[LeeWay]

I haven't seen any more info on that one yet.
Likely too new. I imagine we will be seeing more and more of these hit the market in the next few years. It may indeed pay to wait to automate.

here is a pretty good site.
Cobots Guide | Making Sense of Collaborative Robots

[WOTDesigns]

Thanks, Ray.
I know yours is out of the way and I also like the PDB. The fact that you can quickly change collets is an excellent feature. Makes installing a new collet or R8 tools a non issue.
I also like the way the arm stores out of the way.
I think I have about a year left to pay on the lease for the Pulsar and then it's our's. I will be keeping it and using it for some of our less needed parts, R&D etc.
I will be getting something new or used, but it will need a PDB and tool changer.
I have been considering the Servo Pulsar with your equipment on it. I am setup already with tooling and vises for that.
I am also considering buying a small HAAS or perhaps a Trak 2OP. Both are fairly expensive, but with tool changers. Better accuracy and speed, though since I upgraded to Path Pilot, the Pulsar meets our needs. Parts are probably cheaper on the Pulsar as well. I would be looking at about $20K as opposed to at least double that for tooling on the other two mills.

I mainly mentioned the robotic arm just so that it could possibly do double duty to change tools and swap parts. It would have to be strong and agile though. No doubt about that. I think it could live on the left rear outside the enclosure though where it would store tools and blanks to load and unload.

I would think long and hard about the 2oP before buying. They limited the heck out of the control. No macros, tool offsets aren't stored on the machine, small envelope. I was set up to lease one and finally realized after sending sample programs it couldn't run my parts. They want it to ONLY be second op. The controller is practically an Allen Bradley input interface.

On the plus side... They do their own leasing. VERY easy lease. Put a good chunk down but the payments were cheap enough and you re up each year or buy out.

Having run a Haas control for the last couple years I'm hosed. Even my Fanuc control from 1996 have more options and better interface than the 2oP. It's for second op... Only.

Sent from my SM-G935V using Tapatalk

[LeeWay]

That is the kind of information that is very valuable. There is a dealer in Pensacola and they were wanting to bring one by and demo it next week. They are taking it to Mobile anyway and wanted to stop by and show it here too. We are really too busy and it is too soon for us to have a demo, though that would have been cool.
I did mention to them that I might be interested in a lease, but I would be more interested in just buying a used machine outright if it was cost effective for us. Of course, they do not give out any details until they are shaking hands. I don't want to waste anyone's time on speculation.
It is a shame that they limited this machine. The small screen and minimal controls had me curious. If the buttons ain't there, what's missing. It does have a small envelope, but so does the Pulsar and it fits our needs well.
I will be researching this a lot more before I decide on what we need sometime next year. As I mentioned, the mill isn't the bottleneck right now, but still needs to be automated to allow for more automation in the future. Who know's, I may design a gadget that everyone wants and would need more machine time.
Thanks again for the heads up.

[LeeWay]

Well, I scored a Scorbot ER III off Ebay very reasonably. $750 including freight shipped on a pallet. It is a teaching robot and I will be learning with it. It didn't have a controller and that is likely a bonus anyway since I will be operating it with Arduino. There are several of these bots already being powered under Arduino control just at first glance on the internet. That and the fact that I found the original manual

http://www.theoldrobots.com/book45/ER3-Manual.pdf
for it with all the specs and diagrams including schematics was the reason I went for it.
Once I get it operating, there are already a few things I would like to try. The first would be to add some sort of sensors to add to the safety of the arm. Not sure if that will work, but I can certainly add switches to the arm on both sides to estop or feedhold it.

I think version 3 of this bot only has lie 6 or 8 position encoders. Later versions have increasingly higher resolution encoders and it appears that was just about the only updates to them until like version 8 or 9.
I have the Arduino starter kit already and an Arduino Mega something coming. I know it will need drivers as well. More research for that. I will be setting up an assembly work station with this. I have ideas on how to make it all work very well from a screw chute to part placement and tool manipulation.
This machine will certainly teach me some necessary skills that will be valuable moving forward.

This does come with a small conveyor and a turntable. Don't know that I would use those, but at least the conveyor has an extra servo and encoder on it.
This one is named Bud, but he will be answering to a new name at some point. Not sure what. Let's have some ideas on a name for him. Or is it her?
Here are a few images.

Attachment 341374


Attachment 341376

[Steve303]

Good buy Lee. Can't wait to see what you can do with it. I can see a few uses for the conveyor like moving parts to a second workstation.
Steve

[SCzEngrgGroup]

Lee,

The motors on that are Polulu gear-motors. To be blunt, they are crap. The encoders are low resolution (typically 16-64 PPR, depending on the specific model), but since they run through high-ratio gearboxes, the effective resolution is much higher - as much as 6400PPR at the output shaft, depending on the gear ratio. So, resolution should not be an issue at all. Those gearmotors are, both electrically and mechanically, very fragile, and very easily damaged. A short current overload will fry the armature, and mild mechanical overloads will break the gears. You NEED to drive them with H-bridges with current sensing, and and don't allow them to be over-driven. I would strongly recommend using series resistors as a current limit fail-safe. Or better still, simply replace them with something more robust. The arm looks mechanically reasonably robust, though I'd bet you find that "precision" is not within its vocabulary, as there will be considerable backlash in all the axes, all adding up to a considerable overall positional uncertainty. I think positioning will be "approximate" at best, making operations like getting a screw into a hole hit or miss at best.

Regards,
Ray L.

[LeeWay]

Thanks Guys. Ray, something like this?
https://www.amazon.com/L298N-H-Bridg...idge+for+servo

I assume I can stack these on the Mega board? I would need 3 of these, correct?

I will know a lot more about what I have when it arrives though. Probably take a week or two to get here.
By then I hope to have a few arduino lessons completed.

I think it will be a good starting point at least for a learning tool. I am certain that it will be good enough for that. It only has a 1 kilo weight limit, so it isn't a monster, but much more for the money than those printed robots using RC servos. Some are completely just store bought camera mounts put together.
Do you have a source for some better motors for this, Ray? I would like to do some research on some.

[SCzEngrgGroup]

Lee,

Send me your e-mail by PM and I'll give you some recommendations.

Regards,
Ray L.

[LeeWay]

I will be using an arduino for this project to start with. Mainly because I already have some Arduino stuff and one of my goals is to learn how to use it. The second goal will be to interface it with the arm of course. I am also researching control software for arduino and some on Linux. Linux makes sense because I think it can control 9 axes as well as having other outputs.
I already have a laptop with linux installed. Gettuing my feet wet with linux and Path pilot helps to some degree.
I will be sure to report any progress here. If anyone knows of some decent teaching software on any platform, just post here. I would like to research as much as I can. Even Microsoft is getting into the robot control thing.
Not sure how good that is, but it is what I am most familiar with.

[LeeWay]

Do you guys know what type of motors or joint bearings these little collaborative arms are using?

Maybe harmonic drives? Planetary gears? I would like to take a closer look at how they work. This scorbot has a caveman (in robot years) style wrist mechanism. It certainly functions but could be done much better. In fact the later scorbots do have all refined joints. Maybe I can find a manual for those.

[LeeWay]

Answered my own question.
https://www.youtube.com/watch?v=nj1vO3cP7ug

Now these are not hobby class mechanisms.

[jfong]

However you can find harmonic drive gear heads cheap on eBay if you look long enough. Received these two HD gear heads for $55 last week. These are true zero backlash cfs series harmonic drives and not the hpg planetary version. I have a hpg one that I use as a 4th axis. The hpg works well but does have some backlash.

These came with Teknic brushless servo motors even. I've already tested the servos out on the bench. Not really sure what I will use these for. They were too cheap to pass up.

[LeeWay]

Those are nice. Great score. They look like a usable size too.
I have searched ebay some, but more to see what they were, how they work and what sizes might be available. Those two might be perfectly sized for the swivel base and shoulder joint on an arm. Maybe smaller ones in the rest of the arm.
I seriously doubt I would get into these unless they fell into my lap. There are other ways to make more compact joint than the Scorbot er3. That said, just simply researching those right now. I think I will have my hands full just getting the Scorbot moving.

[jfong]

As far as getting the Scorbot moving....since you want to learn and use Arduino, you need to make several servo drivers. One of the easiest brushed encoder servo drivers with open source Arduino code is this one
https://www.youmagine.com/designs/dc...ware#!comments

I tested this out a couple years ago and was able to drive a small servo relatively easy. Short video here
https://youtu.be/aauSVJ1OSt8

There has been some updates to the code but haven't tested the recent version. I use regular gecko servo drivers for my CNC but that is a pretty expensive solution. You can get lucky on eBay and get good servo drivers for cheap. I once bought 8 new applied motion SV7 brushless servo drivers for $20 each. They also can drive brushed motors too. Standard step/dir inputs too.

If you need help with the arduino stuff, just PM me. I've been working with micro controllers for many many years.

[LeeWay]

Thanks a bunch for that. The code will certainly help. I have three arduino's. The original, the mega and a Due. I will do some testing and learning on the little one and use the mega for my spring winder project. The Due will be for the robot. I will be using H bridges too.
Just been going over some of the more simple basic tutorials and while I can certainly handle them, I have never used a breadboard before. I always twisted wires and taped them up or used screw terimals. I did solder when I knew what I had will work.
These breadboards are pretty nice to be able to layout the circuits on. They are handier than a pocket on a shirt.
Now what I really need is time. It's always the case. Can't find the time to implement that time saving machine.

[LeeWay]

I did score a few decent parts on Ebay. Not too bad on the cost. Planetary gear boxes.
It should do great for a shoulder and elbow joint. Has a lot of very useful parts and switches etc. I am curious to see what the motor drive board looks like. Looks like steppers with encoders on these two motors. I don't foresee any issues using a mix of steppers and servo's.
I have also been thinking and during my research, it occurred to me that for our assembly bot project, a scara type bot might prove to be better and more accurate than a flailing arm type.
Maker is coming out with one, but honestly you can get the industrial ones off ebay working and with the controllers and software for about the same price. The maker type would be relatively easy to build yourself with off the shelf parts. Any ideas what that silver canister is on the bottom stepper?

Attachment 341932

[Novakon]

This is some pretty cool stuff. The difference between men and boys is the price of the toys!

John

[LeeWay]

Yep. Some cool stuff, John.
I have always been impressed with guys like Ray that can design the stuff that they do. I have learned a lot over the years, but generally steered clear of electronics, especially controlling them. I have been stalled out at least a few times now on projects that require this kind of knowledge, so no time like the present to start tackling it. I am also behind the curve with servo's and encoders. One will remedy the other.

Still wondering about that cylinder between the motor and blue gearbox. Brake? Clutch? Doodad?

[daniellyall]

pull it to bit

[LeeWay]

I will figure it out when it gets here. Not sure how long it will take.
I think this was some lab equipment robot or at least part of one. As far as surplus robot parts go, it's good quality stuff. Just about perfectly sized for what I am wanting. I will probably just use the Scorbot to get it interfaced with Arduino. I will have an easier time doing it with that I think. Then move onto designing one to use specifically for assembly tasks.