Restoring a BP series II CNC

[unterhaus]

I asked for this part of the forum and here it is!

I bought a BP series 2 almost 3 years ago. It didn't have any control hardware left, just the cabinets. I'm putting brushless servos on in place of the missing steppers. Much of the hardware is done, mostly electronics and control left. Not having it done is driving me nuts.

It's going to be run using EMC. www.linuxcnc.org


I'll post pics soon.

[unterhaus]

Sorry for no pics, couldn't get a camera. I've been working on the wiring. I have the spindle axis motor wired, suppose I should check it for functioning. I'm using a PacSci brushless servo with a SC752 drive. I have the fixed ram version of the Series II with a ballscrew built into the spindle.

The X axis has a PacSci motor and an AMC controller. The Y axis is all electrocraft. I always wanted matching motors and AMPs, but it just hasn't happened. I also intend on putting a servo on the knee, but it's a monster. I'll get around to that later (yeah right).

To start, I'll be using the encoder built into the motors to position the machine, but later I plan on adding linear encoders.

[unterhaus]

I made a revolting discovery tonight. The motor I had been planning to use on the X axis does not have hall feedback as I thought, but has a tachsyn. Either I have an identical motor sitting around with hall feedback, or I'm going crazy. I was trying to figure out why the amp kept faulting. I measured the power supply to the feedback, and the 5v was reading 2.6v. Then I hooked up a smaller motor that does have halls and it spun like it was supposed to.

I'm Happy to see something move though.

[NEATman]

unterhaus-
I'm in the start of a simmilar process on a Bridgeport V2XT. It's a 1993 machine, mechanically in excellent shape, probbably due to the fact that it always had computer problems. :-)

Anyway, I'll be following your postings carefully, as I am interested in how the linuxcnc.org controller software works for you. I wonder if this poses a problem as I think the amplifiers that you will be using will have the analog ±10volt control signal.

I am researching (in all my spare time, yeah right) the right type of controller with all of the options that I want for this machine (spindle control, direction reversal, mist coolant, etc.) I haven't found much yet that is compatible with the analog signal amplifiers, yet is still affordable. Do you know if the Linuxcnc will allow you to disable the amplifiers and use the encoder signals as a Digital Readout for manual milling? What will you be using for a "breakout" board to interface with the computer?

I'll keep looking for pictures of your machine. In the meantime, here is a pic of my machine. I want to make the machine manual/CNC, and I have already modified the covers and added handles to the ends of the ballscrews. Also notice the 5hp 3 phase motor that is on top of the cabinet, it's the 3 phase converter for the spindle motor. If you are interested in the handle adapters that I made, they may work for your machine. If you are interested, I have them modeled in solidworks, and I could send you the file.

I apologize for the ramblings, but it's nice to see someone else on here working on a Brideport.

Thanks,
NEATman

[unterhaus]

NEATman,
Unfortunately, I have a totally different bed setup than you. For example, the X axis ballscrew is in the saddle, so the handle doesn't move with the table like it's supposed to. I have seen people put handles on, maybe I will if the CNC doesn't progress fast enough, but I'm feeling like things are starting to move again.

It's funny that you replied to this when you did. I had just come to the conclusion that I needed some extra circuitry to interface to my drives, in the form of relays and such. But right now, that's all out the window unless I reconsider my new direction.


As I mentioned in my previous post, I just discovered that my X axis motor didn't match the drive I intended to use. So I went looking for drives in my collection. It turns out that I have some drives that would take either analog control signals or step/direction. So now I'm thinking seriously about going step and direction. That means I'm now in the market for a parallel port breakout board. Previously I was going to use an 8 channel analog output board with 48 digital I/O's which I bought on Ebay. I was going to use that in combination with a Kaluga encoder interface board. The disadvantage would be I would have to write and maintain my own EMC driver. This way, I just have to spend money. There are a lot of schematics out there for the limit switches, now I just have to find them.

The problem with analog control signals for drives is that I am cheap. The most commonly used board for this purpose is the Servo-To-Go. They've been promising a PCI board for years now, but the available card is ISA. It's around $800 without breakout boards. You can get the Phoenix type boards for around $50 each, and you need two. There is also the Pico Systems board from Jon Elson. There is also the vigilant board supported by EMC for around $600 with breakout boards. http://vitalsystem.com/motion/index.htm
I was seriously thinking about that board. If you want a lot of digital I/O something like this is the way to go, although EMC will support a second parallel port.

As far as EMC working as a DRO, I don't think it will. I'll ask on the mailing list.

Thanks for the picture. I'm glad I'm not the only person with a Bridgeport covered in parts. I have to work on storage space now that I'm going to try to start moving the machine.

Eric

[ESjaavik]

I used EMC with steppers for my TOS FN20 conversion.
The drives are 9A and 12A 70V from Phytron. No breakoutboard, as the drivers have optoisolated input/outputs.

So far I'm not happy with this solution. I can get it up to F500 before it droops out with following error. This on a 1GHz Pentium. Just dragged in a Shuttle PC with 2GHz today to see if I can tease it higher. You'll get the same problem with step/dir inputs on servo drives I guess. The pro is you can make electronic handwheels by driving the inputs directly.

I may switch to servos later, but have to see if my drives are compatible with EMC. As my drives have their own current/speed/position PID loops that may be a problem since EMC has this built into it.

Leave out the handwheels unless you can make them like the ones originally mounted om my mill. They were without handles, just disks with no protrusions. And with a dog-cluch and a spring so they would only follow the spindle by the drag force. (The mill had a quite fast auto-feed originally) The handwheels can be a real danger on mills of this size!

I will install either electronic handwheels or a joystick. I just have to see first whether I go for servos or keep the steppers.

The limit switches: I must admit I haven't installed them yet. And so far I've hit the clamps/vice a couple of times, but not the limits. That's when you realize it's just a computer and it cannot see your clamp. If you tell it to go straight through it, it will. But it's highly unlikely even a 15mm cutter will come out on the other side.

[NEATman]

unterhaus/ESjaavik
I have been researching Rutex, as they offer a 25 pin breakout/mother board R990mb board for $58, and a relay board R990io for $68. Also, they offer the R991H servo drive for $98 each, which converts step and direction to analog ±10v. This way I could retain the original drives. If one should burn out, I could always buy a complete brush drive from Rutex, and it would plug right into the motherboard inplace of the R991H.

As far as software, I'm considering Mach2, because it has the neat feature of a touch probe to map and copy parts. Also, it seems to have a pretty good rating among the threads that I have here. I am awaiting a response about the question of using the computer as a DRO and doing manual machining. Thanks for asking about the manual option with EMC. I'm interested in the answer.

Has anyone else ever done this??

If this is possible, I will definetly go this way. I make lots of simple parts that I don't want to have to program, as it would be a waste of time. But, I want to have the ability to generate 3d contours and many other intricate shapes, along with engraving.

As far as the limited I/O on the parallel port, there may be a way around that with Mach2. I was reading the manual last night, (available at www.rutex.com a 4 mb download and 200+ pages) and they talk about a keyboard emulator that were originally intended for video games. They allow additional keys/switches to be programmed as hotkeys, so they could be connected to you additional I/O. I need to re-read it again, but it sounds feasable. Also, for the price, the breakout board, relay card, and three drive signal converters, $420 is a pretty good price. There is a review of this combination in the review section of the site.

Have either of you considered this solution?

As far as the machine covered in parts, I think it's more common that we think! Something about all flat surfaces collecting junk...

Keith

[unterhaus]

Originally Posted by ESjaavik I used EMC with steppers for my TOS FN20 conversion. The drives are 9A and 12A 70V from Phytron. No breakoutboard, as the drivers have optoisolated input/outputs. So far I'm not happy with this solution. I can get it up to F500 before it droops out with following error. This on a 1GHz Pentium. Just dragged in a Shuttle PC with 2GHz today to see if I can tease it higher. You'll get the same problem with step/dir inputs on servo drives I guess. The pro is you can make electronic handwheels by driving the inputs directly.

I never thought of driving the inputs directly, I was just skipping over the part of the manual where it talked about controlling the drive with an external encoder. I might try this. I was going to use an encoder interface board to get the handwheel into the computer.

Originally Posted by ESjaavik I may switch to servos later, but have to see if my drives are compatible with EMC. As my drives have their own current/speed/position PID loops that may be a problem since EMC has this built into it.

I think this can be tuned so it will work.

Originally Posted by ESjaavik I will install either electronic handwheels or a joystick. I just have to see first whether I go for servos or keep the steppers.

I've been using a Haas a little, and I must say that the handwheels are really almost essential.

Originally Posted by ESjaavik The limit switches: I must admit I haven't installed them yet. And so far I've hit the clamps/vice a couple of times, but not the limits. That's when you realize it's just a computer and it cannot see your clamp. If you tell it to go straight through it, it will. But it's highly unlikely even a 15mm cutter will come out on the other side.

Maybe this is why nobody has a schematic of how their limit switches are hooked up.

[ESjaavik]

I agree that some kind of handwheels are needed. Maybe a joystick, but I never tried it. It would of course have to be an analog one so not only the direction can be controlled, but also the speed. And separately for X and Y, so any diagonal move can be done. That's a pro. The con is that a handwheel is more familiar.

Either of them can be put in a separate box and placed where it is easy for you to observe the tool and workpiece. An Estop button should of course be included on this box. Anyhow, that's refinement. For now I bring the keyboard with me around the mill. It gets soaked with suds and chips, but the cheap ones are $5 each, so no big deal.

A handwheel can be an encoder into a microcontroller that just emits step and direction impulses depending on where you moved the handwheel. This would have the same "feel" as a handwheel without the force feedback that you have with a mechanically coupled handwheel. There are a couple of electronic problems by driving the motors by handwheels. They will generate voltages back into the drives. They don't like that. Sometimes to the extent that they get p***ed and quit working.

Disconnecting the motors from the drives during hand-milling is too much of a hassle, and doing it is easy to forget.

Have to go back milling. I'm squaring off a cast iron stump with no initial shape. It's a pig to clamp down!

[unterhaus]

Keith,
I've thought a lot about the Rutex products, and been very tempted to buy from them. Every time I went to their web site to look the price of the step/direction to analog converter it was just too much money for me. It's a personal problem, but I'd rather spend twice as much and get a full analog servo interface. I'm pretty sure I've seen people using the converters. I remember answering questions about them on the CCED email list. Lots of people buy systems with motors and amps but no control, so it makes a lot of sense to get the system up quickly. In my case, I had a blank slate and went with my prejudices. I'm a controls engineer, so naturally I wanted the highest performing system that I could tweak to my heart's content. Now I just want to get things running, things are looking a lot different, and I probably would have been a lot better off had I just spent the money on Rutex or Geckos and had a running mill a year or more ago.

I didn't get an answer about using EMC as a DRO yet, however the first step would be to have some sort of encoder interface with the computer. Step and direction systems don't provide this, this is another reason why a full servo system with the feedback routing through the computer is attractive. Of course you can always drive it by hand through the computer.

If I get my system running with step and direction, I'll probably try Mach2 to see how it works.

Eric

[unterhaus]

Keith,
I asked about the DRO. Assuming you get an interface that reads encoders and has an EMC driver, you can do it. The only trick is to set the following error high enough that it doesn't estop, i.e. the error has to be bigger than the travel of the machine. You could have another parameter file with a normal following error for CNC work.
Eric

[NEATman]

Eric-
Thanks for checking that out for me. I am a mechanical engineer, and the controls part is where I struggle. Where I work, we create multiple axis extremely precise custom machines, but we have highly trained controls people to design and install that portion. (the machine I just finished designing is 12 axes, 5 of which are air bearing). I'm still learning, but from what I can tell, for what I intend to do with my machine, I think that the Rutex set-up will work out well. Plus, It'll sure beat the 286 PC that is on the machine at the moment. I am thinking of staying with the old drives, as the motors that are on my machine are higher amperage than the rutex drive can output. (65A @112V) Plus, it would save me $300, and I would still have the option of selling the used converters, and buying a new amp.

I am interested in the EMC interface. Is there a way that I can demo this without being a software engineer? From what I had seem on the EMC website, that programming went right over my head. I this something that you have tried out already? Or is this something that needs to be tweaked and set-up for each individual machine?

I had thought that the following error values would be a problem. I guess that the other option is to also install software that is like a 4 axis DRO. I could use a parallel port switch box to change between CNC and Manual, and just switch between the different software. Do you see any problems with this strategy?

Say, You don't work for Aerotech do you?...

Also, my parents were going to name me Eric, until they saw my red hair. I had a bad enough temper without being called Eric the Red.

Keith