Really interesting story.. I love the way you described the things...

Now is the time to make all the electrical connections. We used a junction box on the machine into which all the cables from motors and switches are connected. From here two 4 meter cables join it with the controller box through metallic flexible pipe. Power for all the three motors is supplied through a single 12 core shielded cable. All the three encoder signals and the proximity home switches go through a single 20 core shielded cable in a separate flexible pipe. Connected the motors one by one to the break out board and everything seems OK. Now connect all of them together and the motors start creeping motion by themselves. A quick check in Mach3 pin assignments. Switched to "Step low active" on all the three motor outputs. Creep slows down, but does not stop. Connect one motor at a time, the creep stops. Connect them all together, it starts again. Hmmm, seems like a design fault. System is catching electric noise from somewhere. Finally Mr. Nooruddin figured out that combining all the encoder cables together is causing it. Khalid brought out a nice coil of 4 core shielded cable and each encoder is re-wired separately. Motors are stable now. Calibration of all the axes and the home switches took another 15 minutes. Whole day is consumed in this exercise. 2 days still remain in the deadline. We decided to do the test machining the next day and try the vertical mount with actual simulation of our work environment on the last day.

The second last day of the project is a Friday. We have less working hours on Fridays because of the Muslim prayers break on Friday afternoon. All the calibration and configuration of the machine was done the previous day. I do some jogging tests and adjust the max. feed and acceleration on all axes. Next thing is to test a dry run. Loaded the roadrunner and hit the cycle start. Machine finished the dry run without a hitch. Now test the system noise with the encoder index pulse signals. Uh Oh. Something's wrong again. Axes shifted almost 1mm during the program run. Ran the test again. Same results. Panic time! Checked the wiring closely. Found a loose ground connection. Fixed it. Ran the test again. This time it's only X-axis which is loosing steps. Mr. Nooruddin spent some time with the breakout board circuit and components. There was a spare channel for A axis on the BOB. Shifted the X drive I/O's to that channel and that fixed it. The machine is all ready to go and produce some chips. But we have reached the end of day. We'll have to do all the trial runs on the last day.

It's the 15th day of our project. Backlash on all the 3 axes is checked. It is less than 0.01 mm. Not bad. We run a quick setup for hooking up the hydraulic spindle with our controller. The 4 core shielded cable from Khalid comes in handy once again. After checking all the stuff for the last time we shift the machine to the machine shop. A mounting plate for the machine is ready there. We decide to do the first test run on the mounting plate itself. The machine is bolted onto the mounting plate and a 3/4" 4-flute endmill tool of HSCo is mounted in the spindle. Checked the spindle rotation direction. It is rotating in the wrong direction. Switch over the supply pipes and now the direction is OK. A 50x50x20 mm pocket machining G-code is generated on Mastercam and loaded in the controller. Centered the plate, touched it with tool and all the three axes made zero. Depth of cut is set at 1 mm and we start the cut. Machining feed rate set to 200 mm per minute, spindle rpm is 300 (won't go above that). It makes a couple of cuts. So far so good. (see the video in the first post of this thread). After 3 or 4 cuts we notice an undercut on the corners. We check the structure deflection on all the slides. It's not noticeable with a dial indicator. So it must be tool deflection due to extra long adapter and conventional milling cuts we programmed. All my doubts about the machine structure not being strong enough are washed away. Must congratulate Khalid for carrying out a job perfectly.

Checked the load on the servo drives during cutting. It goes upto a max 30% on all the drives. Our guesses for the motor size have also paid off. Now the only job remaining is to mount the machine vertically (the way it was designed to be used) and do the machining which it is supposed to do. We'll do it after lunch. This was the meal I enjoyed the most on that trip .

Machine is disassembled once again after lunch and moved to the stripper shop. There they have assembled a 130 mm thick block on top of a work bench. Our machine's mounting plate is welded on to it and then machine is bolted to that. Another program is made in Mastercam for machining a tapered wall pocket of 160x80x90 mm. I am interested in measuring the motor loads in the vertical mounted position. In this position, the Y-axis is dragged down by gravity and any movement against it would yield the most load. 16 mm 4-flute HSCo tool is installed in the spindle. Job centered and zeroed with respect to tool. This time we have programmed climb milling cuts in Mastercam. This way it won't get dragged into the work-piece at corners. Speed and feeds are set as they were in our previous setup and start the cycle. Mastercam gives us the estimated machining time of 9 hours. With the manually operated machines, they had completed this job in 50 hours in the test run. It is close to sunset and we won't be able to complete all the 9 hours machining today. We decide to do half an hour of machining and then complete it after the weekend. The machine was run for almost an hour and 6 mm deep pocket was cut during that time. Machined surface finish looks good, sizes are OK too. Mission accomplished.

We will leave town in the night and return back home. All the credit goes to Khalid for carrying out the project in the assigned time. All my doubts about meeting the deadline and the performance of the machine proved wrong. Another project executed perfectly by Khalid.

A video for the final test of the machine is here.

Specs of the machine:

X-axis travel - 200 mm
Y-axis travel - 192 mm
Z-axis travel - 95 mm

X-axis max rapid - 5000 mm/min
Y-axis max rapid - 5000 mm/min
Z-axis max rapid - 4500 mm/min

Resolution on all the axes - 0.005 mm

Timing pulley reduction on
X-axis - 10:30
Y-axis - 10:30
Z-axis - 11:30

Ball screw pitches
X-axis - 5 mm
Y-axis - 5 mm
Z-axis - 4 mm

Servo motors/drives on all three axes
400 W, 3000 rpm, Delta - B2 series

Motion controller - UC100 from cncdrive.com

Breakout board and logic design - KaraCNC (www.karacnc.com)

CNC software - Mach3 V3.43.066

Controller PC - Dell Inspiron Laptop, 2.4 GHz, 4MB ram with Windows 7 Ultimate

Spindle - Hydraulic spindle from Climax, 300 rpm max

Thats a very nice info Zafar... Here is the rough cut at full depth by the machine. 10 degree for 70mm and the remaining is 37.5 degree upto next 20mm. Total depth is 90mm. No blunt of tool, no changing of tool.. No coolant... Almost took 10Hrs...

Khalid,

Good to see the final result. A correction from Khalid. With manual machining, it took 70 hours for rough metal removal and another 20 hours to hand grind the pocket walls to the desired finish. So a total of 90 hours. Compare this with 10 hours and a better finish achieved.  Almost 90% time saving. What are you going to do Khalid with this extra time .

I am thankful to Zafar and Team for their untiring efforts with me in this project. We have very good experience with you and shall keep you engage on other projects. I am really gratful for taking the time from your busiest schedule and accomplished this task professionally. I wish i could have expertise like you.