Hello Guest it is March 19, 2024, 03:45:27 AM

Author Topic: Accurate homing  (Read 26321 times)

0 Members and 1 Guest are viewing this topic.

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Accurate homing
« Reply #30 on: August 21, 2007, 06:43:44 PM »
I think I found something to help. In config, system hot keys, DRO select can be a scan code (and key or combo key, a or alt a) I think.
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Accurate homing
« Reply #31 on: August 21, 2007, 06:55:40 PM »
Yup, that works. It will take some getting use to. Assign a key to the scan code. If you go to the Gcode window, the only way I have found to get to the other DROs is to use the mouse. Once on a DRO, use the right arrow key to go to other sections, Feedrate, spindle, machine DROs. If more than one dro in the section, use the up or down arrow keys to get to the one you want. Left arrow key takes you to the g-code screen. Stay away from it.

Brett
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!
Re: Accurate homing
« Reply #32 on: August 22, 2007, 12:59:13 PM »
Thanks,

Been busy making money.  Give me a couple days to try this out.

-Ted
Re: Accurate homing
« Reply #33 on: August 22, 2007, 11:08:31 PM »
I tested opto-interrupters.  It turns out that they are amazingly accurate.  See:

http://www.vinland.com/Opto-Interrupter.html
Bertho
Re: Accurate homing
« Reply #34 on: August 23, 2007, 07:00:00 AM »
I don't know what made me think of it, but is the NISAT 6 axis controller still available?  If you want  a high end system for free, try that one, if it's even still around.  I remember running accoss it years ago, atleast 5 years, only catch is it runs under Linux, and I highly doubt it has a support forum such as Mach, but what other company does?  It was supposed to set the standard for PC controllers, issued by the NISAT, and was free to download, fully functional.  Does anyone remember this or even know what I'm talking about?
Michael

vmax549

*
Re: Accurate homing
« Reply #35 on: August 23, 2007, 09:11:17 AM »
Are you referring to EMC2 under Linux. It is a very interesting format IF you are a linux guru. It lately has been updated from the original EMC. It is better than the original but nowhere near MACH in user freindly. But it does have an interesting s curve axis accellaration curve tuning. (;-) TP

(;-) TP

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: Accurate homing
« Reply #36 on: August 23, 2007, 01:46:03 PM »
This thread really hit a sensitive spot with me.  After working for years on many types of equipment, seeing how large machinery manufactures can make a really fantastic machine and then REALLY screw up the safety related aspects of the design.  I've worked on a $300K seam welder that the mfg fitted with N.O. mechanical home/limit switches.  When the cabling wore through due to poor cable management (another beef of mine) , the machine would not stop when the head homed and consequently $2,000 (or more) worth of TIG and plasma torches were ripped off the machine.  This situation was not only costly but dangerous for the operator.  When I asked the mfgs. engineers about reprogramming their controller to work with N.C. switches I was told that their controller DID NOT have the ability to work with N.C. switches, and they looked at me like I had two heads.  My point is that safety of the machines operator (or bystanders) and secondly the machine should be paramount.  It's important to understand what the different safety system are for and what they should do (and not do).  Here are a few links to that effect:

http://www.designnews.com/article/CA6250678.html
This article talks about a drive system that allows the drives to remain powered in a safe stop condition.  There are two very relevant paragraphs toward the end:

Quote
Peabody distinguishes between a safe stop and an e-stop. A safe stop allows a machine to return to a known position from which it can commence working without an elaborate reset process. An e-stop removes power to the machine immediately, stopping it but possibly scrambling programs. After an e-stop, a machine needs time to gather its wits.

With robots, an operator can push a safe stop button and a locked perimeter guarding will remain that way until the robot has come to a controlled stop. Only then will the guard unlock.

Here is a small PDF from a E-Stop mfg.  This paragraph titled "How shall an E-stop stop the?
http://www.baumerelectric.com.au/pdf/ENG_Chapter_E_Stop.pdf

Now there are different regulations depending on your location and/or the machines function.  Different industries have different requirements.  The gist of it is that an E-stop should safely stop the machine as quickly as possible and remove power to the drives.  With our small table top machines we can be a bit lax and not spend much time worrying about this.  When we build bigger and faster machines we need to take it seriously.  Stopping a machine (through feed hold or similar functions) is not the same as an emergency stop (limit).  Other features in machine controls like feed hold, homing sensors and soft limits are there to keep us from hitting the limit switches or the need to hit E-stop.  I guess my big point is, we should not try and design around the intended function of an emergency stop circuit.   Use it for only the correct purpose and be safe.  OK, I'll get off my soap box now... :-X




Happy machining , Jeff Birt
 

Offline bowber

*
  •  216 216
  • Kirkby Stephen,Cumbria, UK
    • View Profile
Re: Accurate homing
« Reply #37 on: August 24, 2007, 07:20:29 AM »
My CNC is a small industrial unit, most likly used for training originally, it's a Denford Easimill.

I've just recently removed a Ahha system from my cnc and installed new drivers and Mach3, my system was setup with the Estop removing power from the drives and sending a signal to Ahha, so it lost steps anyway. It did remember the dro readings when shutdown but not when it crashed, which it seemed to do any time it got a bit of code it didn't like.
Mach does have a persistant DRO setting so when you shut it down and then restart it it does show the same position but with microstepping drivers you lose position anyway when the power is removed from the drives, not sure how this would work with feedback from encoders but again I'm not sure I'd trust the position.

I use micro switches for Homing and it's repeatable < 0.0005" so I'm happy with that.

So as has been said I think Estop is just that, an emergency! and I would never trust a computer or electronics to control that stop as they may have been the reason for the Estop.

Steve
Re: Accurate homing
« Reply #38 on: September 01, 2007, 12:20:25 PM »
I think my oldest son said it best" POP, If you build your machine and program your code to the same level of accuracy and dependability that you WANT the software to maintain then you will rarely have trouble(;-) "

The Estop and limits are there for a last measure to HELP prevent damage to the operator and machine not save your project from disaster

ANY time you estop or limit stop a machine, and it does not matter WHO made it, you should  reference it back to home to verify that it is still accurate.

If you are running into situations that require constant estops then you need to whack the programer in the head(;-). If you are constantly hitting the limits then you need to whack the operator in the head(;-)

There IS a SOFTLIMITS you can set up that will HELP  prevent the limit crashes. Heck it will even show you the softlimit boundary on the screen when you load the code. If your part goes outside the limits the operator should see it. Can't ask too much more from Mach(;-)

HUM i'll have to test the estop and limits to see if MACH really looses position. I know I have estopped and resumed without loss of position.  I don't recall having tested the limits situation.

(;-) TP


I get E-stop occasionally from the safety swing open perspex guard on my Super X3. I don't lose position. It only happens when I am changing a tool when not moving. No problems resuming, nothing lost. But we cannot assume microstepping is not screwed if power to stepper system goes off. If the E-stop just stops the drives without killing their brains, then there should be no problem, else reference it again. My limit (read reference) switches are optics, and resolve position edge to within the same narrow (1 step (not microstep) on the stepper I think) and repeatbility is excelent. Tight window on refence switch edge is important. Mach3 seems to get it right.
Re: Accurate homing
« Reply #39 on: May 17, 2010, 11:08:59 AM »
You are mostly concentrating on e-stops and accidental limits, and that is not the point. When you turn off the machine from one day to the next or for the weekend you want to home again so the the machine continues to run with accuracy (just as if have never turn it off). In the case of a lathe, the only way you can be precise is using the index pulse from the incremental encoder.

In my case I will use Teco drives. I read the documentation and I see the servo drives come with preprogramed home routines which use the index pulse. Has anyone interfaced mach and teco drives so as to take advantage of the routinces which come with the drives?