Adhesive dispensing machine forming adhesive bead, can LinuxCNC solve this?

Hello,

I am using GRBL1.1 to start experimenting and understanding about router machine before concluding on which commercial version to use.
We have a adhesive dispenser, dispensing adhesive with air pressure controlled by solenoid valve. Now the issue I am facing is, when closed the solenoid valve, the pressure is still present for sometime causing the adhesive to flow causing the formation of big adhesive drop at the end.

I tried to continue its motion after closing solenoid valve but can't see much improvement.

I thought if I could continue its motion in background while closing the solenoid valve, but couldn't find any such threading support in GCode programming.

Another suspect I have is real time statement execution, I mean I feel as I am using GRBL, the command executions are not quick and so by the time dispenser start moving again after closing the valve, there is already a big chunk of adhesive formed.

G01 X35 Y0 Z-10 F1200
M05   ; here I am closing the valve
G01 X70 Y0 Z-10 F2000  ; here its continuing to move

Please let me know if my analysis to problem is correct and if professional Real time system (linuxCNC) will surely help in solving this issue.

If my analysis is not correct, any help is appreciable. 

Regards,
Vandana

I can't help much with your motion question, but is it possible to install a second solenoid valve as close to the dispenser as possible which will vent remaining pressure when the main valve closes?

If the main valve is normally closed, and vent valve is normally open, they could be triggered with the same signal.

It would also help if the main valve is as close to the dispenser as possible to reduce the line volume and residual pressure.

Just a thought...
 

Actually its not possible as the valve size and shape is selected to accurately dispense the adhesive for desired line width. Also the dispenser needs to be kept at minimal distance from surface to ensure height of dispensing doesn't affect the actual line due to its viscosity.

If you can at least tell based on your experience on, how much difference will come in real time operation when compared GRBL1.1 with LinuxCNC or Mach3 or any other good commercial system, it will be helpful

I suspect any 'stutter' or delay in task execution may be related more to your hardware selection than to the control software.

I've not used grbl other than for 3d printing.  I have used Mach3, Linuxcnc, and 2 or 3 other PC-based CNC control systems.  As well as many different 'commercial' systems (Fanuc, Siemens, etc.).

At the speeds and computations/lines per second you are likely to be using none of them stutter or delay or have any noticable latency unless there's a hardware issue or a programming error.

Even a 'dumb' old 3d printer with a 16-bit board running an old version of grbl will be smooth if the steps/commands per second are kept low.

Having said that, LinuxCNC is probably more capable of a custom or 'odd' control scheme than any of the others.  Specifically, you could program a dispensing move, shut off the valve, and then have it continue moving for a period of time - not just a distance.

Ok you mean to say I should check on the motor, solenoid valve and the lead screw and then we should be able to achieve?

It would be great if you can suggest the good quality motor, solenoid valve and lead screw, which really can help.. I hope I am not asking for too much..

Thanks in advance..

Everything needs to be checked.  Control hardware, motion hardware, dispensing hardware.

You need to post many, many more details on the current system & components before anyone here can offer specific suggestions.  Photos of the equipment are invaluable.

And, frankly, if what you're posting isn't interesting nobody is going to respond.  You're asking a bunch of hobbyists (though some are frighteningly clever) to help you redesign & troubleshoot a commercial motion control system.  Think about how you might tempt various folks in this forum to get interested enough to spend their time thinking about your problem(s).  That's why I responded - my brain jumped right to "why not bleed off the pressure?".

Specific machine details (how does this thing move, and what is moving? belts? ball screws? magic?)
Size of system (is this huge?  tiny?)
Air pressure required to achieve desired flow
Adhesive viscosity (roughly, not specific)
All current components in use, and what is good/not good about them.

Keep in mind that there are many commercial firms around who specialize in process optimization just like this.  If this is really for a commercial application it may make sense to hire someone as they are likely to solve it faster (and thus get you cashflow sooner) than a DIY approach.  Not that a DIY approach is wrong - especially for prototyping - but don't forget that buying your way out of most problems is the way to go.

The one thing you can't get back is wasted time.

Once again I really appreciate your detailed reply.

I have little knowledge of embedded. Trying to make the things happen.

We are just entering into this market. Getting the right person is really a difficult task here, we are trying in parallel.

I will try to find what best component I can find here when compared to what we are using..

If I fail to achieve what we need, then again will get back..

Thank you for all the support..

can you clarify this statement?
"
Another suspect I have is real time statement execution, I mean I feel as I am using GRBL, the command executions are not quick and so by the time dispenser start moving again after closing the valve, there is already a big chunk of adhesive formed.

G01 X35 Y0 Z-10 F1200
M05 ; here I am closing the valve
G01 X70 Y0 Z-10 F2000 ; here its continuing to move
"

if i am understanding that as i think i do let me know if the below matches your problems and why the solution is not working.
G01 X35 Y0 Z-10 F1200 (everything going fine line width perfect)
M05 ; here I am closing the valve (the second the valve shuts you see a big blob get put in half way in your line of glue at the 35 mark)
G01 X70 Y0 Z-10 F2000 ; (no substantial amount of glue seen for this move.)

if the above is correct then its possible the valve shutting causes a pressure spike increase the output for a brief moment and before the next move can occur.

G01 X35 Y0 Z-10 F1200
G00 X45 Y0 Z-10 (rapid move max speed can be a g1 with the highest possible feed rate)
M05 ; here I am closing the valve
G00 X70 Y0 Z-10 (rapid move max speed can be a g1 with the highest possible feed rate)

most machines have max acceleration so the idea above is to get the machine moving a fast as it can before you shut the valve off and it shoots out the extra glue, then to continue moving until the glue stops flowing.

so what should happen with the above code is
perfect glue line till 35
thinned out glue line from 35 to 45
valve shuts shooting out extra glue. with the increased speed this will hopefully be placed over a larger area.
45 to 70 should spread any glue over left after the blob.

Your understanding is absolutely correct and I would definitely try this solution, it surely will improve.. I didn't think of this solution. However I may have to change the motor as I felt due to weight on lead screw, it had already reached its maximum speed..

Thanks a lot for the idea..