The Fanuc encoders are designed to handle the incredible noise from welding. Tig produces even more noise than this. You can see in the video that even the rail(shuttle) the robot is riding on was told to move and that encoder is much further away. Because all the motors(axis) are moving during the crash, this movement was created by the controller CPU. High frequency weld current running through the robot, due to improper fixture ground, could cause the cpu to perform erratically. The originator of the video said they had to change some cabling to fix it. I believe the trunion(fixture) was not fully grounded and the robot control cabling became a shared weld ground.
Inadequately sheilded welding stuff for welding use. The encoder is innocent until liquid, a surge, or bearing and bushing shrapnel invade the sensor wheel area.
@@ERICtheLATE Your answer gave me an endorphin rush lol i maintain and teach these types of robots but without the welding aspect to it and when he said "Why?" my head starting rushing through any possibilities and then i came across your comment and was like Ah haa!! 👍
As a mechanical engineer intern, I had the same experience a month ago... I had the chance to revive an old Kawasaki laser cutting robot without the laser head. To demonstrate its capabilities to my boss, I duct taped a pen to the end of the arm and taught it to write down the company's name. I left it there for the weekend, turned off, and when my boss came in on monday, I wanted to surprise him. Of course the encoder battery died, so it forgot its zero point, and smashed the pen into the table with max force... I've never been more embarrassed in my life before.
On our university we've got a ~30 year old ABB robotic arm with a stained LCD (think of Game Boys with their screens going out) and the encoder battery dead too. If you unplug it, be sure it's at most a day or less before power goes back, or else you calibrate the damn thing all over again. I don't wanna know what's gonna happen after this pandemic with a year and half of no classes there.
Saw a video about the future of automation and they were talking about combining humans and robotic machines to provide the best of both worlds. They showed a factory with a lot of robots running and I was like Nope! If you think that you're going to have humans working with robots on the same station simultaneously, then you know nothing about robots. Unfortunately I was already 20 minutes into the video before discovering that these people don't know what they're talking about. I know of no factory where humans are allowed inside of a robot's work area, due to safety considerations.
@@jfan4reva there are these things that companies are calling "co-bots" which are meant to interact with people in a working environment without causing harm if someone gets too close. I saw a demonstration at Hannover Messe in León México, but as of now, they're more like a gimmick still in development... I'm no engineer, but that's my grain of salt which might be totally wrong.
@@jfan4reva Collaborative robots, they are specially made for that purpose, they are not your average industrial robot. In general, don't know about the specific video you saw however.
I saw the same reaction on Cnc, some encoder's cable pick up the noises comming from the torch and the controler though he is out of position, so he move the robot to correct. Hope you solved the problem.
@@zakel5750 seems like there should be some kind of "sanity check" to the encoder positions so if there becomes a large and physically impossible error the system halts instead of trying to blindly correct for position. Also, you would think that robotic arms designed for use in high emi environments would have optical encoders and balanced cables to reduce interference
When it come to cnc crashes or "f***ups" I've found there's an "oh-moment" 50% of the time. Never under estimate a humans ability to overlool their mess-ups
I used to maintain P-100's and M-3 Fanuc robots. (Long time ago!) These sort of things were almost always wiring issues. (Except with one certain brand of plasma cutter. They made so much odd impulse noise you couldn't shield the system well enough. A change to a different brand solved that problem.) Our ABB robots were really solid welders, but they were package purpose built systems, not integrated systems. I'm still amazed at just how fast robots can move during a crash and how hard they can hit. Glad I don't have to figure this out. I don't miss those headaches!
If someone is standing within the reach of the robotic arm then it is their fault for tempting fate. You should always assume from the moment you power the arm that it is hell bent on killing everyone and destroying everything within its reach as fast as possible.
Looking at the dry run, I was thinking "wow, that robot has some serious accuracy issues". Then I figured that the continuous oscillation was not Parkinson's but the way the weld head is supposed to move while welding.
Yeah, wider beads and pretty caps run with a specific oscillation to expand the puddle and move it around using the electromagnetic field produced at the tip. Its probably the main reason I'm a terrible welder unless its TIG.
The edge in that "Why." It's definitely frustrating when you have a recurring problem in a complex, EXPENSIVE machine, and deadlines to meet and a mountain of ongoing problems to solve. I wonder if it took some heavy damage during that little mishap, or if the problem is recurrent enough, accumulated damages.
Did you not notice the red flashing light after the incident - so somewhere in the system it had detected an error - so hopefully the error level is set low enough to not do permanent damage to anything too expensive.
This is why EMI / EMC protection is required and very important on drive, signal cabling, and digital interfacing of such robotics. If all else fails a watchdog should kick in, i'm sure the watchdog did kick in, here in this scenario - stopping a more catastrophic error from occurring. Usually that can be achieved by looking at commanded position and encoder feedback, when the error is out of range, to trigger a full halt on the machine - most likely what was seen in this video. There must have either been a cable fault or EMI issue from the welder where electrical noise caused havoc on the arm controls, as soon as the arc started, the machine went crazy... Lucky no one got hurt.
Judging by the dates on the comments, this video got picked up by the algorithm. Amazing demo of why there are safety perimeters around industrial robots. And how dry runs don't show everything.
As a mechatronic technician, I'm sorry ut a You are wrong. This is an example how not to programming a robot. The safety stuff around the robots is for the apes who don't understand that a robot goes straight forward without cheking if something is in his way. Excuse my english and grammar, I'm European and no one teach me the American Accent
@@anatomie6340 As an engineering specialist for automation, with couple of years of coding and developing in robotics, I'm sorry ut a You are wrong. You are an example that you just don't know, that you don't know... It's probably lack of experience.
When i was doing some welding job in my bike, the automatic light in the garage repeatedly turns on eventhough there is no moving object in front of the sensor and it doesn't see me and the welder at all. Welding machine causes very strong interference around itself both electrically and magnetically.
@@MarinusMakesStuff Yeah I'd bet you're right. We had a grounding issue. We believe it caused the encoder to pass incorrect position info and it "corrected" directly into the side of some racking. Luckily it happened before our down shift where we're able to make repairs without operations asking us "How much longer?" every 30 seconds.
I took my Elkie into a car stereo place when I was 19 to get the stereo working. I opted to go a bit more powerful with it, got a 200W amp put in. I got the keys back after paying, got into my truck, started driving. The noise was horrendous. The stereo had a whining sound that was directly linked in pitch to the engine RPM. I knew exactly what it was, but instead of fixing it myself, told them they need to. They had it for about 45 minutes when I finally lost my patience and asked if I could go in to show them what was wrong. They let me, and I simply disconnected the amp power wire from the battery and re-routed it so it WASN"T SITTING RIGHT NEXT TO THE BLOODY DISTRIBUTOR CAP and the problem was fixed. Wires induce currents and accept outside currents. You got a lot of current flowing through the mig feed that's probably leaking electromagnetically across to the arm's motors.
Me: TH-cam algorithm: You just gotta see this blast from the past, I've been saving it up just for you Reality: youtube hijacking videos from unknown creators and sending them viral to make lots of advertising revenue they don't have to share
The welder generated electrical noise that was picked up on the encoder circuit. This caused the robot servo drives to "compensate" for this positional error. Ensure your jig, robot, controller, cable shields and other controls are properly grounded, especially with large bonding straps to minimize impedance. Try to avoid creating ground loops, when possible, as these can actually work against reducing electrical noise. Also check that your welding source is properly isolated and is not shorting back through the arm on a high(er) impedance path. Lastly, while probably not related, make sure that the encoder backup batteries in the base of the arm are good. I've seen errors on order arms from various maintenance deficiencies where the batteries seemed to let the controls "ride-through" a slightly flaky connection.
Seems to be a lot of thoughts the welding electromagnetic fields are causing this but maybe as it looks like there are many other welding positions on the automatic fixture manipulator it may only be on this one scenario. I wonder if there’s a welding setting that’s got a typographical error like 200 instead of point 200? That only becomes active after the torch is enabled?? I have had many calls for water jets cutting weird and the issue was the tool compensation was set at .4 instead of .04 causing it the be about 3/8” off. Just something else to check into as a potential solution.
Check your ground cable on your welding machine, make sure you have a good ground. It looks like and soon as your start welding or strike an arc you are shorting out.
I agree with other commentors that improper ground is likely the culprit. If you advance the video to 1:41, and click the "." (period) key three times, you will see a light on the robot originating from the fixture or under the fixture. Press the "." (period) key once more, the arc from the wire to the work piece begins.
Don't turn your geo metro into a robot. My guess of why the robot freaked out. the high voltage created a magnetic field that messed with the Halifax sensors ( or what ever sensor the robot uses for position awareness) .
I used to program Fanuc robots of all sizes at a GM taillight manufacturing facility. Hot melt dispensers, parts shuttles, die part extraction, etc. The sheer speed and power of these machines is awe inspiring and to be revered. Linear tip speed in relation to machine size darn near defies physics. Absolutely defies any and all reason. Haha Man... they’re capable of tremendous speed and acceleration.
that awkward moment when integrator angel on the left shoulder advises you to ring out dgnd line, while technician devil whispers to halve hf ignition's frequency to see if it goes slower. (and in per-hour pay they switch places)
looks like SABO robotic arm... i dont know why it needt to shake like that... i understand that Z or whatever welding pattern.. but why does it not have some oscilator that would be much cheaper to repair than the robot ?
I wonder if, after the arc started, this was the first time the robot had run at full speed :) I think it should enter Wimbledon as I'm sure it was faster than Emma Raducanu.
Was it the electric /magnetic field generated that messed with the sensors ? As it seem to happen exactly as there is a current surge at the starten of the weld
Any noise on the rotary encoder lines will cause this due to the encoders signal being disrupted. Most of them are pulse type sensors and most welding uses pulse, and always is using a rectifier, which creates a pulse.
I am no robot expert or anything but it appears to be a ground issue because as soon as it stuck the arc it wigged out so that would be where I started. I only mention this because I think it would be interesting to learn what actually caused the crash if I am wrong.
If a dry run works then why use the teach pendant to do the actual run. Try running from the main console not the teach pendant. That looked like a controlled crash rather than noise induced axis runaway. If that crash is repeatable and in the same place then it is unlikely to be noise effecting the plc,that would cause random movements in any axis. Also the plc would likely have error flags. What error /alarms are shown? We need more information if you want remote diagnosis. Industrial machinery repair technician.
I'd be curious about that current your pushing through those cables attached to that arm, looks like you sunk a bolt.into that cable cover.... you sure that bolt, did not wear through some of the internal cables.... I'll bet you have cable drama.... looks like a terrible dressout...
The fact that it only does it while it's welding should help you solve the problem. But honestly, you could have a combination of problems. Question I have is, is this it's original setup? What has been replaced from it's original setup? Did it run continuously before? Troubleshooting questions you need to ask yourself
🤨👉 "You've been replaced by the Fast & Furious Funac 10,000 🤖 ...now get out !" 😔 Later that day... 😒 "Er, so...um, what extra perks and benefits did you have in mind ?" 😏👌
Why? EMC! Almost sure a ground loop picking up the EM welding pulse. The emitting loop area can't be reduced coz of the robot arm, so it is the receiving loop area that needs to be pushed to zero. That's why ground loops are really a NO-NO in this theater.
The Fanuc encoders are designed to handle the incredible noise from welding. Tig produces even more noise than this. You can see in the video that even the rail(shuttle) the robot is riding on was told to move and that encoder is much further away. Because all the motors(axis) are moving during the crash, this movement was created by the controller CPU. High frequency weld current running through the robot, due to improper fixture ground, could cause the cpu to perform erratically. The originator of the video said they had to change some cabling to fix it. I believe the trunion(fixture) was not fully grounded and the robot control cabling became a shared weld ground.
Bullshit you have no idea what you are talking about.
@@tonyabbot6771 I have been servicing Fanuc robots for 27 years. What is your opinion of what happened?
@@bradfitz999 I think he was kidding 🙄
@@tonyabbot6771 Yeah, Tony, what's your expert opinion?
@@pchansen100 I think you're right. Now that I read it as sarcasm it's kinda funny.
this guy : "why..."
me : i feel you
Indeed, the disappointment was palpable.
A lack of electrical isolation if you ask me. As soon as that arc touched off the s*** went berserk
@@dustinmcdermont699 that's what i was thinking. It ran the dry run perfectly. I fight with machines for a living, and i can feel this guys pain.
The same with my diy cnc plasma cut, when the arc fire, the controll board go crazy...
As a hobbyist, that *why* was brutally felt
Me signing up important documents
How did it go?
Classic
"Ok were gonna do a dry run first"
Or my debit card
I'm a miserable welder; but I think this robot has me beat at bad welding.
neah, still better
Don't be too harsh. It was probably up all night drinking again and just immediately fell asleep once it had to do some actual welding.
Please cheer up quickly ;)
I'm sure there are worse than you - like those that were welding when they fell of the structure and died.
Inadequately shielded encoders for welding use
Exactly
And controller board goes wild. Some "engineers" have little imagination about noise level in that applications.
Inadequately sheilded welding stuff for welding use. The encoder is innocent until liquid, a surge, or bearing and bushing shrapnel invade the sensor wheel area.
@@ERICtheLATE potahto potayto.
@@ERICtheLATE Your answer gave me an endorphin rush lol i maintain and teach these types of robots but without the welding aspect to it and when he said "Why?" my head starting rushing through any possibilities and then i came across your comment and was like Ah haa!! 👍
That pause, then "why." Not laughing at his broken robot, but damn, comic timing was strong. 😂🤖
Successful test of the red warning light.
As a mechanical engineer intern, I had the same experience a month ago... I had the chance to revive an old Kawasaki laser cutting robot without the laser head. To demonstrate its capabilities to my boss, I duct taped a pen to the end of the arm and taught it to write down the company's name. I left it there for the weekend, turned off, and when my boss came in on monday, I wanted to surprise him. Of course the encoder battery died, so it forgot its zero point, and smashed the pen into the table with max force... I've never been more embarrassed in my life before.
On our university we've got a ~30 year old ABB robotic arm with a stained LCD (think of Game Boys with their screens going out) and the encoder battery dead too. If you unplug it, be sure it's at most a day or less before power goes back, or else you calibrate the damn thing all over again.
I don't wanna know what's gonna happen after this pandemic with a year and half of no classes there.
@@Kalvinjj either some teacher is kosi a weekend, or aome students are going to get a surprise lesson in dealing with old tech hands on.
you could have saved face by saying "this is what will happen if I don't get that raise we discussed"
Bet you supprised him all right.
devil's in details. sad you malfunctioning robot didn't write company mane on boss'es corpse. but when you read it is says 'kill all humans'
Good reminder that human flesh is the softest part in the workshop.....
Saw a video about the future of automation and they were talking about combining humans and robotic machines to provide the best of both worlds. They showed a factory with a lot of robots running and I was like Nope! If you think that you're going to have humans working with robots on the same station simultaneously, then you know nothing about robots. Unfortunately I was already 20 minutes into the video before discovering that these people don't know what they're talking about. I know of no factory where humans are allowed inside of a robot's work area, due to safety considerations.
@@jfan4reva there are these things that companies are calling "co-bots" which are meant to interact with people in a working environment without causing harm if someone gets too close.
I saw a demonstration at Hannover Messe in León México, but as of now, they're more like a gimmick still in development...
I'm no engineer, but that's my grain of salt which might be totally wrong.
@@jfan4reva Collaborative robots, they are specially made for that purpose, they are not your average industrial robot. In general, don't know about the specific video you saw however.
@@jfan4revapeople don’t realize there’s absolutely no benefit to being human when considering the potential of machines
Really dont know. Possibly a short somewhere, but we did fix and repair a few cables that possibly were grounding out.
I saw the same reaction on Cnc, some encoder's cable pick up the noises comming from the torch and the controler though he is out of position, so he move the robot to correct. Hope you solved the problem.
Does this robot have through arc seam tracking?
Sounds really like a grounding/shielding issue, since the crash happened as soon as the arc started.
@@zakel5750 seems like there should be some kind of "sanity check" to the encoder positions so if there becomes a large and physically impossible error the system halts instead of trying to blindly correct for position.
Also, you would think that robotic arms designed for use in high emi environments would have optical encoders and balanced cables to reduce interference
@@BycraftWelding even if the traking is as wrong as it can be, it would never do this. In a flat weld it can only track vertical anyway.
Whyyyy???
The reaction is priceless! 👍🏻
You know it wasn't you fault when you say "why" after. If you know exactly what did it the second after it happens, it's your fault.
/CNC programmer
This is very true
When it come to cnc crashes or "f***ups" I've found there's an "oh-moment" 50% of the time. Never under estimate a humans ability to overlool their mess-ups
The dry run was fine, to me that would suggest the programme was OK. Maybe a fault elsewhere?
@@ACERASPIRE1 It was OK until it started welding - my money would be on some electrical interference from the welding head
Any electronics fail
/ CNC electronics repair
I used to maintain P-100's and M-3 Fanuc robots. (Long time ago!) These sort of things were almost always wiring issues. (Except with one certain brand of plasma cutter. They made so much odd impulse noise you couldn't shield the system well enough. A change to a different brand solved that problem.) Our ABB robots were really solid welders, but they were package purpose built systems, not integrated systems. I'm still amazed at just how fast robots can move during a crash and how hard they can hit. Glad I don't have to figure this out. I don't miss those headaches!
Looks like you had some EM feedback that messed with your sensors.
I think so to...
I really love the collaboration shown among technical folks like us.
Wow that was fast. Imagine if it came straight at your face.
If someone is standing within the reach of the robotic arm then it is their fault for tempting fate.
You should always assume from the moment you power the arm that it is hell bent on killing everyone and destroying everything within its reach as fast as possible.
@@jackaw1197 It is what they all dream of when not switched on.
@@sichere The literal arms of skynet.
Yep. And that's not even a particularly fast one. They're lethal in the wrong hands.
Looking at the dry run, I was thinking "wow, that robot has some serious accuracy issues". Then I figured that the continuous oscillation was not Parkinson's but the way the weld head is supposed to move while welding.
Yeah, wider beads and pretty caps run with a specific oscillation to expand the puddle and move it around using the electromagnetic field produced at the tip. Its probably the main reason I'm a terrible welder unless its TIG.
This was more suspenseful than most movies.
The edge in that "Why." It's definitely frustrating when you have a recurring problem in a complex, EXPENSIVE machine, and deadlines to meet and a mountain of ongoing problems to solve. I wonder if it took some heavy damage during that little mishap, or if the problem is recurrent enough, accumulated damages.
"Robots never complain, take bathroom breaks or lunch, etc, you just program them and they run all day"
@@kealanocarroll9738 Lol yep until they don't.
It’s unlikely that it was permanently damaged.
Did you not notice the red flashing light after the incident - so somewhere in the system it had detected an error - so hopefully the error level is set low enough to not do permanent damage to anything too expensive.
@@smeric28 end effector is definitely a goner though
Also possible there is a grounding issue with the robot or welder where voltage is flowing where it shouldn't be going.
This is why EMI / EMC protection is required and very important on drive, signal cabling, and digital interfacing of such robotics. If all else fails a watchdog should kick in, i'm sure the watchdog did kick in, here in this scenario - stopping a more catastrophic error from occurring. Usually that can be achieved by looking at commanded position and encoder feedback, when the error is out of range, to trigger a full halt on the machine - most likely what was seen in this video. There must have either been a cable fault or EMI issue from the welder where electrical noise caused havoc on the arm controls, as soon as the arc started, the machine went crazy... Lucky no one got hurt.
Possible short circuit. Fried the board at the first sign of electricity. Improper ground?
Improper ground, I would say
Judging by the dates on the comments, this video got picked up by the algorithm. Amazing demo of why there are safety perimeters around industrial robots. And how dry runs don't show everything.
As a mechatronic technician, I'm sorry ut a
You are wrong. This is an example how not to programming a robot. The safety stuff around the robots is for the apes who don't understand that a robot goes straight forward without cheking if something is in his way.
Excuse my english and grammar, I'm European and no one teach me the American Accent
@@anatomie6340 As an engineering specialist for automation, with couple of years of coding and developing in robotics, I'm sorry ut a
You are wrong. You are an example that you just don't know, that you don't know... It's probably lack of experience.
@@JoeDhirk Yes it is definitely lack of experience...if you define Experiance as the sum of all mistakes someone did;)
On the first dry run going back and forth i had thought the robot was programmed by Micheal J Fox
This fast?! Must be welding current on the motor. Never seen something like that
I agree, either interference/insulation problem, a short or grounding issue.
They actually do move that fast; however, I would have to believe that the welder's power is somehow interfering with the servos.
When i was doing some welding job in my bike, the automatic light in the garage repeatedly turns on eventhough there is no moving object in front of the sensor and it doesn't see me and the welder at all. Welding machine causes very strong interference around itself both electrically and magnetically.
@@MarinusMakesStuff Yeah I'd bet you're right. We had a grounding issue. We believe it caused the encoder to pass incorrect position info and it "corrected" directly into the side of some racking. Luckily it happened before our down shift where we're able to make repairs without operations asking us "How much longer?" every 30 seconds.
I took my Elkie into a car stereo place when I was 19 to get the stereo working. I opted to go a bit more powerful with it, got a 200W amp put in.
I got the keys back after paying, got into my truck, started driving. The noise was horrendous. The stereo had a whining sound that was directly linked in pitch to the engine RPM.
I knew exactly what it was, but instead of fixing it myself, told them they need to.
They had it for about 45 minutes when I finally lost my patience and asked if I could go in to show them what was wrong.
They let me, and I simply disconnected the amp power wire from the battery and re-routed it so it WASN"T SITTING RIGHT NEXT TO THE BLOODY DISTRIBUTOR CAP and the problem was fixed.
Wires induce currents and accept outside currents.
You got a lot of current flowing through the mig feed that's probably leaking electromagnetically across to the arm's motors.
Me when I try to walk down the carpeted stairs quickly in socks.
Me:
TH-cam algorithm: You just gotta see this blast from the past, I've been saving it up just for you
Reality: youtube hijacking videos from unknown creators and sending them viral to make lots of advertising revenue they don't have to share
Advertising on youtube? Never seen such a thing ;P
It's truly the wild west of copyright and intellectual property rights with video content.
Positive from welding torch derivation with signals wires from robot.
dokładnie :)
As someone who welds occasionally but not a lot, can confirm, test beads are nice, when you try to do actual work, CHAOS. 😂🙃
EMI from welding arc. that's why.
Every Robotprogrammer be like: "Ouch, know this feeling" 😂😂😂
Indeed 🥺😂
The welder generated electrical noise that was picked up on the encoder circuit. This caused the robot servo drives to "compensate" for this positional error.
Ensure your jig, robot, controller, cable shields and other controls are properly grounded, especially with large bonding straps to minimize impedance. Try to avoid creating ground loops, when possible, as these can actually work against reducing electrical noise. Also check that your welding source is properly isolated and is not shorting back through the arm on a high(er) impedance path.
Lastly, while probably not related, make sure that the encoder backup batteries in the base of the arm are good. I've seen errors on order arms from various maintenance deficiencies where the batteries seemed to let the controls "ride-through" a slightly flaky connection.
Love how fast these things can move, it's incredible.
Wow ya think the power pixies took a path they weren’t supposed to?!
Seems to be a lot of thoughts the welding electromagnetic fields are causing this but maybe as it looks like there are many other welding positions on the automatic fixture manipulator it may only be on this one scenario. I wonder if there’s a welding setting that’s got a typographical error like 200 instead of point 200? That only becomes active after the torch is enabled??
I have had many calls for water jets cutting weird and the issue was the tool compensation was set at .4 instead of .04 causing it the be about 3/8” off. Just something else to check into as a potential solution.
any high frequency getting through shilding?
And that folks, concludes your daily reminder that you are a squishy, slow moving, fragile organic being and that machines.... *are not.*
HV from welder messing up the signal line?
man: "why?!"
robot: "absolutely not."
I have had my own disasters with robotics so I feel you.
Let me guess. Encoder went crazy because of the welding arc?
My teachers always told me that it wasn't the robot crashing by itself, it was my bad programming.. starting to think it was the robots all along
I felt that why he said “why?!”
Why? Because it wants to be a dancer, not a welder dammit!
*_Отправьте его в отпуск, пусть отдохнёт!!!_* 🙄🙄🙄😎😎😎😎😎😎
Check your ground cable on your welding machine, make sure you have a good ground. It looks like and soon as your start welding or strike an arc you are shorting out.
Me: testing code.
Customer: Opens application:
I hate when my elbows slip too
I agree with other commentors that improper ground is likely the culprit. If you advance the video to 1:41, and click the "." (period) key three times, you will see a light on the robot originating from the fixture or under the fixture. Press the "." (period) key once more, the arc from the wire to the work piece begins.
Don't turn your geo metro into a robot. My guess of why the robot freaked out. the high voltage created a magnetic field that messed with the Halifax sensors ( or what ever sensor the robot uses for position awareness) .
I used to program Fanuc robots of all sizes at a GM taillight manufacturing facility. Hot melt dispensers, parts shuttles, die part extraction, etc.
The sheer speed and power of these machines is awe inspiring and to be revered. Linear tip speed in relation to machine size darn near defies physics. Absolutely defies any and all reason. Haha Man... they’re capable of tremendous speed and acceleration.
Something must not have been grounded properly. As soon as the weld started it crashed. I think electrical noise interfered with the arm.
Horizontal axis drop down & electric trip.....
that awkward moment when integrator angel on the left shoulder advises you to ring out dgnd line, while technician devil whispers to halve hf ignition's frequency to see if it goes slower. (and in per-hour pay they switch places)
It's been 3 years. Let me guess. The problem was grounding or a power related issue indirectly connected to grounding issues.
Judge: Sign yer marriage certificate here please!
Me:
Ahhhh, now I see why you need to stand well back when the machine is working.
Me, practicing to talk to the girl (perfect)
Then Me, talking to the girl (crash)
That robot just completely electrocuted itself
I love his succinct query...
Well, at least the red blinky warning light worked well when it all went to crap!
Actually a good question in the end.
Sheesh the robot scared it's self.. It obviously wasn't ready for all those sparks... hehehe..
looks like SABO robotic arm... i dont know why it needt to shake like that... i understand that Z or whatever welding pattern.. but why does it not have some oscilator that would be much cheaper to repair than the robot ?
Electrical induction through unshielded cables?
I wonder if, after the arc started, this was the first time the robot had run at full speed :)
I think it should enter Wimbledon as I'm sure it was faster than Emma Raducanu.
For once, it's not "expensive machine crash" clickbait.
As to "why?", it's because it knew you wanted it too much.
That’s Jim Carrey trying to write that the pen is red when in fact it was blue
Robot is like: AH! BRIGHT LIGHT!
Was it the electric /magnetic field generated that messed with the sensors ? As it seem to happen exactly as there is a current surge at the starten of the weld
Any noise on the rotary encoder lines will cause this due to the encoders signal being disrupted. Most of them are pulse type sensors and most welding uses pulse, and always is using a rectifier, which creates a pulse.
If you count the frames from when the welder started and when the arm started moving, it was 0.23 seconds.
I am no robot expert or anything but it appears to be a ground issue because as soon as it stuck the arc it wigged out so that would be where I started. I only mention this because I think it would be interesting to learn what actually caused the crash if I am wrong.
One more phobia. It can happen in manually mode (T1), when you teaching them
If a dry run works then why use the teach pendant to do the actual run. Try running from the main console not the teach pendant. That looked like a controlled crash rather than noise induced axis runaway. If that crash is repeatable and in the same place then it is unlikely to be noise effecting the plc,that would cause random movements in any axis. Also the plc would likely have error flags. What error /alarms are shown? We need more information if you want remote diagnosis. Industrial machinery repair technician.
These are not the droids you are looking for...
Although it's just a robot, it kind of hit me :(
robot's just trying his best but he's sick and needs help
Things getting worse when the robots take this literally
Why did you set that up to be CNC welded in the first place? That looked like a hand job to me.
Looks like it shorted itself out somehow. Grounded not wired properly somewhere?
I'd be curious about that current your pushing through those cables attached to that arm, looks like you sunk a bolt.into that cable cover.... you sure that bolt, did not wear through some of the internal cables.... I'll bet you have cable drama.... looks like a terrible dressout...
If you go frame-by-frame you can see the arc in the cable just before the tip turns on
I think that's just rolling shutter, you can see it again at the last frame of weld flash where it's only visible on the bottom half of the screen.
I’m a CMM programmer. I can’t tell you how many times I’ve said “Why?!”, just like he did.🤣
God damn that's incredibly dangerous
The fact that it only does it while it's welding should help you solve the problem. But honestly, you could have a combination of problems. Question I have is, is this it's original setup? What has been replaced from it's original setup? Did it run continuously before? Troubleshooting questions you need to ask yourself
Me watching this going wtf part 2? And the last second was like ohhh Dayum yeah he dead.
Looks like someone needs one of those metal hoops with the cable threaded through twice to stop any interference.
Electrical noise problem maybe? This must have been debugged, does anyone know the result?
I tried to develop some monitoring electronics for a welder once.
I gave up because the EMF it generated was a nightmare
Looks like the spiritual successor to ED-209
This is a good reminder that you shouldn't fuck around with industrial robots and motors, even when they're correctly functioning.
🤨👉 "You've been replaced by the Fast & Furious Funac 10,000 🤖 ...now get out !" 😔
Later that day...
😒 "Er, so...um, what extra perks and benefits did you have in mind ?" 😏👌
If welding was like talking to other humans this would be me.
HAHA, this sucks, I know your pain. Sometimes I think its just me that deals with this kind of stuff.
the robot got spooked by the welder spark
This is me every time trying to print documents in a hurry.
So, what caused it?
...subscribed. Please post more stuff.
The frustrated "why" moment that every technician knows.
"What the fuck" and "how the fuck" usually come next
EMI WINS. Flawless victory!
Of the three colored lights, a red light means I'm dead
Why? EMC!
Almost sure a ground loop picking up the EM welding pulse.
The emitting loop area can't be reduced coz of the robot arm, so it is the receiving loop area that needs to be pushed to zero.
That's why ground loops are really a NO-NO in this theater.
The robot had a heart attack during work, because he’s been made to weld all the shitty jobs no one else wants to do.😂😂😂
Robot: Goodbyes cruel world.
Guy: Why?