with anything.

I am wondering if at that exact location a loose wire is twisted just the right way, or wrong way.

Kenneth

[harryball]

I checked, no obvious loose wires, parts etc... If there were I would have expected the same problem on 3.5.20

I've ran the surfacing routine about 8 times after the downgrade to 3.5.20 without a problem. I'm installing a new spoilboard so I surfaced my current board down to 1/8" and have to surface the new board. Works fine in 3.5.20

/RB

[myxpykalix]

harry,
Have you tried making a new surfacing routine with the new version of software? If i understand correctly you were using the old surfacing routine with the new version and had the problem.
Just for grins make a new routine with the new version and try it. Just like how some older files won't open in newer versions of programs that might be the source of the glitch.

[robtown]

Will it work just doing a straight up CR command?

[harryball]

The cut file was generated by VCarve Pro. I'm not opposed to regenerating the code in Aspire but I find it disturbing to think old cut files may not be compatible with newer control software. I can't imagine that's the case.

I have not tried the CR command. I'll give that a shot on an aircut.

I'm also waiting to hear what Shopbot has to think. I sent it to support with the file.

/RB

[Gary Campbell]

Robert...
I also have been experiencing some unpredictable, but now repeatable odd occurrances with 3.6.1 that may or may not be related to your problem. I have sent a file also...
Gary

[jerry_stanek]

I just tried 3.6.1 and have had slight problems with it. I will be going back to an older version.

[gabepari]

Try editing the part file and tweaking the numbers a few thou. I've had numerous part files that had a "bad" combination of moves that would "kill" the software. On all occasions, changing the coordinates that the bot moved to, or from, solved the problem. Sometimes you can reduce your tolerance, I see you're running 6 digits after the decimal, and that will eliminate it as well.

I think it is a math error induced by rounding.

[scottcox]

I think Gabe is spot on. I had a very similar issue and after eliminating everything else, I modified the file by rounding the numbers on that particular move. Then it worked fine. So I modified the post-processor to show only four digits after the decimal, recalculated the path and it's ran fine ever since. Using 3.6.1 by the way.

[butch]

Robert
I have similar issue. They had me turn off the limit checker, and that helped. I will look at rounding down the number of digits and see if that helps.
thks
Butch

[ted]

Hi Guys,

Robert's problem was a bug in the new software resulting from an internal speed calculation that just went slightly out of range. We were able to fix the problem shortly after we received a copy of the file. The fix is in the latest version of the software that is posted on the website, Sb3.6.2, and is now there ready for you to download and try.

Gary Campbell's problem, for which we also have the file, is a different issue that we believe is close to being cleared up as well. It will probably take a fix, but is a very unique situation that I don't believe is likely to apply to anyone else.

(Robert's problem would probably have been fixed by doing something like the procedure that Gabe describes. But, I would prefer it if we can get such files that have repeatable faults, and get the problem fully fixed.)

Jerry ... sort of an extension of previous comments. If you can send us a problematic file, it will be very helpful. The 3.6.x software has been in beta for a long time and been well tested. We are pretty confident that any problems are quite limited. It is pretty robust and has lots of new features. I'm sorry that you've had problems, but I don't believe there can be many ...

Please note that from now on (because of the security constraints imposed by Vista, to which we are trying to now be compliant), that to send us the helpful ShopBot.ini settings file or Problem.log it will be easiest to use HELP > Send Support Files; or HELP > User Data File View (if you want to check them out). When we're testing files, it is always best if we can do it with your settings file so we get an exact duplicate of the computer conditions.

Butch, I can't help but feel you must be dealing with a different issue and that you are getting a message related to limit switch triggering if we have suggested you try with limits turned off.

In general: If I could talk for a moment about support. Here at ShopBot we try and distinguish between problems that are related to the technical side of the tool actually running right, and questions that have to do with the craft of CNC work. Sometimes, of course, there is a fine line. But we'd like to suggest that when a problem really does seem tool-technical related, you work with us first. We work hard to check for phone messages and emails after hours and get back to people as soon as we can. There is a risk that we may never hear about a problem if it gets buried here on the forum -- there is just so much material here that we can not monitor it all. And, we are also concerned that no matter how well meaning everyone is, there is enough difference between ShopBots and between people's understanding of the issues, that this is not really an efficient place to get this first level of tech issue resolved. Just my plea ... I realize ShopBotters will be ShopBotters ...

FYI ... we have been pondering a new approach to support that uses a combination of a blog site and twitter to get people expedited information and responses to problems they might be having. We conceive of this as partly interactive like the forum, but a more focused environment where tech support and the ShopBotter with an issue will take the lead, but others can jump in. Because we'd like to get video integrated with this, the new Google Wave system may also have usefullness.

Ted Hall, ShopBot Tools