Day Forty-Nine

So here are many of the pictures that I kept meaning to upload the past few months. Sorry!

Above: The claw of our mechanical arm.
Lower Left: Backside of Arm. Lower Right: Front-side of Arm.

Lower Left: Encoder with Servo.              Lower Right: Close-up on Encoder Case.

  Below: Arm and Microprocessor Board while I was 

working on my Quadrature Encoder Program.

 Above: The driver board that had originally come with the Arm.

Below: Side View of Arm.

 Above: Better view of claw with the pneumatic servo.

Below Left & Right: The encoder without its protective cover.

 Below: Better view of my setup for the encoder program. (4 images)

Below: Images showing the MMBe on the microcontroller board with current wiring. (3 images)

Day Forty-Eight

After 2 hours of testings we, Dave and I, have come to the conclusion that our MMBE is junk. It cannot tell what the status is anymore, neither forward nor backwards. I have a bunch of pictures to upload for Day 49. I'll Also get a copy of our newest program for that day.

Day Forty-Seven

As I promised, CODE!


Interface Test #1

CON
  _xinfreq = 5000000
  _clkmode = xtal1 + pll16x


  counthi = $27
  countlow = $10
  speedy = $80


OBJ
  debug:"FullDuplexSerialPlus"
  bob:"MMBe"
                                          
PUB Main
 Comm


 repeat
   Display
                                                                                             
PUB Comm
  debug.start(31,30,0,57600)
  bob.start(3,4,2400)
  waitcnt(clkfreq+cnt)
  debug.tx(16)


PUB Display | key, forward, reverse, power, xf, xr, status, a, b
  xr:=0
  xf:=0


  repeat until xf == 1 or xr ==1
    debug.str(string(13,"press F or R for direction"))
    key:=debug.rx
    debug.tx(13)
    debug.tx(key)
    debug.dec(key)                              
    debug.tx(13)


    status := bob.getstatus                                     
    debug.dec(status)  


    if key == 102  
      xf:=1
      debug.str(string(13,"Motor is moving forward"))
      bob.setdir(0)
      bob.set(200)
  
    if  key == 114
      xr:=1
      debug.str(string(13,"Motor is moving in reverse"))
      bob.setdir(1)
      bob.set(200)


    if key == 115
      bob.motorstop
                                
 if xf == 1 or  xr == 1                        
   debug.str(string(13, "yay it worked"))
   
Interface Test #2

CON
  _xinfreq = 5000000
  _clkmode = xtal1 + pll16x
   
OBJ
  Debug : "FullDuplexSerialPlus"
    MMB : "MMBe"


VAR
  byte counthi, countlow, Speed, dir, trip
  word MoveCount, TotalCount                                


PUB Main
 Comm


 waitcnt(clkfreq+cnt)
    Totalcount := mmb.readHiCount
      Debug.dec(Totalcount)
      Debug.tx(13)


 repeat 
   MoveCount := debug.getdec
   Speed := debug.getdec
   dir := debug.getdec
   mmb.setdir(dir) 

   outb[15..0]:= MoveCount
   counthi := outb[15..8]
   countlow := outb[7..0]
   Debug.tx(13) 
   Debug.dec (mmb.encode(counthi,countlow,speed))


   repeat 
     Totalcount := mmb.readHiCount
     Debug.dec(Totalcount)
     Debug.tx(13)


     {if totalcount > 10000 and trip == 0
        trip :=1
        mmb.encode(counthi,countlow,60)} 
  
PUB Comm
  debug.start(31,30,0,57600)
  mmb.start(3,4,2400)
  waitcnt(clkfreq+cnt)
  debug.tx(16)


PUB Display | key
  mmb.set(128)


  repeat 
    key:=debug.getdec
   
Speed Control

CON
  _xinfreq = 5000000
  _clkmode = xtal1 + pll16x
   
OBJ
  Debug : "FullDuplexSerialPlus"
    MMB : "MMBe"


VAR
  byte counthi, countlow, Speed, dir, trip
  word MoveCount, TotalCount

PUB Main | x
  Comm
  mmb.speedcon_mode
  x := 0
    movecount := debug.getdec
    outb[15..0]:= MoveCount
    counthi := outb[15..8]
    countlow := outb[7..0]


  repeat 
    x++
    mmb.speedcon(counthi,countlow)
    waitcnt(clkfreq*2 +cnt)
    mmb.motorrev
  
PUB Comm


  debug.start(31,30,0,57600)
  mmb.start(1,2,2400)
  waitcnt(clkfreq+cnt)
  debug.tx(16)

Day Forty-Six

We pulled up my good old encoder program that I shared a few weeks back. Started modifying it so that it would work with the MMBE program for motor control. (Which I WILL post tomorrow, I promise!) We, Dave and I, cut out much of the originals of both programs in the effort to give signal to a specific point. The MMBE program gives the signal; the encoder program tells where the arm is.

Day Forty-Five

Worked the kinks outta the program except one, it only works in the forward direction but not reverse. Dave H and I started working on maintaining speed while under load. It didn't work. Our teach said that we may have to go back to my original encoder based program to actually make good progress. The MMBe has aspects of itself that limit what we NEED to do.

Day Forty-Four

Today we, Dave H. and I, re-wired the microcontroller for neatness and safety for the circuit. In this way, we do not have lots of looping wires going hither and thither. Then we retested the circuit to verify that everything worked as it did on Day 43. 'Twas a success. Tomorrow, Day 45, we will be working on cleaning out the bugs in the Forward/Reverse program that Dave H. wrote the beginning of last week.

Larry J. and John are working on getting the pneumatic hand/claw working on our arm.

Day Forty-Three

So we didn't actually work on that program today. Larry, John and I worked on getting the pneumatic intake from one of the non-operational arms onto our arm. It was screwed on TIGHT. Afterwards, we looked for a quick connect system for that intake. The intake is to operate the pneumatic solenoid for the claw/hand of the arm we are working on. The searching and removing and fitting took about 45 minutes. The remaining 5 minutes of class and the next 105 minutes after that was spent working with the MMBE chip on another microcontroller board. I was working with Dave and the teacher, Michael, to get the counter portion of the program reading at the right time. Mostly my part was watching and learning, as this part of the project I had not been working on. I have a decent idea as to what they are wanting to happen, but I'm hoping I'll get a clearer answer on Day 44. At that time, I'll write it down here, along with a cleaned up code of what the Michael and Dave have achieved so far.

Day Forty-Two

I worked with Larry J. on his interface program, we were hung up so long on how to get the text output to not flicker or scroll. Eric F. had to give us the clue we needed to fix that. Also the previous problems we were having on Day 41 was due to a damage crystals on the board we were using. Hopefully we make good progress on Day 43 for me to be able to post something regarding that!

Day Forty-One

Today Larry and I started writing a program that would allow for push button control of each servo. One button for the servo number and then in conjunction with that, specifically while the aforesaid button is pressed down, either the forward or reverse button. At the moment, we are just trying to get it to display strings when both buttons are pushed together. When we have a whole, albeit basic, program written, I'll post that.

Day Forty

Not much done. I assisted Dave H. with his program. Specifically with syntax and program flow. As it is not my program, I do not have a copy of it to post. The purpose of the program was kinda lost on me also. He just asked me what he needed to write to perform this or that action, and I responded.

Day Thirty-Nine

Connected the new wires I needed to make. Plugged in the additional encoders. I just had to add outputs from the newly added encoders into my program. It worked perfectly. (Though it did take about 40 minutes to get the wiring for the breaks to be disabled/powered at the same time, ugh) Now I am able to read direction and amount of movement from all 5 encoders at one time, as well as, being able to disengage all 4 breaks at once so the arm can do whatever it is we want it to do! As my end of the week project is now completed, I shall help Dave and Larry with their programming of the MMBEs.

Day Thirty-Eight

I need new wires to connect the other encoders, via the umbilical terminal to my microcontroller. The existing wires are all breaking down. So I made new wires today. Not terribly productive towards the end project but will help me get to this week's end goal of being able to read from all 5 encoders at once as well as being able to release all 4 breaks at once. 

Making new wires, hand crimping on new male and female ends... VERY time consuming. 

Day Thirty-Seven

Had a meeting as to what to do this week. I am to continue on my Encoder project and assist others in interfacing with what has been labeled, for better or worse, as my encoders. I'll start by wiring the brakes into one line, then connect the umbilical to the parallax microcontroller. THEN, I'll start inputting a repeat of my original program, which i posted a week or 2 ago, to run and track 2-5 encoders at once. I'll be adding 1 encoder at a time. Getting the 2nd encoder to work with the first in the existing program is going to be tough. Encoders 3-5 will be easy after I learn how to put in the 2nd encoder.

Day Thirty-Six

Today I was called off from my encoder work to help the other group integrate my encoder programming with their MMBE (Motor Mind B-enhanced) program. Purpose was to see if the MMBE could take the encoder input and give a processed output. No such luck. They were all working still on relearning to write in SPIN and making an interface program to work between the computer and the MMBE. So, I just spent the day reminding and advising Dave H. and Larry J. regarding their SPIN programs. I offered to Eric F. but he likes doing all his programming alone. Even though I had solutions to his frustrations. Oh well. It was a wasted day for me, but it helped the others progress in our project.

Day Thirty-Five

Now that I have my plan, I will be creating a wire packet (essentially 2 CAT5's with 1 extra strand) for a 17 wire tidy output to go into my board. I'm considering soldering my break lines together so that i need only 13 wires. Giving me 3 spare wires for later on. If I complete attaching the connector tips to the 16 strands of wire before time is out, then I'll start wiring it up and then modify the code. After the code is modified, tested, and works, then I'll post that.

Day Thirty-Four

Today I spent my time designing and planning a new wiring scheme for a multi-encoder program. Which I'll fiddle with later. It's basically the same as the last program posted just with more variables and more wires. Fun!

Day Thirty-Three

Today we had our weekly meeting. We decided what needed doing: interfacing the motor, learning more about the Motor Mind controller cards that Prof. bought for us.

Day Thirty-Two

Studying for PLC II exam, didn't work on the Project. Exam is today.

Edit: I aced my exam, woohoo!

Day Thirty-One

Studying for PLC II exam, didn't work on the Project.

Day Thirty

Researched on how to do an efficient wiring setup for the arms rather that the direct encoder input/output wires. Too messy. At the end of the day, I'm still lost as to which wires I'll actually need to hook up. Will continue when I have time, maybe Day Thirty-Three.

Day Twenty-Nine

Today I worked on wiring and programming in multiple encoders. I spend most of the first hour wiring it in. Silly me, didn't check to see if the connectors were attached fully to the input wires. Waste of 45 minutes. The next 20 minutes were spent determining what data from the encoder was required for the program. Tests have shown that the program needs all 3 channels from each encoder to be fully operational.

The wires just for 2 encoders was already at 10. On Day 30 (which will hopefully be Wednesday [a non-class day]), I'll relearn the connector plugs to cut down on the required wires for operation. Then on Day 31, I should be all set to set up and execute 2+ encoders with the program with a much more tidy wiring setup.

Day Twenty-Eight

Went over the specifications and schematics of the Motor Mind B - enchanced board with Eric Ard to determine programming requirements for when we receive the boards. (Teacher purchased a board for each group.) As I was doing that all of class period, I have not yet worked on getting the 5 encoders to work simultaneously with my board/program. It's put off for Day 29.

Day Twenty-Seven

Now that I have a working program for the direction and count, I will be furthering the code/circuitry by adding more encoders. Woot! This will require my figuring out if both Channel-A and Channel-B are require for the initial program, then adding channels from 1 additional encoder. After that is successful, I shall keep adding encoders (and wiring, egads!) until all 5 encoders are wired in and counting (with direction) perfectly.

Hmm I said I would upload pictures and haven't yet... Maybe monday, as I will need to upload the final wiring and circuitry for this program.

Note: Technically tomorrow is Day Twenty-Seven, but class is cancelled for Day Twenty-Seven and Twenty-Eight. Thus, I shall work today as if were Day Twenty-Seven! ^_^


Edit: didn't get it all figured out, will continue on Monday.

Day Twenty-Six

Tinkered with that Quadrature Encoder program from my previous post. I'm fixing the bugs in it. Hopefully it'll work completely or at least to some degree before 3pm today.

Edit: it's now 2pm and I've gotten the program to work. It shows current position and the change in position. (YAY!! we have direction now!) The change in  position is either positive or negative number with the amount of change from previous to current position. The output is FAST. So fast, in fact, that I need to slow down the output on the screen, just to be able to read it. O_O  Awesome!

Larry helped me with getting the program to output to the Terminal Program.

Day Twenty-Five

Had meeting today on what needed to get done this week. I'm going to continue working on incorporating the Z-channel to show direction of movement, with A- and B-channels doing the counting. I think.

Edit: Scratch the Z-channel; it wont tell direction. Moved onto a tangent. Found a Quadrature_Encoder program that can control 16 encoders at once. I'm using its example code to fashion a usable program from it.


{{
* File: ReadEncWithQuad.Spin
* Name: Palak Shah
* Date: February 20, 2012
*
* Desc: Attempt to read the values recorded by the microcontroller 
*       as inputted by the encoder. 
}}
CON
  _xinfreq = 5_000_000
  _clkmode = xtal1 + pll16x


OBJ
  Encoder : "Quadrature Encoder"
  FDSP : "FullDuplexSerialPlus"


VAR
  long Pos[6]  'Create buffer for 5 encoders (plus room for delta 
               'position support of 1st encoder)
  long AbsPos, Delta1


PUB Init
  Encoder.Start(20, 5, 1, @Pos) 
               'Start continuous two-encoder reader (encoders 
               'connected to pins 20-26)
  
PUB Main
  Comm


  repeat
    AbsPos := Pos[0]                 'Read each encoder's absolute position
    Delta1 := Encoder.ReadDelta(0)   'Read 1st encoder's delta position 
                                     ' (value since last read)
    Output


PUB Comm
  FDSP.Start (31,30,0,57600)
  waitcnt(clkfreq * 3 + cnt)
  FDSP.tx(16)
  FDSP.str(String("'Encoder Reader' by Palak Shah"))
  FDSP.tx (13)
  FDSP.tx (13)


PUB Output
  FDSP.bin(AbsPos, 16)
  FDSP.tx (13)
  FDSP.bin(Delta1, 16)
  FDSP.tx (13)  

Day Twenty-Four

Worked on getting the Z-channel programmed in, not sure how it's supposed to be incorporated, thus failed at this today. Will work on that on Monday (Day Twenty-Five).

Day Twenty-Three

Today I tinkered with the code that does a better increment without using assembly. Tinkered cuz it STILL doesn't work properly. But in this case, while it can keep up with the speed of the arm, both directions doesn't count in their own direction. Like Channel A is ++ and Channel B is --, but it only counts ++ or -- depending on the code setup.

The current code for counting is as follows:
CON
  _clkmode = xtal1 + pll16x
  _xinfreq = 5000000
  Ao = 20
  Bo = 21

VAR
  long counter

PUB Main | newBits, oldBits, direction
  dira[Bo..Ao]~
  dira[1..16]~~
  oldBits := ina[Bo..Ao]
  counter := 0
  repeat
    outa[1..16] := counter/1000
    newBits := ina[Bo..Ao]
    if newBits <> oldBits
      direction := (newBits & %1) * (oldBits >> 1)
      case direction
        0 : counter++ 'one direction works perfectly but the
        1 : counter-- 'other doesn't when the first does work.
      oldBits := newBits

Day Twenty-Two

Learned that using Count++ and Count-- as counters were too slow for the movement of the robotic arm. So I am now learning about CTRA and CTRB as much faster counters to count the pulse edges of the outputted encoder pulses. Prior to this, the code is 5 lines long. And too slow. Gotta work on efficiency of the code. If that fails, then I will hafta self teach (maybe some professor help) myself Assembly for SPIN.

Day Twenty-One

Programming in 2 channels, so that the microcontroller can see forward and backwards. Then with that knowledge we can incorporate Dave/Eric's work with PWM into that. Thus the computer program can control voltages to move the arm servos forwards/backwards.

Day Twenty

Today I got my program to count each pulse, had to use all 16 LEDs on the microcontroller. Now I shall work on making it count forwards and backwards. Taking input from both channel A and B. In this way we can see our exact position regardless of which way we're moving. The circuit is the same for each channel. Programming SHOULD be relatively easy.

Day Nineteen

O_O it blew up... omg.

just kidding.

Today i did get the circuit working, eventually. (took about 1.75 hours to get all the kinks outta the program, wiring and solving my grounding issues.) The LED on the microcontroller now flashes with every pulse. Tomorrow I'll do a test run to see if the 2nd program I wrote correctly counts each pulse. If it does, then I'll have what i set out to do this week done. Yay!!

Day Eighteen

Today I set up the circuit on the microcontroller board to test the encoders. First circuit test was to see if the directed output from the encoder caused the LED light on the microcontroller to turn on. I didn't complete the circuit in time to test this. For creating the circuit, I had no idea what i was doing and got a little help from Dave. On Day Nineteen, I will actually get to see if the circuit works or not.

Note to self: upload pictures.

Day Seventeen

Dave isn't here today, so I cant assist him. So I will work on getting the microcontroller board to light up an LED when it receives a pulse signal. Then, later to be able to count the pulses from each channel (A and B).

http://www.arrickrobotics.com/arobot/edled.html

http://ikalogic.com/tut_closed_loop_spd_ctrl.php

http://www.societyofrobots.com/robotforum/index.php?topic=13192.0

http://www.robotstorehk.com/motordrivers/motordrivers.html

Day Sixteen

I attempted to be useful today, but due to my severe cold, I couldn't concentrate on anything for very long. Low attention span? I will assist Dave and Eric F. on Monday with the program-to-arm-manipulation.

Day Fifteen

I spent the day programming/modifying code. Also, I investigated hardware that could help make the programming simpler.

http://www.parallax.com/tabid/768/ProductID/64/Default.aspx
http://www.electronickits.com/kit/complete/motor/k166.htm

Day Fourteen

I spent all of today's time writing and searching for SPIN code that would allow me to take an encoder's signal and output it to the LED lights on the Parallax Microcontroller. In the end, I have 3 programs that seem like they will work for the first stage of encoder-to-microcontroller programming. After class, I helped Larry make our wires for connecting the arm to the microcontroller. I stripped and flexed the CAT5 wires so they were straight. Then Larry soldered the male and female ends to those wires.

http://www.encodergeek.com/PropellerProgs.html is the site i found the 3 SPIN programs. Just search for "PWM signal".

http://www.nerdkits.com/videos/motors_and_microcontrollers_101/
http://www.electronics-tutorials.ws/io/io_7.html

Unrelated:
http://www.bobblick.com/techref/projects/propclock/propclock.html

Day Thirteen

Today was a break from work. The class spent the class period discussing what REALLY needed to be done moving forward. My group's arm is now in fully functioning condition. I decided that I, along with Dave, would be working more on the coding aspect of this Project starting on Tuesday. Got shot down a lot on my thoughts/ideas. So I'll just volunteer less input from now on. Other than that, I think I'll progress nicely.

Day Twelve

Not much done today as we are waiting for the new tip/connectors to come in. I did look at the programming for the arm. Lotta code. Makes some sense. @_@

Day Eleven

Today the teacher helped Dave and I finalized how many pins we have to have to run what we want to run on our arm [20 pins = (1 motor power, A and B channels) * 5 + brakes * 4 + brake common]. Then our prof. helped us with a pseudo code for the running of a single servo - foward and backwards, and what we needed to find out to get that pseudo code to work. Our group needs to find out which direction is the forward (Channel A before B) and which is the backward (Channel B before A).

After class with Larry, we will hook up the oscilloscope to the arm to determine the directions, the degrees in a single cycle, and the number of cycles in a full range motion.

Elbow - 35 cycles from home to full
Shoulder -
Wrist Pitch -
Wrist Yaw/Roll -
Waist -

John spent the whole time with the other group, helping with the testing of the encoders of the arm.

Larry will be buying the terminal wire male and female ends, of which our group will split the cost. We will solder the tips and ends to the wire rather than crimping it. $12 + S&H (if any), thus approx. $4 each of us.

Day Ten

John brought an air hose with him today, with some tools. Dave, John and I are going to test the pneumatics on our arm. Someone took all our wire connectors from our locker. As such we spend the 2nd half of the class searching the 'net for new wires.

Day Nine

John and I tested all the encoders on our arm. First we learned how to wire it for testing, then we learned what to look for in the test. This took all of the class period as Dave was absent. Afterwards, I put up all the equipment and the arm.

Day Eight

I don't remember exactly, at least at this moment in time.

Day Seven

Our teacher, Mr. Michael Beavers, showed the class how to hook up the PWM driver circuit board. He used multiple alligator clips and 2 DC Power supplies. Each Power supply boosted it's respective polarity. One boosted the positive; the other boosted the negative. He showed us the output on an oscillascope. It was neat.

Day Six

I don't remember exactly, at least at this moment in time.

Day Five

Today our group finished taking apart the arm's panels and cleaned it out some more (probably one more day to complete the clean out). We will lube on Day 7, if needed. Dave, John, and I also tested each dc servo to see they worked in both directions and got the servo breaks to release with the connections directly off the servos themselves.


After class, Larry and I worked further on the wired connections. We figured out which servos corresponded to which motor numbers in manual by testing which pins corresponded to which servos breaks. Through a process of elimination, we now know that Motor 1 is the base servo, Motor 2 is the Shoulder servo, Motor 3 is the Elbow servo, Motor 4 is the Wrist Pitch servo, and Motor 5 is the Wrist Yaw servo. 


To get the brakes to release we need to connect a ground to the Motor pin connector and a positive to the Feedback pin connector as detailed below.


Shoulder
Motor connector = pin 7
Feedback connector = pin 11
Motor #2

Elbow
Motor connector = pin 9
Feedback connector = pin 18
Motor #3

Wrist Pitch
Motor connector = pin 10
Feedback connector = pin 11
Motor #4

Wrist Yaw
Motor connector = pin 12
Feedback connector = pin 18
Motor #5

Day Four

Continued to take apart and clean the arms as much as we were able. We will continue on Day 5. I will upload pictures on Day 5 or 6.

Day Three

Today we took apart 2 of the arms to see which servos controlled what aspects of the arms. Also we looked at what cleaning needed to be done as well as what needed repairs. We decided that we needed specific tools and cleaners for tomorrow. Also I took "before" pictures for when we reassemble the arms after disassembly and cleaning.

Below are links to possible sources of info regarding the arms.
http://www.labx.com/v2/spiderdealer2/vistasearchdetails.cfm?LVid=12555905#MoreDesc
http://www.robotsdotcom.com/used/A255.pdf
http://www.americanrobotsales.com/c500cv2.pdf
http://www.ece.mcmaster.ca/~sirouspour/lab.html
http://stores.aleoutlet.com/Items/201007-11678?&caSKU=201007-11678&caTitle=CRS%20Robotics%20ROBOTIC%20ARM%2C%20Model%20A255%2C%20%20P/N%20RSA-14-103
http://www.eng.auburn.edu/outreach/k-12/robo-camp/crs-a255.html

Day Two

Decisions:
Decided on 5 things that need to be done, specifically 4 groups:


Group 1

1. Location of work
2. Teams of 4 (will organize in February)

Group 2

1. Get detailed information, data sheets/specs, on the arms and remote
2. Inventory of what we have
3. Supply list of what we need

Group 3

1. Clean and lube
2. Test remote, arms, and pneumatics

Group 4

1. Build interface
2. Program manual and auto modes
3. Finishing touches

Timeline:
January
09- Inventory + Datasheets
16- Clean/Lube + Supply List
23- Test Remote, Arms, Pneumatics
30- Same as 23rd


February
06- Build Interface
13- Same as 6th
20- Same as 6th
27- Same as 6th


March
05- Make arms work in manual mode
12- Same as 5th
19- Program the automatic sequence + learn about Pulse Width Modulation
26- Same as 19th


April
09- Same as 19th
16- Same as 19th
23- Same as 19th + Start Finishing Touches
30- Finishing Touches


Edit: After class, Larry and I looked at the arms for the relevant information to get more information on them.


2 Arms by CRS Robotics

• Model #: A251 and A255 (Larry Jolley found the link to the manual)
• Inputs: 5 @ 35VDC 2A
• Weight: 17 kgs. (37 lbs)
• Serial #: RA2219 for the A251 and RA20018072 (Mfg in May 2000) for the A255

2 Arms By Hudson Robotics

• Model #: CRSPlus - SRS-M1A
• Serial #: RA1298 and RA1326

Day One - Starting Out...

Today we, our class, identified what our project would be and what we are going to do to get an A. We will be working on a servo, dc motor, driven arm. We have an arm and a controller but no way to interface the two together - yet. Our end goal is to make the arm pick up a cup with liquid, move it to another cup, and pour the liquid into that second cup.

For day two, we are looking to start up a timeline as to what our daily goals are gonna be.