[3.1.17]  BlueBot Revival: Awakeining an old Robot


is a Japanese Dainichi Kiko 'PT-300H'. Born sometime around 1983, Its a 'scara' style robot with three thetas. Its heavy, its somehow able to 0.1mm repeatably (3.9 thou) at full extension (28 inches). Thats quite good : ]  Each of the three thetas has a brush-dc servo and quite nice encoder. The robot itself was gifted to the cause by some excellent mit alums, dusted off from a basement and re-reverse engineered. Follow along for the trials and tirbulations of waking up a robot and putting some zip back into its axes. 

First Power-On
Electric Motor Adapter Endurance Test Conclusion Image Directory

What now?
High level overview of the controls system for BLUEBOT:
The main supply for the three main servo drives is a monsterous 40A 48V supply. An overload rregulator is used to prevent regen-current from quickly reversing motors. This overload regulator dumps any overvoltage conditions into waste heat. The 24V subsystem feeds control logic and main power for the Z-Axis stepper motor driver. The 12V supply is used to feed internal lighting inside the control cabinet along with the coolant fans to keep the supplies and servo drives cool. Finally the 5V supply feeds logic, the MESA 7I77 and externally feeds the (rather hungry) motor encoders. The servo drives (AMC 25A8 / 12A8) were surplus units, nominally from ebay / laboratory clean-outs. Each of the servo controllers are mounted on an aluminum plate for better thermals.
The actual control box is a bit of an array of parts. Its was difficult to determine 'how much space will everything take up' without actually having all the parts on hand.  A few things were learned during reverse engineering the robot, which resulted in some parts being larger than expected. The encoders each pulled 250mA at 5v, so a solid supply was needed, the electric spring loaded brakes on each axis, each pulled 1A at 24v, so an oversized 24v supply was needed. The regen on the controllers was sizeable enough to cause overvoltage on the main supply, so a shunt regulator was aquired [link] [local datasheet]

Pre-Disassembly: Firing up the Sequencer to see what works

Robot gets a new 'life' 
I lucked out and inherited a beautiful big robot. Unfortunatley I didnt grab a photo of it in the basement where it had been living. It took a bit of heaving and shimmying, but eventiually it ended up in the mighty rob's pickup. It then found its way into a wonderous basement. The robot itself broke down into two parts, its base, which nominally protected the theta 0 motor and provided a surface to work from, and like, the whole robot. Shown in the photo is a base that the robot sat on, made of, what i can only describe as, mega-steel.
Mega steel base meets a concrete floor
A concrete bit and a non-hammer drill made quick work of chewing into the floor, a shopvac was placed right next to the hole during the drilling. A 1/2" diameter hole was drilled first (to limit the loading on the final bit) and finished with a 7/8" diameter bit. This took about 10 minutes per hole, but, admittedly it was overloading the drill. I waited 10 minutes between hole to let the drill cool down. Shopvac and a respirator was used, along with safety glasses as appropriate PPE.
Aaaaand the drill cooked
After the third hole, the trigger controller for the dewalt failed 'full on hbridge short', any battery connected pumped ~40A into the handle and smoke would pour out quite effectively.
A quick drill-repair later...
Admittedly, the drill [Dewalt 9180] was used and had seen a proper beating. I grabbed a replacement drill trigger and a replacement motor. Some disassembly later and the drill was back up and running. Its a nice drill, and fortunatley with the advent of new 20v form factor drills, these have all become quite cheaper. For more details on the drill repair, see here [link].
The robot stands up:
I cant harp on this hard enough, if you want to get whatever your robot / contraption up and running, especially with reasonable acceleration and deceleration, a proper mating to the environment is a must. For this application, I used 'giant frigging concrete anchors'. Structral mating to the environment was an issue on a previous robot [link], and as ideally this would move quicker, there was plenty to do.  Specifcally I used 7/8" diameter concrete anchors [link] that attach to a 1/2-13 threaded stud [link]. Nuts and giant washers were used as leveling hardware. To drill into the concrete i used a combination of masonry drill bits ending up with a 7/8" diameter hole.  To 'lock' the expansion anchor in place, a threaded bolt is used and a nut is tensioned against the expansion anchor, this jams a moving slide into the anchor and helps keep everything pre-loaded. Note the expansion bolts are rated to 8100lbs shear and 5100lbs pullout when used correctly, Oversizing really wasnt harmful in this application.

Bill Of Materials [Including fabrication hardware / consumables]
# RequiredCost Used no shipping
Cost New
Shunt Regulator: SRST50G [Advanced Motion Controls]
The shunt regulator is a rather simple 'If voltage is above X, turn on a clamping resistor, rather quickly' device. I started lt-spicing up a quick comparator model and before i got the chance to finish, i found a
Servo Drive [Theta 0,1]: 25A8 [Advanced Motion Controls] 
I was fortunate to have one of these drives on hand from a lab cleanout, but purchased the second for the second theta. I debated using an existing extra 12A8 Brushed servo drive for the second theta, but determined that 'tuning the servo drives in software is a pain' and decided to only do it once.
Servo Drive [Theta 2]: 12A8 [Advanced Motion Controls] 
I had an extra drive on hand
Stepper Drive [Z axis]
This was purchased 'new'
Stepper Motor [Z axis]
This stepper was pruchased new as well, i opted for a double-ended Nema 23 stepper so an encoder could be added in place for extra precision, and to close the control loop.
Servo Drive [Z axis rotary] 1
Concrete Expansion Bolts 4
Concrete drill bits 2
Base mounting hardware 1
60A contactor 1
Dell PC 1
MESA 7I77 + 6I25 1------347.52
DB25 Breakout 1-----9.59
Meanwell 24V 5A Supply1-----33.99
Din Rail Relay Breakout119.49
Logitec F710 Wireless hand controller (low cost pendant)125.9949.99

From the sponsors:

Hardware upgrades...

Progress: After tweaking with settings and optics here are the progression of images

(There's other photos in the photo gallery)
Concluding Remarks:

From the sponsors:

If you have questions or comments, ask below or send over an email.
HTML Comment Box is loading comments...

(be careful, im not responsible for whatever weird biotics were leftover in your surplus sequencer, wear safety gear!)

Rensselaer Polytechnic Institute 
Electrical & Electrical Power