this project i have started is way out of my skill level, but i intend to finish it, i have came to Chief Delphi to get some assistance.
summery of Project.
Create a 3D Printer.
Acquired an Universal Instruments Axial Inserter (old) with everything.
included (2)PMI Servodisc DC Motor 00-01202-090 Type U12M4 (see page 27)](http://www.potomacelectric.com/Downloads/Datasheets/Kollmorgen/Kollmorgen%20Servodisc%20catalog.PDF) and (2) 152-021-250-11UF encoders. Plus a half ton of other stuff. Including a “PDP-11”
The problem is that i don’t understand how to run the motors and how to program the PDP-11. my solve for these problems is using “Black Jaguar’s” using CAN attached to a desktop PC. also how can the encoder talk to the Jaguar, the encoder has 8 wire output. >> Is there a less expensive and more efficient way to do this? Is this way possible? Anyone live in or around Torrington CT that is willing to help me? Is anyone willing to IM/Video me to help me?
First off, that PDP-11 is a genuine antique and a nice piece of history; don’t wreck it.
Servo motors and Jaguars are not a good mix. You really want a servo motor driver, one that can manage the feedback from the encoders. I also don’t think it’s easy to run a CAN bus from a typical PC.
My best recommendation would be for you to google “servo motor” and learn how they work. Also, it is a lot harder to build something (like a 3D printer) from components you happen to have, as opposed to buying what is best for the task.
Lastly, get a copy of issue 21 of MAKE magazine.
Josh,
Applications of this type would normally turn to stepper motors for accuracy and position at slow speeds. Jaguars are not designed for stepper control. How did a PDP-11 get included in the KOP you received? These servo motors would depend on zero backlash in any transmission they would be connected to for accuracy in write head position.
(Thinks to self…was that ‘up, down, up, up, down, up, down, down, load’ or ‘up, down, up, up, down, up, down, up, load’. Well it must be right the paper tape is moving.)
the PDP-11 was the original computer that was running the matchine i salvaged. another problem is that the reason they were throwing it away was all the backboard chips (i think that is what they are called, but the cards that are in a seprate location that control everything) are burnt, fried, snapped, etc. if they were in working condition i would spend my time looking on how to programe a PDP-11. my guess is that if i bought new cards ($110 a pop, 19 of them) i could get it working the way it was. but that is WAY to much money for just a hunch. I took the pdp-11 out of the machine, because i know what vintage stuff is worth, but this i dont know, so i want to be safe than sorry.
The reason why i thought i had the right stuff for a 3d printer is because the machine moved a PC board (up to 28" square) around and placed components (resistors, cappasitors, IC’s, etc) in the threw holes. so it had to be accurate and repeatable because it did around 400 boards a day, and got calibrated 2 a year. that is just XY movement, for the Z-Axis i planed on using a liner actuator mounted right above the orriginal leed cutoffs (must be dead center because they didnt move) but i am just looking for ideas to move the motors in a pattern from a 3d image.
An interesting dilemma. The PDP 11 I am familiar with was used as an edit controller for a CMX video post production system. The main box was three or four rack units high, with a separate paper tape reader and several frames of interface electronics that talked to remote controllers for the tape machines (called I Squared). In addition to controlling several edit machines it also controlled a video switcher and audio mixer. As I remember it had 16K real core memory.
The way DEC built those things, it is likely the main computer still functions. It is the interface components that likely went bad. If the motors came with the system, then it likely had smart interfaces. The computer would tell it to go somewhere and the interface would respond when the motors actually got there.
There are many motor controllers that are available for PCs that can be programmed in modern software.
one idea of mine (colaberating with Trekkie (dont know real name)) is to use 2 jaguars in parallel to run one motor. and use RS232 to CAN for controling via computer.
Josh,
Jaguars are not to be run in parallel. In all likelyhood, that will destroy one or both Jaguars.
is there another way, i am looking for the simplisity of the jaguar. Built in PID for position control, limit switch’s input, encoder input, preferaby RS232 able. Scott McMahon a TI employee that knows Stellaris projects replied to me with this
It is probably outside of the scope of your project, but the Jaguar’s goal in life was two fold. One, to make FIRST projects easier by providing a platform that could do some of the control tasks internally (call the built in high-speed PID functions rather than writing them). Two, to serve as an educational tool for creating a similar device. If you have the ability (time, etc) to spin your own, Jaguar’s schematics and software could be leveraged. Replace the HBrige and modify hbridge.c to match. I realize that is a tall order, but something to consider if time and resources permit.
does anyone have the knowlage to work this that is willing to teach someone wanting to learn. or can someone point me to reasorces to learn how to do this.
Josh,
The Jaguars have some good current capability. Did you need something that draws currents above the Jaguar spec?
i need 8 amps with a peak of 65 and 45vs dc (per moter)
OK,
Neither the jaguar or the victor can handle the 45 volts. You can buy special victors that are designed for 24 volts and I think the spec on the new Jaguars is also higher than 12 volts. You might have to do some serious research to find motor controllers for that output. 45 volts is kind of odd, usually it will be 12, 24, or 48 volts. Higher voltages are also encountered.
i just read on another forum post that victors handle 48 volts the problem being is that they are $200 each. and when i was talking with a TI (the new owners of lunicy micro) he was talking like it possible to imporve the h-brige and other components to handle the the 48 volts, if anyone knows where to learn, or is willing to teach how to do this large task, i will be willing to listen and learn. Or if you ar willing to do this for me i will be willing to private email ideas.
Josh,
The 48 volt version of the Victor is part number 48 HV, not an 884 as is used on First robots. You are correct that they are $200.
To get this straight, you wish to use the motors linked above that are pullouts from an antique machine. If the curves you linked to are to be believed, this motor can be run at 12 volts with no cooling and a much slower output (less than 1000 RPM) and lower torque (~120 in/oz). You can get higher output torque but you will need to cool the motors during operation.
Have you tested the motors yet at any voltage to see if they actually work? A few of the critical problems with motors of this type is the brush wear, commutator contamination, and magnet structure reduction in field strength. If this was a working machine with these motors when the boards burned up, it may have been caused by motor failure for one of the above problems. I hate to be the bearer of bad news on this.
UPDATE: all motors turn. Used an old power supply to give them 5vlts @ ?amps. They turned and I could not hold them back with my hands. It was very slow, but it was constant no matter what torque was put on them. I am still working on how to read a 7 pin encoder. Also working on how to work a “Z” axis into the design.
I am thinking of just using this power supply because it is rack mountable. And it is constant.
If anyone is interested feel free to ask questions.