Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: nelsonk11 on June 26, 2013, 10:04:57 PM

Title: jog lockup
Post by: nelsonk11 on June 26, 2013, 10:04:57 PM
I am running  Mach3 ver R3.042.040.  Sometimes when running a fast jog (shift arrow keys) the fast travel doesn't stop when I release the keys.  A couple of times this has been very scary.  The keyboard is fine, and it has happened on all 3 axis.  I am running servos with a Galil board.  Any ideas?
Thanks
Title: Re: jog lockup
Post by: Hood on June 27, 2013, 06:48:59 AM
Really need to post on the Galil part of the forum as it will be, if indeed it is, a Galil Plugin bug.

Hood
Title: Re: jog lockup
Post by: nelsonk11 on June 27, 2013, 08:05:06 PM
ok
Thanks
Title: Re: jog lockup
Post by: capnnick on June 28, 2013, 02:45:46 AM
Hi, I'm afraid that this particular bug isn't specific to galil. I just came here to create a topic in search of help for this very problem with my smoothstepper system and I'm glad to see that I'm not the only one experiencing the issue. The knowledge that it effects two different motion control boards may help us get to the root of the problem. Perhaps we can share a thread? I have tried both the latest lock down release of Mach as well as version 3.043.62 and they exhibit the same problem. I have found that the runaway jogging of an axis is 100% repeatable. It will occur any time that you release the shift key before releasing the arrow key during a rapid movement. Nelsonk11, can you confirm that you are able to reproduce the glitch this way as well?

I contacted smoothstepper support and they thought that it was most likely a mach issue. Their theory was that for whatever reason mach wasn't sending the "end jog event" to the plugin if I understood them correctly? Any thoughts?

My system, if it helps:
windows 7 64 bit
Ethernet smoothstepper
Mach3 (latest)
MachStdMill 2.0.6 plug in/screenset

Title: Re: jog lockup
Post by: Tweakie.CNC on June 28, 2013, 04:02:37 AM
Hi Guys,

This thread also relates to the same problem http://www.machsupport.com/forum/index.php/topic,24665.0.htm
To my knowledge, the cause has not yet been proven as to being an SS or Mach3 issue.  :-\

Tweakie.

Title: Re: jog lockup
Post by: nelsonk11 on June 28, 2013, 09:35:55 PM
I haven't tried the release of the shift key before the axis arrow, so I can't say.  I will do that Monday.  Interestingly, I ran multiple parts today and it didn't happen once.  It well could be I released the axis before the shift today. This machine has no external controllers, uses a Galil board and no pendant.  I think the Galil board is a quality board from what I know of electronics and not a junker from China.  It may be a Mach3 bug perhaps.  Yes we can merge threads, although I don't know how to do that.
Title: Re: jog lockup
Post by: nelsonk11 on July 05, 2013, 12:20:01 AM
You're right!!  If I release the shift key the wrong way it always hangs up.  This is a software issue with Mach3.  How does this bug get passed on the Artsoft?
Thanks
Title: Re: jog lockup
Post by: Tweakie.CNC on July 05, 2013, 01:40:09 AM
Quote
This is a software issue with Mach3.  How does this bug get passed on the Artsoft?

Don't be too hasty, I have no doubt it relates to the JogOff(); command but it is yet to be proven if it is Mach3 that is not sending the command or the SS that is not recognising the command under these particular circumstances.

One day it will be resolved but in the meantime at least we know not to release the keys in that particular order.  ;)

Tweakie.
Title: Re: jog lockup
Post by: nelsonk11 on July 05, 2013, 10:56:03 AM
I'm using servo motors with a Galil board--totally different hardware.  The only thing common to both is the s/w.
Title: Re: jog lockup
Post by: Hood on July 05, 2013, 05:01:51 PM
It certainly does seem like Mach is not sending the stop jog signal. I tried it today with the CSMIO controller and get the same thing.
Hood
Title: Re: jog lockup
Post by: garyhlucas on July 05, 2013, 05:31:46 PM
As reported before the xilifeng usb stepper board does the same thing. This is pretty deadly too, great way to break stuff! I am glad my E-stop button actually kills the dc supply to the drives.
Title: Re: jog lockup
Post by: capnnick on July 09, 2013, 04:34:27 PM
So if we have determined that the bug occurs on the xilifeng, csmio, smooth steppers, and the galil controllers I think its safe to say that it is a Mach3 issue. Is this on the radar to be patched in the next release or is there someone that we should email to make sure that it gets the attention that it needs?

It's a very dangerous problem that could get any one with a motion controller board in conjunction with mach3 hurt. When I was first learning to use my machine I had an aluminum plate turn into a flying buzz saw that took out a nice piece of wood from a nearby chair and some of my wall as well because I accidentally plunged a 45 degree face mill into it. It was my mistake that time, not mach3's, but with a bug like this it could happen to anyone.
Title: Re: jog lockup
Post by: Hood on July 09, 2013, 05:34:26 PM
Brett is aware of it and thus Brian will be aware of it. Whether anything is getting done I have no idea, Brett will have to let you know about that.
Hood
Title: Re: jog lockup
Post by: nelsonk11 on July 09, 2013, 10:40:21 PM
I'm not familiar with them--I guess they are the s/ware team for Mach?
Title: Re: jog lockup
Post by: Hood on July 10, 2013, 03:02:08 AM
Brett is support, Brian is the programmer/owner of Artsoft.

Hood
Title: Re: jog lockup
Post by: Hood on July 12, 2013, 08:52:51 AM
Bit of an update here regarding the CSMIO/IP-A
Whilst testing the afore mentioned bug (step mode using shift key for rapid mode) I found a bug with the CSMIO.
To get a slow jog in the CSMIO you alter the feed override pot. When I did that and jogged and then released the jog key the axis would often reverse, see the video here.  http://www.youtube.com/watch?v=GK3JxRhy4ao
 I contacted CS-Lab on Tuesday and they asked me if I could take a video showing the issue, I did so and the next morning I got an email saying they had found the issue and would post an update soon. I just got the update and its sorted. The original problem I was testing for, the one you guys are seeing, is still there and they are aware of it but it does look like it is a Mach issue. Brian and Brett are supposed to be looking for it soon, so fingers crossed.

Hood