Posts by StoopidEngineer

G'evening All - does anyone have experience with using the torque limiting features/commands with Epson robots? I've got a T6 that I'm going to be using to load a machine tool, and need to make sure the part is seated properly. Inconsistent part length (less than 0.010") means the 'load' position needs to be variable, or my tool needs to be compliant in the one direction. Rather than make the EOAT more complex than it needs to be, I'd like to accomplish the same end result in the program.

I'm looking at the 'LimitTorqueLP' (Limit Torque in Low Power Mode) statement - am I simply going to get collision detects all the time if I use this command when loading the collet?

If that's the result, can I turn off collision detect, set the speeds and torques way low, drive to the 'stall' point when loading, grip the part with the collet, release the part from the gripper, turn off torque limiting and turn collision detect back on, then resume the rest of the cycle?

I guess my question is 'what does the Limit Torque function actually do?

Thanks all!

Maybe I'm reading too much into this, but I want to make very sure about this before I just throw power into a cabinet:

I'm swapping a 1994ish LR Mate 100i with an R-J2 controller for a much newer LR Mate 100iB with an R-J3iB controller. I opened the R-J3iB and was pretty surprised to see that the main power terminal blocks are for L1 and L2, so I looked at the nameplate on the controller and it is called out as 220V single phase, 60 Hz. Was this a controller bought with the option, or are a lot of the Mate controllers 220V single phase?

The R-J2 coming off is three phase (but I'm not 100% sure of the voltage) and it looks like we're pulling power from the drop into the CNC cabinet.

So am I going to need a 220V single phase drop to power this? Or is there a setting/configuration somewhere that means I could run this on three phase? I ask because some CNC machines can be wired for 1 or 3 phase, and even automatically detect the power they are wired to.

Dropping a single phase line isn't the end of the world, but our facilities guys are pretty busy right now.

Also, if anyone has the installation documentation for an R-J3iB controller, I would appreciate it. I got all of the standard manuals from the reseller we bought the robot from, but not the install sheet I'm used to seeing with new robots (i.e. the double sided page of 'DON'T DO THIS' stuff Fanuc ships out with new robots these days.)

Thanks all!

Currently have a mechanically shot robot with an R-J2 controller and will be swapping it out for a newer replacement arm with an R-J3iB controller - is there an easy way to take a backup from the R-J2 and upload it to the R-J3iB? I understand the struggles of backing up the R-J2, which I am able to do, but didn't know if I could export programs as ASCII then load them onto the R-J3iB, or if there was some other easier way to do it, rather than have to type it all in on the TP again. Thanks All!

Fanuc will sell you most manuals, but will ask questions if you try to buy others: I recently purchased a basic user/controller manual for the R-30iB controller with an LR-Mate 200iC (I think it was $100) and they didn't ask any questions. When I tried to buy the mechanical maintenance manual and electrical maintenance manual, I had to tell them I have taken Fanuc classes in-person at Rochester Hills, and worked for ASIs for 10+ years before they would let me purchase them. (I think the mechanical manual was $100 and the electrical was $200.)

I know it makes the most sense for the error you're seeing to be the reason your line is down, but I actually have that error permanently displayed on one of our LR Mates, and the robot and cell run fine. The error has been showing for (apparently) as long as we've had the machine- ~8 years. And does not negatively impact anything. I'd probably check other issues on yours as well.

Quote from kluk-kluk

As long as the robot you buy has the same number (the A05B-XXXX-XXXX one) there should be no problem

You will however have to master the robot.

All our robots are second-hand and we have a lot of robots running with different F-numbers.

If you have the possibillity to reprogram the cpu-board and the servodrive is compatible, you sometimes can run a different robot.

We even have a M16-i RJ-3 robot running on a RJ-2 cabinet.

That's good to know - thanks! I really doubt I'll find another 100i that has the matching model number, due to the age, if nothing else, but I'll keep my eyes open. Thanks!

Hey Gang - So I've got an LR Mate 100i High Speed that is completely shot mechanically. It's been in service for 25 years and has run literally tens of millions of parts, but is so whipped it won't reliably hold position and has incremental positioning errors until it crashes, is re-taught....rinse and repeat.

So I'm shopping for replacements, have quotes for a couple of used robots (some drop-in replacements, some newer stuff) but was wondering if anyone has ever swapped an arm without changing the controller.

I've never done much with servicing robots, especially ones this old, so I don't have any experience with it, but my current company is trying to pinch every penny right now, and it looks like I can get an arm for dirt cheap on eBay (yes, I know.) Before I dig too much - is it even possible to have a robot and controller with different F numbers? Am I going to run into non-stop issues?

I'm just trying to offer all the potential solutions I can, or at least be realistic about the options.

Thanks All!

I would strongly suggest considering a more robust mechanical solution (something like Nexen's rack and roller pinion system) rather than what looks to be a standard straight rack (14 1/2° pressure angle?)

At the very least, I'd suggest adding an oiler to the system with a visible/monitored reservoir - something like a way oiler from a CNC machine, for example.

Hey guys - working on a machine tending cell that's got some serious hours on it, and I've run into an issue I can't seem to explain.

LR Mate 100i with an R-J2 controller. Loading a Fanuc Robodrill from a automated tray handler loading pallets. Pallet placement is good and consistent - loaded onto pins and is not moving, and is repeatable between pallets. Robot is securely fixtured and is not moving or rocking during installation. Both single axis moves and multi-axes moves have been tested and verified to repeat within original robot tolerance. (It's actually incredible how well this thing repeats, considering how old it is.) I checked everything with a dial test indicator on a mag base, set up to catch different points and different axis moves - I'm comfortable the robot is repeating well. It performs correctly within the machine, loading/unloading the fixture.

Where I run into trouble is the pallet work - if I had to guess, I'd say the X,Y coordinate of the robot is not aligned with the X,Y of the tray - sometimes the robot is off to one side or another, and sometimes it is dead nuts in line, depending on where it is in the pallet. The pallets have not been verified by me, but have been in service for years without issue - no new pallets, no changes to the program. But there was a pretty serious crash a couple of weeks ago, and they've been fighting with it ever since.

There is also some math to compensate for pallet locations, because they're not quite an even pattern, but again, they've run literally millions of parts through this system with all the same hardware and software, and are only have trouble after the crash. On the R-J2 (my first time working on one) there are no Tool/User/or Jog frames set up. The pallet system is...not what I'm used to on the newer stuff, but seems to be fairly straightforward and make sense.

I guess my question is: if we managed to crash this hard enough to throw off the X/Y alignment of the robot coordinate system to the tray coordinate system, and there aren't and have never been any frames set up in the robot, is there a way to re-teach/correct the alignment of the robot to the tray?

If this was a new robot running PalletTool or something similar, I know the process for teaching the pick point coordinate system (using the teach plate w/3 points) but this is all working in the same frame in the program.

I can get some screenshots of the TP in the portions where its looking at the pallet later if that helps.

Thanks all!

That's the thing - we don't use PMC (nor have I ever seen anyone use it that I'm aware of) but I feel like it could be used a lot more than it is. And the class happens to take up the other 3 days of the week that I would be going out for a 2 day class, so I didn't know if it was worth checking out.

I've had good luck with Fanuc training at Rochester Hills as well, but I've heard the other facilities don't do as good of a job.

Hey gang - Title says it all: Is Fanuc PMC training (at the Rochester Hills Facility) worth the time and money?

I've got a couple of certificates from previous/basic trainings, have done some of their online stuff, and am actually asking about this one because I'm looking at going out for a couple of days for a maintenance class, and the PMC class would fill out the rest of the week. It's a bit of a drive, so going for just 2 days doesn't make a ton of sense to me.

LMK if you've taken it, or if you think it's useful at all.

Thanks!

Hey all - looking for some help. I'm not with an ASI anymore, so I don't have the 'build and price' option available in my CRC website. I'm looking to get a price on a CR series robot, and I know the base price, but I don't even know what options to ask for at this point.

Pretty typical assembly applications, nothing vision (right now) or anything else that's crazy. Would like to know what i need to have the robot be a standalone unit (and act as a PLC) and also have it work with a PLC (AB Micrologix)

Help me understand what I don't know at this point.

Thanks all!

Hey Gang - I'm far from what I would call a Fanuc expert: I've been to a couple of Fanuc training weeks at Rochester Hills, and have done a handful of integrations. This question is really rooted in curiosity more than anything.

Recently, I've been working on a couple of Universal Robot integrations, and I very quickly got sick of the limitations of programming it in the UR GUI on the teach pendant. Unfortunately, their offline program is only a Virtual Linux machine that runs the teach pendant software - nothing in the way of simulation, etc. And it has all the same limitations as working on the physical teach pendant in the GUI.

Because of what I saw as the lack of basic functionality in the programming (it's not possible to call a sub-program from within a sub-program, for one) I was advised by UR that the best course for more advanced programming was to do it in their base robot language of URScript, which is based in Python. I've learned enough Python to be dangerous, as well as work around some of the issues I was having with the UR GUI programming, but I found myself wondering:

Just how useful is being able to do some/any programming for a Fanuc in KAREL? I heard with the move to the R30iB controller it is now an option you have to purchase just to be able to access anyway, but I've never had a reason to even look at the KAREL programming in the past. The only time I can think of when it would be useful is when I was doing some PalletTool programming and the needs of the application were pretty far outside what PalletTool would handle out of the box.

It seems that TP programming (either on the physical Teach Pendant or in RoboGuide) is way more capable than the UR equivalent. Is it typical or expected to have to get into the KAREL programs at some point anyway? How useful is it to be able to do? Any good examples of projects that got into the KAREL programming, and why they wouldn't have worked with out it, or why it just made it easier?

Thanks all!

I asked Fanuc about that when I was talking with them, and they apparently only sell these as a pair for the LR Mates. They're a Beta motor, so it's not like they're a fortune, but it would have definitely been nice to just have to swap the encoder.

Fanuc also doesn't service these, so if someone knows of a reliable group out there that does, or has a process for testing these, I'd be interested to hear about it.

Thanks!

I had an encoder go bad on an LR-Mate (SRVO-138) resetting power did the trick for a couple of days, but by the time I found out about it, cycling power had no effect. Swapped the motor/encoder, did a single axis master, all is well.

But now I have a motor sitting on my desk with a big orange NFG sticker on it, and I'm wondering if it's worth getting into the encoder end. I did pull the cover off just to look, and nothing on the board looks smoked, there's no obvious damage to the rotor, and if I didn't know it was bad, you'd never know it was bad.

Anyone ever troubleshot one of these? Is there any point in even attempting to figure out what went, or trying to repair it?

The robot is 10 years old, but it's a clean environment and it's not a 24/7 operation.

It's not my money either way - I'm just hoping to learn something, and potentially put a motor on the shelf as a spare.

Thanks All!

Hey gang, trying to wrap my head around total cost of ownership for a robotic weld cell, and I need some help.

For those of you who have experience with them, what's the actual cost per 8 hours for a typical robotic MIG weld cell, including consumables (wire and gas) with a welder running ~65% of the time? I'm not trying to get to the penny (although I am being specific on purpose) but would like to get a decent idea if possible. Order of magnitude (are we talking $500 per 8 hours, or $2,500 per 8 hours?)

Thanks All!

-Jason

For the curious: I'm an Automation Engineer looking at ROI on a weld cell for a customer who's a friend, but have't ever run a weld cell in production.

Good Afternoon All,

This is something I've been wondering about for a while, and although I think I know what the answer would be based on somewhat parallel industry experience, I'd like to see if anyone has any direct experience.

I'm a CNC guy, and have run CNC machines on rotary phase converters for years without issue (although a lack of maintenance on the RPC can cause issues with line balance and/or voltages.) Now I'm wondering about running a robot on a phase converter when 480V 3 phase isn't available.

Anyone ever tried it successfully, or unsuccessfully?

Thanks!

Cobenson - thank you very much. That's a lot cleaner solution than what I was thinking (because I just don't know any better yet.)

Thanks!

Depending on how large and/or complex your part to be deburred is (and a lot of other variables) you may want to consider using a program like AutoDesk's PowerMill Robot, which is actual CAM software with a plug-in for machining with a 6+ axis industrial robot. You set up your cell, drop in your part, and program the operation just like you would a CNC machine.

From my experience using a robot to deburr, I would advise against it if the requirements are for tight tolerances or good surface finish. We had a customer request a precision deburr of a part that was manufactured on a simultaneous 5 axis machine (extremely complex geometry in hardened tool steel) with very tough surface finish requirements, and the robot we were using was not nearly rigid enough to deburr at the speeds we needed with the programming options we had available to us. Both of these issues might have been avoided with the software I described (or one of the other CAM programs capable of this) but they are not cheap at all.

I agree that the mechanical fix is the right one, but the direction I'm getting is to fix it in the robot program. (This might be driven by already having spent money we didn't have budgeted to fix a problem caused by the customer running a product not in the original scope.)

I'll take a look at changing tool frames on the way to the pallet, because I think it actually already does this, based on my observation today.

Thanks All!