It's a 1984ish Wells Index 810, complete with 1980's CNC controls. I'm going to ditch the old electronics and retrofit modern motor drivers and use PC control for CNC. This mill was built as a CNC so there is no way to manually operate it w/o using the CNC. I'm pumped, I've looked forward to this for a long time! :grinpimp:
Looks like you've got your work cut out for you. Not even a quill lever???
Programming a z axis is not for the meek...have you looked into what software type programming is in it currently??? Some of the old stuff while primitive can also be real easy to use. Just a matter of being able to understand/navigate through it..
I was reading through the literature for programming the unit and I don't want to even try messing with it. It uses punched tape, but it also has a serial interface that you can use from a PC if you want. You must load the operating system tape when you power it on, too. I don't want to, I'd rather use the PC to run it No quill lever, it's 100% electrical
do you know what you are going to use to run it yet? i have been using mach 3 for a few month now and i am able to figure it out. i don't have anything to compare it to but if i am able to figure it out, it must be easy
Mach3 and Gecko 320 servo drivers! Errr I mean yes, I think I know what I need. If it weren't for Mach and Gecko I never would have even considered something like this for home use ($$$$). I haven't figured out which break out board to use but there are a few choices. This mill has limit switches and zeroing switches as well so I need inputs for those, too. I can't get 3 phase power, either, so I need to get a VFD or some other such device.
JM -- note the posting is in WI. I'll be in Milwaukee this week, but no way to get a machine like that back in my luggage! :flipoff2:
Up in WI I'd expect he might have stolen it for under $2K.... down here in Texas, hard to find anything like that. When all the oil went bust in '95 there were a ton of surplus machines. I bought my old Index Milling Machine (Index 555 w/power feeds) for $2K then, but it had a nice $800 DRO fitted to. Built in the early 1960's it still is a nice machine.
Nice find, and looks like a clean machine. On the RS232 interface, often the interface chips ($5 item) are blown but replace a couple of IC's and you're in business again. One guy (name slips me) makes opto isolation boards for this purpose (he is over on a Machinist board) as it is so common. So if you got a CNC machine due to $5 chips being blown.... I hate you! :flipoff2:
I started ordering parts. I ordered 3 G320 Geckodrives (www.geckodrive.com) and a VFD to run the 3-phase spindle motor on single phase 240vac. I am sifting through all the various manufacturers of breakout boards and it's looking like it will probably be a Campbell combo board (http://campbelldesigns.net/Combo-board.php). I don't have specs on the servo motors, but I know the controller on the machine now runs up to 100VDC, the Geckos are only good for 80VDC, so I need to find a power supply. I could make one but it's the same cost to get one with a nice mounting plate and some voltage regulators for +5 or +12vdc.
This weekend I'll start stripping what I can and cleaning the light rust off the ways. It's got a Bijur mist coolant pump so I'm sure that needs to come apart. I might just set up a flood system so I have some chance of keeping the shop a little cleaner. It also has a pressure oil system for all the ways and half of the connections leak so I need to address that this weekend, too. Hopefully I have all the parts by the following weekend so I can start wiring the panel!
Kwik-switch 200 actually. Seems like a little bit of an oddball but there are a lot of tool holders on ebay and I got a box full. I got a couple of collet holders as well but no collets so the search is on for replacements. I also got some kinda fancy thing for measuring tooling:
I plugged it in and the large indicator runs up and down. What I don't know is if the tool and holder are supposed to fit into the tapered hole somehow?!?!
And I ordered the breakout board today from Campbell. I talked to the guy on the phone today and he says the USB Smooth stepper should be out in the next couple weeks so that would be nice to add on.
yes that's a quick setter put the holder down in the taper it will hold it in place and allow you to tighten and loosen tools and set the heights using the DRO. You should also be able to zero it out at any time and use that to adjus the height if neccessary
I ordered one from Bob Campbell today. It was between those two but I decided the extra cost on the Campbell board was worth not having to wire up two more power supplies, and it looks like it will have a daughter-board setup for the Smoothstepper when it comes out, which should be fun and exciting
I still need to figure out the power supply for the servos but I think I've got that nailed.
Just a little update. I have enough stuff now to make the thing move, so hopefully I can do a little wiring this weekend. I'm using a 1500VA toroidal transformer, which will be good for a rectified ~72vdc. I was going to go bigger, but I can add another if I need to after testing. The guy I bought it from made it sound like 1500VA would be enough, but we'll see. In theory I should be able to supply 60 amps with the Geckos at 20 amps each.
Naval Jelly worked awesome to remove the rust on the ways!
HP D530, P4-2.66ghz, 512MB ram, 40GB hdd, CDROM, full complement of I/O ports, $155 from http://www.geeks.com. I was surprised, they shipped it right away and it's a very nice machine for way cheap $. I bought the XV15 new back in the day, NEC made some kick ass monitors:
Control cabinet is cleaned out:
I'm going to reuse most of the components in the power cabinet-contactors, stepdown transformer, etc. This is also where the VFD will live (blue thing sitting in the bottom):
Here's a shot of the Gecko G320s and the Campbell break out board. I can't get over how small these Geckos are. The Campbell board has relays on it, will run 0-10vdc to control speed on the VFD, and has its own isolated DC power supplies. It's a nice board. The G320's are good for 20amps @ 80vdc. I considered Rutex and others that will run higher voltage (my servos will do 100vdc) but I'm not willing to pay 5x or 10x the cost of the Geckos for 20vdc.
Here's a shot of a G320 next to the original servo drivers.
I need to weld up a shelf to mount on the machine to hold the monitor, keyboard, and mouse. If all goes well I should be able to get the axisisisess moving this weekend.
Yeah I know... My home desktop PC is P4-1.8Ghz though so it seems like a screamer! It should run Mach most excellently. As luck would have it I bought two more very similar PCs (Dell, however) from a local business. Same specs but they were $25 each!
I just got in from the shop and have some progress to post! We had a few good weekend days without rain so I had to get take care of some other things and didn't make much progress on the mill... until this weekend.
I was able to fire it up tonight without the PC and with just the X axis powered up and it worked as expected. I can't force the motor to move more than a couple of encoder counts.
I ended up with about 75vdc for the servos, close to the 80 volt max rating on the Gecko drives.
I had to build an e-stop and servo power switch box. I will be using it for other indicators and switches in the future. I reused the contactor that ran the servo power with the old hardware. The electrical box is a $7.95 A/C disconnect with everything inside removed. The Allen Bradley switches came from Ebay.
Here's a shot inside the control cabinet. Lots of extra space in there. You can see various power supplies on top. 5 and 75vdc, a stepdown transformer for 120vac to run the PC and fans. Below that shelf is the breakout board and the geckos. I have a smoothstepper paid for and should be shipping this week, so that will give me more I/O for the limit and home switches among other things. It's not nearly as messy as it looks here. All I have to tie down is the encoder and limit switch cables.
I need to stock up on useful things like a vise, hold downs, cutting tools, and some more measuring tools before it can make parts for me... but it runs! The first real parts I want to make are some Mog 406 wheel/brake adaptors to run chevy 8-lug stuff. And a new knocker for my back door!
Looking good man! Glad to see one of us is making progress!
Did you bother taking the time to draw a new wiring diagram, or are you just "winging it"? If you drew anything up, I would be interested in seeing your limit and e-stop circuits. I started drawing a diagram for mine in Acad, on the assumption that eventually something is going to go tits up, and I am never going to remember how I wired it, so it will hopefully help. In reality, the drawing is mostly done in it's first generation, however it takes up 35x17 inches of Acad virtual space, and is fairly difficult to track anything down in as well! I am going to take some time soon to seperate it into modules so it is printable and makes more sense!
Anyhow, looks great, gotta post some video when you get chips flying.
I have all the wiring diagrams scratched onto graph paper I did have a plan when I started, and my cabinets worked out really well so things came together pretty well.
The estop isn't 100% done. I want to wire it up to Mach knows when it's been hit and I need to wire it into the VFD, but right now I just use it to power the servo and encoder power supplies. The green momentary runs a latching relay for one side of the A/C and the estop is on the other side. It's simple and it's how the original was done, too.
I am on the other end then, I have everything in place so e-stop shuts down the VFD, All BOB IO, drivers, and signals Mach as well through one of the PP inputs. It sounds like you are already short on inputs though, due to the servo encoders, so my scheme wouldn't help you anyhow.
The thing I am having trouble wrapping my little brain around is how to set up the VFD to trigger an e-stop on a fault. Seems easy on the face of things, as most VFD's have a relay or two which can be programed to signal a fault condition, but my hang up, is mine signals a fault when it is not running. If I put the fault relay in the e-stop circuit, until mach starts the spindle, the drive will be in fault condition, and until the e-stop is satisfied, mach won't come out of e-stop. There are ways around it, but nothing I really like, it involves someway of bypassing the VFD relay to take it out of e-stop, and then rembering to switch it back in when the machine is up, or using a seperate input for it, of which I only have one left...
Have you thought about spindle shut down with the axis drives still running? I don't really see anyone even mentioning it remotely in their builds/retrofits, but it seems to me it should be incorperated. Especially on a machine the size of which we are working on. You could easily serriously damage something if the spindle shut down and the axis' were still running and doing heavy cutting at the time... I think probably a lot more than just a broken EM.
My goal with the e-stop is to make the spindle and the table stop moving. Since I am using servos, a runaway table is much "easier" than it would be for steppers. Killing my servo power supply takes care of that. As for stopping the spindle, I want the VFD to know when the estop has been hit so it stops. It has some dynamic braking built into it, but the spindle motor currently has a magnetic brake on it that I am trying to decipher. I think I might be able to run it on DC, or maybe I can find a replacement that isn't 3-phase. In any case running the estop to the VFD's estop signal should be fine for me. I don't plan on putting anything else on the estop. In this case, when the estop is hit the ways will stop instantly and the spindle will take a smidgen longer to spin down, so the part maybe be hosed but the endmill should survive.
As for I/O, I have one of the first Smoothstepper boards, so I have I/O galore The drivers I have take the encoder signals, they don't run straight to the PC.
Well I ran the first program on it today! I got the VFD wired up and working nicely with Mach as well. I made that roadrunner they send out with Mach about 50 times, doing various tuning. I got it running around at 180ipm with pretty good acceleration. I do have some kind of electrical noise on the Z axis limit switch once in a while so I need to figure that out. It might not be noise, it could just be the old contacts being weird. It only happened once. Here's what's left:
1) install the Smoothstepper
2) wire up a real err/res circuit so mach knows about gecko faults
3) mount the PC somehow. There is enough room on the cabinet but it will get really dirty sitting there.
4) buy tooling!!!!
I've been using the Smoothstepper for a while now. I've run the mill for 1/2 days at a time cutting roadrunners out of the air for various testing reasons. I'm very happy with the Smoothstepper.
I am not so happy with my Geckos, however. They seem to like to fault for no reason. They also seem to like to stay faulted for no reason, too. I have a brand new one that is now stuck in a fault state and I cannot get it to react to save my life. So now I have one that takes two reset pulses to start up, one that will not start up at all, and one that faults for no reason but will reset fine. The 4th one (yes I bought another one to mess with) is on the Z axis and likes to make X and Y fault when connected to the power supply. Disconnect it and the problem goes away. Some kind of electrical noise I'm sure.
I'm going to place an order for Granite Devices VSD-A drivers tomorrow. I considered these drives when I started out but I thought Gecko was the gold standard. The GD's have much more functionality and seem like more robust parts. I don't need to hassle with a seperate encoder power supply with them, they will use my differential encoders, they will run a higher voltage, they have PID tuning software that runs over USB (no scope needed for PID tuning), etc. They just look great. They are 2x the cost of the G320, however, and they are going to cost 10x as much to ship. And I have to rewire all the drivers. Dammit!
Hey, if you didn't throw out the old axis drive motors or other control stuff, I might be interested in them... I've got one that still runs on what looks like the stuff you took off.
Well this is odd. I just bought a Wells Index 810 that is missing electronics. I intended to replace the electronics and drives so i thought i would do a little research. I never would have thought another member would have already done exactly the same thing. I am going to start on this in 3 weeks. Any tips for me? What is this hobbey place where discounts can be had? Can i have a pictures of every square inch of what you have wired up. It would really be a huge help not to have to research and just install and wire and parts.
Well this is odd. I just bought a Wells Index 810 that is missing electronics. I intended to replace the electronics and drives so i thought i would do a little research. I never would have thought another member would have already done exactly the same thing. I am going to start on this in 3 weeks. Any tips for me? What is this hobbey place where discounts can be had? Can i have a pictures of every square inch of what you have wired up. It would really be a huge help not to have to research and just install and wire and parts.
Hah! I learned a lot, too much to probably put into this post, but:
- The 810 uses DC servos with optical encoders. The screws are something like 5 TPI iirc (I can doublecheck that). The spindle motor and servo power supply were both 3-phase, but everything else ran single phase 110. I am running a VFD for the spindle (3HP) that is controlled by Mach. I reused a step-down transformer to get 110vac and that runs my breakout board (Campbell designs combo board), Smoothstepper, way oil pump, and probably a few other things I'm forgetting. The servo power supply uses some huge toroid transformer I got on the interweb, about 76 vdc and lotsa amps. The servo supply and VFD are both using single phase 240vac.
- I started with Gecko 320 servo drives and they had a lot of problems faulting on startup. Once they were up they ran great, no issues there, but if I messed up and faulted one they would not reset. It got bad enough I couldn't even use the machine, so I swapped them out for Granite Devices VSD-A drivers. Night and day difference! Zero problems, software tuning, etc. The G320 faulting was a "common" issue and Gecko didn't know what was doing it. Well they finally figured out why, you need to run a 47k resistor between the negative motor output and incoming power ground. Needless to say the G320s are on the shelf. Granite Devices has built a new drive, the VSD-E and I would HIGHLY recommend taking a look. They also give hobbyist pricing. Knowing how nice these are now, I never would have bought the G320.
- The VSD-A drives use differential encoders, the G320s do not. The encoders on the 810 are actually differential but inside the housing only 1/2 the leads are connected, so they worked with the G320 only. I popped the covers open and sure enough the PCB inside had everything it needed for differential operation. I just ran wires from the screw terminals on the housing to the right locations on the PCB, just like the factory would have. the labels on the housing are even there for differential operation. I had to use the 'scope on them to be sure which was which, I can dig up my notes about the wiring if you need to make this change.
- I have been using a SmoothStepper since the first run was produced, and it's a huge improvement over the parallel port for Mach. Mach with the PP would just crawl. With the SS it is smooth and the PC doesn't seem to even be busy now.
Ahhh what else am I forgetting... Basically I cleaned out the whole electronics cabinet and replaced it all. Any questions?
I've got more pix, I'll try to find them. I can always take more
Well that depends what you're trying to do. The G320's are 80vdc/20a. The VSD-A is 150vdc/15a. The motors on my 810 are rated at 120vdc/24a (verified), but they were only being driven at 48vdc (again this is coming from memory). Amps=torque, volts=speed. My mill cranks along pretty well and certainly works fine for what I do with it. I am able to do 100 ipm rapids with conservative accelleration and haven't had problems cutting anything. But I also haven't tried to max it out yet (shop rebuild all winter).
The reason the OEM drivers were so large was because they were 1970s technology. They had a lot of discrete components and it was all through-hole stuff and that just takes board space. In this photo you can see the 3 driver boards down below with a fat power supply capacitor:
Andy
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Related Threads
?
?
?
?
?
Pirate 4x4
18.7M posts
366.4K members
Since 2000
A forum community dedicated to custom off-road vehicle owners and enthusiasts. Come join the discussion about trail reports, builds, performance, modifications, classifieds, troubleshooting, fabrication, drivetrain, and more!