Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: mariod on July 27, 2012, 07:43:35 AM

Title: Input 1.
Post by: mariod on July 27, 2012, 07:43:35 AM
I'm new to Mach.  We are having trouble integrating safety circuits into Mach 3.  We are trying to use input 1 as shown in the general options, but find that it does not inhibit manually jogging the axis.  It may work with gcode program execution, but does little in manual.  To follow safety protocols, the machine needs to be locked down completely when the safety circuit is active.  An EStop is not an option as the safety circuit is tripped frequently (in the form of a light curtain), and it would be a nuisance.  Any suggestions?  Thanks.
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 11:54:09 AM
I would use the Input 1 feature as it will stop Gcode program execution without positional loss  AND use a simple brain or macropump to turn OFF Jogging when the input1 is active.

(;-) TP
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 01:32:21 PM
Input 1 will stop instantly just like pressing Stop so loss of position can occur if you use it whilst in motion.
 What you could do is have a Brain or macropump see an  input (Not input 1), call a feedhold, then  have it switch a relay to enable input 1. That would inhibit code but not lose you position. You could also then have the brain or macropump look at the Jog Control LED and switch it off (if its on) and that way you will not be able to manually move and axis either until switched back on.
Its a bit of a round about way but its the only way I can think of to achieve what you want.
Hood
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 06:08:27 PM
I thought that the input 1 just blocked the execution of the Gcode program to keep it from running?

(;-) TP
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 06:12:28 PM
It does but if there is motion the steps are stopped with no deceleration and again started with no acceleration when Input 1 becomes inactive again.
Hood
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 07:39:23 PM
Testing here it seems to work ok using the input#1 function. Stops with deaccel. while Stop stops instantly
Could be wrong though, (;-) TP
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 07:43:46 PM
Get your scope out and try again ;)
Hood
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 07:59:58 PM
Just did it myself see if you agree.
Scope of initial move then three Input 1 activations then finally a feed hold.
Note the acceleration at the start then no deceleration or acceleration on the three Input 1 activations/deactivations and finally deceleration on the feed hold.
Vid link is here, will keep it up for a while.
http://www.youtube.com/watch?v=jKiltZQ1CBQ&feature=youtu.be

Hood
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 08:06:59 PM
Also here is a post where it was discussed and as you will see later on Art confirmed to me there is no decel.
http://www.machsupport.com/forum/index.php/topic,12918.0.html

Hood
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 08:17:18 PM
HOOD I cannot make out the vid very well I will test the machine back at the shop on Monday.

Testing here at the house i set the velocity wide open and the acell all the way down and watched the dro decay as it sped up and slowed down. The STOP came to an instant stop.

IF the input#1 does not use the accel/deaccel it is useless for a safety. Might as well just Estop it.  

From the safety aspect then MACH3 cannot be made to be safe as you must stop the machine including the spindle which CAN be done BUT at the loss of position and restarting is NOT one of mach3 strong points. Machine with large look ahead tends to run on a bit before stopping with feedhold.

Just a thought, (;-) TP

(;-) TP
Title: Re: Input 1.
Post by: Hood on July 27, 2012, 08:29:45 PM
I suspect what you are seeing with the DROs running on a bit is there is a slight delay to its operation (at least when emulating) , so in that respect it is not like pressing the stop as that is immediate.

Hood
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 08:47:12 PM
You could be right, the longer it is not having an active machine here at the house that runs MACH3 the more I seem to  forget about how it is suppose to work(;-)

Probably about time to sit back and watch the newbies play with it .

(;-) TP

Title: Re: Input 1.
Post by: Overloaded on July 27, 2012, 10:09:45 PM
I almost responded to the original post, glad I didn't though. Was going to suggest a macro instead.
Emulating, looked fine here too, but made me wonder what the heck its function is actually intended to be. ? ?
I picture a mill or robot with an enclosure.
Open the door or gate while a program is running and it stops (instantly) most likely loosing pos., close the door and it resumes ? ? no reset requested ? ? ?
  Imagine a robot and the door closes behind you, you'd be eaten alive.
Seems like it should just prevent the activation of any code or program cycle while IN #1 is active, otherwise EStop !
As far as disabeling jogging, while the door/gate is open (In 1 active) you might want to do some work with the pendant or what not. So you would NEED jogging enabled.....but no chance of a Cycle Start.
Any idea of its intended function ?
Russ
Title: Re: Input 1.
Post by: BR549 on July 27, 2012, 11:52:48 PM
I think the original intent was IF the safeties were not armed THEN the machine would NOT run the G code program. That part still is there. I think the back end got added on as an after thought. If you open the door the machine stops with Estop.

Probably the best option as you cannot really depend on feedhold to always stop quickly. So you cannot stop the spindle UNTIL motion stops. So there is a period of unsafe condition after opening the door.  I think running with the buffer you have little choice but to either run the buffer out OR dump the buffer and stop instantly.

AND you have to remember Mach3's  hobbist background. I don't think it was a top priority at the time.

Just a thought, (;-) TP





Title: Re: Input 1.
Post by: Hood on July 28, 2012, 02:18:00 AM
Mach4 will be interesting to see on a machine, Feedhold is supposed to be instant, fingers crossed :)
Hood
Title: Re: Input 1.
Post by: Overloaded on July 28, 2012, 07:13:15 AM
I think the original intent was IF the safeties were not armed THEN the machine would NOT run the G code program. That part still is there. I think the back end got added on as an after thought. If you open the door the machine stops with Estop.

The afterthought was with good intention then, but must not have been verified to work properly, cause it dont do that.
It just does a slam bam hybrid feedhold, then resumes ... and if the spindle is running, it don't even stop. (sim)
Seems to be a dangerous liability the way it is. Surprising.
I would not recommend using it for obvious reasons.  :)
Nasty bug.  >:D