Here's my approach to the drawbar problem. The spring applies 625 lbs working pressure at 5/8" compression. The stop bolt is adjusted to give full release at the quill's hard stop and will eventually be replaced with a motor driven stop that retracts except for tool changes. So far it works great manually. Takes pretty good pressure on the lever but not excessive.
Threads are 3/8" x 32 tpi on both ends of the OEM drawbar which is necked down to about 3/16 for most of its length. A 1/2" diameter extension was simple to make. McMasterCarr has taps. Used green loctite.
Threads are 3/4" x 32 tpi on the shaft. I cut those on the lathe as the tap costs about fifty bucks. No loctite needed.
Anybody doing servo controls for these things?
I just bought a pair of Parker ZX series controllers for the x and y axis to run the Z610 series servo motors and it looks like those will work fine, using a serial connection and Parker's Xware terminal software on the PC but that's just 2 axis and using their X code. We have made a number of straight moves. I still have the Galil 2160 and the breakout boxes and I suppose for the other axis I'll need those.
The other servos are Pacific Scientific (2) and I have a MPC (Motor Power Company of Italy) brushed 60V, 7.7 Amp gearmotor on the knee: Penta 5M 30 series. I plan to use the PS servos on the quill and rotary table. Can anyone suggest servo amps for those two? I'll worry about the finer points after I have it all up and operational I suppose.
I'm building a Hydrapoint right now. I'm using servos, but not a commercial off the shelf solution. I'm using 24v 2hp Battlebot type motors on 2.5:1 htd belt drives for the X and Y. I'm using a Kflop board to handle all the motion control so I dont need complete servo drivers (Kflop closes the control loop) just the power stages since the encoders go to the Kflop and it just sends a PWM and direction signal to my H-bridge boards.
For the knee I have replaced the acme screw with a 1.25" ground ballscrew screw from a Mazak I got on ebay for a steal. I'm using one of the battlebot motors there, but have it on a 6:1 reduction belt drive (120 tooth HTD pulley barely fits inside knee). The X and Y screws are from a bridgeport kneemill conversion and fit the hydrapoint great.
Sometime in the next month or so I'll start a thread here on the zone with pictures documenting the build. I'm still putting it together, but I just acquired a bunch more parts and am anxious to get it going. I looked at your pictures posted in 2010 quite a few times while thinking of how to build the drive mechanical parts.
Hey Marvin, thanks for posting and if you'll post a link here to your thread I'll follow it too.
I've learned a few things on the control side. Most importantly, the Parker Compumotor ZX servo drives are a poor choice for a coordinated axis machine as they are only capable of accepting ASCII commands in a proprietary command language over the serial comm link and you cannot tap into the servo amp sections without considerable reverse engineering sans the benefit of schematics. In short, damned near impossible. They are single axis controllers as well so even if you crack the code on the command language you still cannot effectively coordinate the axis. Earlier units ("X" I think) had amp inputs. ZXF had "Follower" capabilities of dubious benefit. So in due course these will go back on the market, to be replaced perhaps by AMC 40 series amps if I can get them at a price I can stomach.
I expect to go with AMC amps on all 5 axis (X,Y,Z, knee, and rotary table). As far as I can tell the B25A20AC is a good choice for the 3 smaller motors. The Galil 2160 controller and 2900/2908 combo I have on hand should complete the picture I hope, and then perhaps Mach4 or something similar on the PC for the CAM interface.
Eventually I expect to tie in the spindle VFD, some sort of control on the shifter, brake and Vari-drive on the Bridgeport head, and I'm installing proxes now for limit switches. The quick change is also on the list. I'm using it and the carousel now but swapping the tools manually until I reassemble the arm. (It's a very weird and strange animal)
It's a very beefy machine for a Bridgeport type and the fully supported table is a distinct advantage for rigidity. Eventually it will be a very good machine. Just not there yet. Lately we've been using the working drives though for an experimental process so nothing is likely to be changed on it for a couple more months while we wait for better equipment to replace it.
Just a quick update. I bought four AMC B25A20ACQ amps, which I know will run the smaller motors and should also run the Parker 610s fine if I don't crank them up too high. I've got the amps, Galil controller, PC, and PacSci motors on the bench and optical encoders on order. We've been running the mill on the Parker ZX controllers. They work fine if you don't demand much from them. Still have to get everything to play nice, and need to mount a motor on the quill. Then if it all comes together properly I'll look at setting up the rotary table on axis E.
I'm back.Been running the mill in semi-manual mode for a few years now and just recently found an affordable deal on the servo amp for the knee. So I pulled the Galil stuff out of the storage cabinet to have a look. I'm enrolled in a tech class this summer for computer managed machining and hoping that experience will give me a chance to refresh my memory and make a bit more progress. I had the Galil setup working on the bench at one point but I have some really significant backtracking to do. Just recently ran across this link, I thought I had list it.
The mill currently has a couple of single axis controllers on the x and y but I recently added a 4 axis DRO which has been most helpful. (and also demonstrates that my ballscrews have backlash... who knew?) The tool change works but is something of a pain, there is little reason to think it can't be much better.
But we'll see what the new generation of machine tool controls looks like. Maybe there will be something in there that I will want to use. I am still clinging to the idea of having a set of rotary knobs or handwheels that I can spin to simulate manual operation of the machine, preferably with computer tracking for repeatability. But I don't think the concept has caught on yet.
Reply with Quote
