Re: Idiot Proofing a Proccess
You could add a simple switch that can read Fixture A or Fixture B. You could in the ladder write to G54.0 and G54.2 for either fixture. Then in the program you read #1100 and #1101 respectively. So If #1100 eq 1 then you have fixture A If #1101 eq 1 then you have fixture B. I can help you with the ladder if you need it.
Re: Idiot Proofing a Proccess
I think I am understanding what you are getting at, but I'm not sure. Can you explain this a little more for me? Thanks!
Re: Idiot Proofing a Proccess
At the fixture you will add something that will be sensed by switches. Based on which switch is being made that will indicate which pallet is there. You will have to find two free inputs that the controller can read. (X-Bits) Most machines have options that are not currently being used but the inputs are readily available. In the logic we read the input as a normally open contact and feed G54.0 coil in the ladder. The CNC sets macro variable #1100 to one when G54.0 in the ladder goes on. It does the same for #1101 when G54.1 comes on. You would write the logic so if both inputs are on you alarm out. So in your program it will read IF [#1100 EQ 1] GOTO 10. Line N10 would run the program for fixture 1. If #1100 was set to 0 then it would read the next line IF [#1101 EQ 1] GOTO 20. Line N20 would run the program for fixture 2.
Does that make sense?
Re: Idiot Proofing a Proccess
...yes, the way to build a machine is all logic. drdos time is a different story. Thx dr
2 Attachment(s)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
drdos
At the fixture you will add something that will be sensed by switches. Based on which switch is being made that will indicate which pallet is there. You will have to find two free inputs that the controller can read. (X-Bits) Most machines have options that are not currently being used but the inputs are readily available. In the logic we read the input as a normally open contact and feed G54.0 coil in the ladder. The CNC sets macro variable #1100 to one when G54.0 in the ladder goes on. It does the same for #1101 when G54.1 comes on. You would write the logic so if both inputs are on you alarm out. So in your program it will read IF [#1100 EQ 1] GOTO 10. Line N10 would run the program for fixture 1. If #1100 was set to 0 then it would read the next line IF [#1101 EQ 1] GOTO 20. Line N20 would run the program for fixture 2.
Does that make sense?
Yes, that makes perfect sense. I've heard of this being done but never been a part of something like this before. Am I looking in the right spot on where the switches will be wired to? Will these switches be just a basic two-wire switch?
Now if I am looking in the right spot, and it is a two-wire switch, would it be safe to say that I would wire one lead of the switch to say input 388 and then the other lead to 363?
And if I am correct in all that, what would the next step be?
thank you for all your help!
2 Attachment(s)
Re: Idiot Proofing a Proccess
Ok, I think I have this somewhat figured out. If I am looking into this correctly, I would have one wire from the switch going to #7 for +24v, and another wire going to #342 to engage x10.2 when the switch is closed. And then for the second switch, I would have one wire going to #7 for +24v, and another wire going to #343 to engage x10.3???
If this is correct, my next question would be the proximity switches I am looking at have 3 or 4 wires on them, what would the purpose of the other wires be? ground, or something else?
Re: Idiot Proofing a Proccess
Ok, talked with an engineer and have everything all figured out on the wiring side of things. Now, for the software changes, I will definitely need to be talked through how to update the ladder and anything else needed to make the control see and understand the switches
Re: Idiot Proofing a Proccess
Hy, if available, or afordable, use feed skip functions, like probing or torque skip / kindly :)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
deadlykitten
Hy, if available, or afordable, use feed skip functions, like probing or torque skip / kindly :)
I have gone in pretty deep with the switches and am ready to get them purchased and installed. Can you explain how a feed skip would work for me? the machine doesn't have a probe, so I can't verify the fixture by probing if that is what you are getting at.
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
ADVCOLE
Ok, talked with an engineer and have everything all figured out on the wiring side of things. Now, for the software changes, I will definitely need to be talked through how to update the ladder and anything else needed to make the control see and understand the switches
I am fluent with Macro B so getting the program itself to work will be no problem, I just need that #1100 and #1101 variables to be mapped to the switches.
Re: Idiot Proofing a Proccess
hy advcole, i just had a bit more time, and i read all the conversation : even if there was a probe, is waaaay faster to use a switch
Great idea drdos :) pls, will you share a few such switches examples ?
if possible, until then, to avoid mistakes, after machining, simply put all your work in line, thus to see the order : ababa ... still, mistakes can happen, but maybe will help
if you succed in implementing the switch, then, for the future, think of a :
... hidraulic clamping system to minimze in machine time for switching plates , and/or outside the machine, to reduce preparatory time; please, will you share a few photos with your setup ?
... solution to automize plates going in/out the machine / kindly :)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
ADVCOLE
I am fluent with Macro B so getting the program itself to work will be no problem, I just need that #1100 and #1101 variables to be mapped to the switches.
Can you post your ladder file?
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
ADVCOLE
Control: Fanuc 18i
Background:
We are running a high volume of a family of parts, so minimizing spindle idle time is key. In order to make things go fast and smooth, I made a subplate that allows hot-swapping of fixtures. I have two different fixtures, A and B. Fixture A holds 3 sets of parts in the family, and fixture B holds the other three sets.
Problem:
One program controls the machining of both fixtures A & B with an M00 in the middle to hot-swap them out. It's only a matter of time before someone pulls Fixture A out, goes to lunch, comes back, and then somehow puts Fixture A back in when it should be Fixture B.
Current Solution:
after the M00 there is a 4-second dwell, then an operator message "#3006=1 (Run Fixture B?)". The idea is to cause a break in their autopilot mode so that they can confirm the correct fixture. I am still worried about a problem arising though.
Possible solution:
Barcode the fixture and have a barcode reader that inputs the fixture number into a macro which then runs the corresponding program??? I assume I would need an rs232 barcode reader as this machine doesn't have USB, then as far as parameters go and how to get the scanner to communicate with the machine and finally get the scanned data into a variable is where I am lost.
Has anyone tried anything like this before? any other simpler, idiotproof solution I am overlooking?
...yes. on hill for climbers
Re: Idiot Proofing a Proccess
...send drdos a check for 500usd...money talks
Re: Idiot Proofing a Proccess
hy machinehop :) about the $$$, i was going to say that maybe you should let them figure it out
and then, another idea came : advcole, get 1000 from your boss, and 50-50 with drdos :) i think is a better advice, isn't it ? haha
also, there is another possible solution to identify pallets : run a short face / drill operation, that should exist only on one pallet, while on the other should cut air
check extra load, and if it's there, then is pallet-a, else pallet-b
*requires :
... load monitor
...... or
... custom application
and another idea come : negation of idea from above, thus, check is there is no load, then is pallet-b, else pallet-a
*requires :
... torque-limit, which is not torque-skip, thus is not a skip-function
...... or
... no load detection function
i don't know how / if those are possible on fanuc, but is a good idea, because :
... costs with skip function > cost with switch > cost zero if custom_spec is there
... time required for detection with skip ( tool change, probe, etc ) > time required for detection with switch ( on - off ) > time for detection with custom_spec ( coded inside the program, or monitoring )
* thus if spec is there, then it will run in background ( adding 0 time to your cycle time ), and it will only require configuration
i'm also proofing processes, and i use methods as those described above, and others
as long as those specs aren't there, then what drdos said is the best approach / kindly :)
1 Attachment(s)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
drdos
Can you post your ladder file?
drdos, here is the ladder file. I took a quick look at the ladder on the machine, and it appears to have things mapped to G54.0-G54.7 and G55.0-G55.7. As far as I can tell, those are all related to an APC system, which we don't have on the machine, so anything related to an APC can be overwritten to make room for the two switches. Thank you so much for all your help with this!
2 Attachment(s)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
deadlykitten
hy advcole, i just had a bit more time, and i read all the conversation : even if there was a probe, is waaaay faster to use a switch
Great idea drdos :) pls, will you share a few such switches examples ?
if possible, until then, to avoid mistakes, after machining, simply put all your work in line, thus to see the order : ababa ... still, mistakes can happen, but maybe will help
if you succed in implementing the switch, then, for the future, think of a :
... hidraulic clamping system to minimze in machine time for switching plates , and/or outside the machine, to reduce preparatory time; please, will you share a few photos with your setup ?
... solution to automize plates going in/out the machine / kindly :)
deadlykitten, here is the switch I will be using-
https://www.automationdirect.com/adc...y/dw-ad-713-m5
As for the fixtures I am not using hydraulics, but rather a vacuum chuck that I built, the system allows for very quick, repeatable fixture changes, and holding forces of over 2,500 pounds.
Below are pictures of the vacuum chuck, and the venturi vacuum system. The fixtures locate onto the chuck using a round and diamond pin. Sleeves on the chuck receive those pins and locate to about .0002"
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
ADVCOLE
drdos, here is the ladder file. I took a quick look at the ladder on the machine, and it appears to have things mapped to G54.0-G54.7 and G55.0-G55.7. As far as I can tell, those are all related to an APC system, which we don't have on the machine, so anything related to an APC can be overwritten to make room for the two switches. Thank you so much for all your help with this!
Just to make sure that I won't be stepping on toes, what bits are on for Keep Relay 6?
1 Attachment(s)
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
drdos
Just to make sure that I won't be stepping on toes, what bits are on for Keep Relay 6?
I think this is what you are looking for? K06- 00000000
1 Attachment(s)
Re: Idiot Proofing a Proccess
Here is the updated ladder. I tied X10.2 to fixture A selected and X10.3 to fixture B selected. #1000 = 1 fixture A in machine, #1001 = 1 then fixture B is in the machine. If both switches are ever on at the same time then you will get an alarm and the machine will not run. If you set Keep Relay K12.0 to a 1 it will ignore these changes.
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
drdos
Here is the updated ladder. I tied X10.2 to fixture A selected and X10.3 to fixture B selected. #1000 = 1 fixture A in machine, #1001 = 1 then fixture B is in the machine. If both switches are ever on at the same time then you will get an alarm and the machine will not run. If you set Keep Relay K12.0 to a 1 it will ignore these changes.
Thank you very much! Now I just need to get my boss to order those switches. In the meantime, while I am waiting for the switches, I assume I can just put a jumper between +24v and x10.2 or x10.3 to verify the ladder file and that the variables are switching between 1 and 0 as they should?
Re: Idiot Proofing a Proccess
Quote:
Originally Posted by
ADVCOLE
Thank you very much! Now I just need to get my boss to order those switches. In the meantime, while I am waiting for the switches, I assume I can just put a jumper between +24v and x10.2 or x10.3 to verify the ladder file and that the variables are switching between 1 and 0 as they should?
yes. I recommend that you do that...
