83 (Buy video camera) is done.
Data display user interface 80
Test, improve data display 82
Video acquisition, synchronization 84 (does not depend on 81)
November 10, 2007
11/09/07 Lee Sam
Time in 9am
Time out 3pm
1) Managed to make the LPC3180 (Master) and the MAC7111 (Slave) to communicate via SPI on both directions - one direction at a time. No external hardware is used, but the Slave Select line for the MAC7111 is generated using a GPIO pin on the LPC3180 and toggled only once every couple of SPI data transfers, to syncronize the transfer process.
2) DMA code for the LPC3180 to send data to the MAC7111 is written and verified working. The code is tested for multiple size of DMA transfers, from 1 byte to 50bytes.
Time out 3pm
1) Managed to make the LPC3180 (Master) and the MAC7111 (Slave) to communicate via SPI on both directions - one direction at a time. No external hardware is used, but the Slave Select line for the MAC7111 is generated using a GPIO pin on the LPC3180 and toggled only once every couple of SPI data transfers, to syncronize the transfer process.
2) DMA code for the LPC3180 to send data to the MAC7111 is written and verified working. The code is tested for multiple size of DMA transfers, from 1 byte to 50bytes.
November 9, 2007
Yingyi Tan 11/9/07
Completed projects:
task 14
Dependencies:
12 depends on 11
14 depends on 12
13 depends on 14
Current projects:
Helping to install the satellite boxes (drilling holes in robot). I believe is under task 27.
task 14
Dependencies:
12 depends on 11
14 depends on 12
13 depends on 14
Current projects:
Helping to install the satellite boxes (drilling holes in robot). I believe is under task 27.
11/09/07 Ullmann Kevin
4:00 - 5:45 PM (1.75 hrs)
1. Worked with Stephane on video (~1.2 hrs)
2. Photoshoot (~.55 hrs)
1. Stephane showed me what he has been doing with the video part of the graphics program. After he gave me a basic understanding of how everything works, we discussed various ideas for improvements and possible solutions to problems, including tie-ins to the current program.
2. Due to the pretty shirt I was wearing I was chosen to be submitted to a long period of squating, during which I accidentally knocked over the robot. We should definitely give the robot a name so that for that last sentence I could have said, 'I accidentally knocked over (Jimmy)' Jimmy of course is just a placeholder name.
1. Worked with Stephane on video (~1.2 hrs)
2. Photoshoot (~.55 hrs)
1. Stephane showed me what he has been doing with the video part of the graphics program. After he gave me a basic understanding of how everything works, we discussed various ideas for improvements and possible solutions to problems, including tie-ins to the current program.
2. Due to the pretty shirt I was wearing I was chosen to be submitted to a long period of squating, during which I accidentally knocked over the robot. We should definitely give the robot a name so that for that last sentence I could have said, 'I accidentally knocked over (Jimmy)' Jimmy of course is just a placeholder name.
11/9/07 Gregory Falco
1:15-2:45PM
1. Article Found (30 min)
2. Fixed Blog (15 min)
3. Lab Meeting (45 min)
-----------------------------------
1. I found Three Uses for Springs in Legged Locomotion on a internet service called Sage. I created a trial subscription that will last until Nov. 30 2007 so if anyone needs an obscure Mechanics article, I can get it for you.
2. I did some blog maintanance- ie. corrected formatting and forgotten labels
3. Took notes on meeting- will arrange for lab info session where members and prospective members will get to hear about the history and future of the Biorobotics lab.
1. Article Found (30 min)
2. Fixed Blog (15 min)
3. Lab Meeting (45 min)
-----------------------------------
1. I found Three Uses for Springs in Legged Locomotion on a internet service called Sage. I created a trial subscription that will last until Nov. 30 2007 so if anyone needs an obscure Mechanics article, I can get it for you.
2. I did some blog maintanance- ie. corrected formatting and forgotten labels
3. Took notes on meeting- will arrange for lab info session where members and prospective members will get to hear about the history and future of the Biorobotics lab.
11/08/07 Sam Lee
Time in 2pm
Time out 5:30pm
1) Worked with Jason to better understand the SPI problems and to sort out remaining compatibility issues. We suspect that we might actually be able to get away without generating our own slave select - which might make things alot easier
2) for the receving and sending data in the slave (MAC7111) the slave select seems to only be need to be asserted the whole time - ie there is no need to toggle between frames. this is beneficial as we can use a GPIO pin to toggle in software after each block of DMA transfers.
Time out 5:30pm
1) Worked with Jason to better understand the SPI problems and to sort out remaining compatibility issues. We suspect that we might actually be able to get away without generating our own slave select - which might make things alot easier
2) for the receving and sending data in the slave (MAC7111) the slave select seems to only be need to be asserted the whole time - ie there is no need to toggle between frames. this is beneficial as we can use a GPIO pin to toggle in software after each block of DMA transfers.
11/08/07 John Buzzi
1:00pm-2:30pm
1)Finished machining the satellite processor boxes (1hr 15 min)
-------------------------------------------
1) The microprocessor mounting boards were nearly done by the end of my previous session however holes still needed to be drilled in the last two boxes. Using the same method of centering each half of the box individually and drilling holes in specific locations I was able to finish the boxes. Drilling the mounting holes in the robot itself is the next step in the process.
1)Finished machining the satellite processor boxes (1hr 15 min)
-------------------------------------------
1) The microprocessor mounting boards were nearly done by the end of my previous session however holes still needed to be drilled in the last two boxes. Using the same method of centering each half of the box individually and drilling holes in specific locations I was able to finish the boxes. Drilling the mounting holes in the robot itself is the next step in the process.
11/06/07 John Buzzi
*Note: I havn't had a chance to write my blogs after i worked this week so i am catching up.
2:00pm-5:00 pm
1) Further discussed the satellite processor mounting with jason (45min)
2) Machined the processor boxes (2 hrs)
-------------------------------------------------------
1) Jason and i decided that it was necessary to remove all material that wasn't structural from inside of the sattelite processor mounting boxes. Also it was decided where to drill the mounting holes that will attach these boxes to the robot. We also concluded that in order to attach the top and bottom of the boxes as well as hold the processors in place we would need to use a system of screws and standoffs. This also meant additional holes needed to be drilled for this purpose as well.
2) I used the mill to first go through each half of the boxes individually and remove all excess plastic that was not structurally necessary. The boxes also needed mounting holes to attach to the robot as well as holes so that one half of the box will be able to attach to its other half and not fall apart. This process involved centering the mill over the box using the edge finder and then drilling holes at the appropriate locations.
2:00pm-5:00 pm
1) Further discussed the satellite processor mounting with jason (45min)
2) Machined the processor boxes (2 hrs)
-------------------------------------------------------
1) Jason and i decided that it was necessary to remove all material that wasn't structural from inside of the sattelite processor mounting boxes. Also it was decided where to drill the mounting holes that will attach these boxes to the robot. We also concluded that in order to attach the top and bottom of the boxes as well as hold the processors in place we would need to use a system of screws and standoffs. This also meant additional holes needed to be drilled for this purpose as well.
2) I used the mill to first go through each half of the boxes individually and remove all excess plastic that was not structurally necessary. The boxes also needed mounting holes to attach to the robot as well as holes so that one half of the box will be able to attach to its other half and not fall apart. This process involved centering the mill over the box using the edge finder and then drilling holes at the appropriate locations.
11/05/07 John Buzzi
*Note: I havn't had a chance to write my blogs after i worked this week so i am catching up.
12:30-1:30
1)Discussed satellite processor mounting options and battery mounting with jason. (1hr)
-----------------------------------------------------
1) We went through each of the 3 sections of the robots body and determined how to mount the satellite processor boxes. The plan was similar to the one i developed on my own however we thought of a modular system for mounting the battery packs on the nose of the robot.
12:30-1:30
1)Discussed satellite processor mounting options and battery mounting with jason. (1hr)
-----------------------------------------------------
1) We went through each of the 3 sections of the robots body and determined how to mount the satellite processor boxes. The plan was similar to the one i developed on my own however we thought of a modular system for mounting the battery packs on the nose of the robot.
November 8, 2007
11/08/07 Ullmann Kevin
8:30 - 10:00PM (1.5 hrs)
1. Adding functionality to changing variables (.5 hrs)
2. Researching 'machine gun problem' (1 hrs)
1. I made it so that when the user wants to change the variables that the list of variables and their respective indices also appears. This took me longer than expected because the window would freeze; I was able to resolve this problem by making the the msgbox 'modal'.
2. I searched online and within Matlab help to try and fix the problem of too-rapid key presses not functioning properly. I was not able to find much useful information. I am currently looking into the function ginput to see if I can alter it so that the CPU does not wait for user input to continue running the function.
1. Adding functionality to changing variables (.5 hrs)
2. Researching 'machine gun problem' (1 hrs)
1. I made it so that when the user wants to change the variables that the list of variables and their respective indices also appears. This took me longer than expected because the window would freeze; I was able to resolve this problem by making the the msgbox 'modal'.
2. I searched online and within Matlab help to try and fix the problem of too-rapid key presses not functioning properly. I was not able to find much useful information. I am currently looking into the function ginput to see if I can alter it so that the CPU does not wait for user input to continue running the function.
11/09/07 Yingyi Tan
5.00pm - 6.15pm
1. Re-soldered ribbon cable to sensor board (0.75h)
2. Assembled robot back together (0.5h)
---------------------------
1. The previous ribbon cable was not cut to the right length. As a result, it was tilted to one side. Also some parts of the insulation was broken and a piece of electrical tape had been used to cover the exposure. The old cable was removed and a new one soldered on. Each wire was of a different length to allow the ribbon cable to lie flat against the board.
2. The coupling between the right hip box and the center box was put back on to the robot. Loctitde was used on all screws to ensure they stayed in place. All parts of the robot was pieced together.
1. Re-soldered ribbon cable to sensor board (0.75h)
2. Assembled robot back together (0.5h)
---------------------------
1. The previous ribbon cable was not cut to the right length. As a result, it was tilted to one side. Also some parts of the insulation was broken and a piece of electrical tape had been used to cover the exposure. The old cable was removed and a new one soldered on. Each wire was of a different length to allow the ribbon cable to lie flat against the board.
2. The coupling between the right hip box and the center box was put back on to the robot. Loctitde was used on all screws to ensure they stayed in place. All parts of the robot was pieced together.
11/08/07 Hajime Furukawa
4:30 - 5:30
1)Board Modification and Mounting overhead (1h)
---------------------------
0) Sensor boards are plugged into the plug-a-pods. I believe some of the functionality is being tested with the sensors currently.
1) Discussed and thought of what to do next, which is mount the boards to the boxes. To do so, we must take out the connector on the side of the plugapod and instead connect a ribbon cable that spot. This is so that the mounting holes on the sensor board can be used. Then connectors will connect to the end of the ribbon cables.
1)Board Modification and Mounting overhead (1h)
---------------------------
0) Sensor boards are plugged into the plug-a-pods. I believe some of the functionality is being tested with the sensors currently.
1) Discussed and thought of what to do next, which is mount the boards to the boxes. To do so, we must take out the connector on the side of the plugapod and instead connect a ribbon cable that spot. This is so that the mounting holes on the sensor board can be used. Then connectors will connect to the end of the ribbon cables.
11/8/07 Gregory Falco
1:45-2:45PM
1. Corrected Blog (20 min)
2. Corrected Accounts (40 min)
----------------------------------------------
Several lab blog entries lack labels therefore are not capable of being tracked. I assigned proper titles and labels to those that lacked them.
There have been several late updates to the account info on webfinancials. They don't seem to update it at the end of each month and there is significant lag so I included recent charges for our account summary.
1. Corrected Blog (20 min)
2. Corrected Accounts (40 min)
----------------------------------------------
Several lab blog entries lack labels therefore are not capable of being tracked. I assigned proper titles and labels to those that lacked them.
There have been several late updates to the account info on webfinancials. They don't seem to update it at the end of each month and there is significant lag so I included recent charges for our account summary.
11/07/07 Lee Sam
Time in 9am
Time out 3pm
1) Narrowed down the problem of data being received by the slave - the infomation is going into the receive FIFO, however, it is not triggering the interupts. this problem is mostly a coding error - working to solve this.
Time out 3pm
1) Narrowed down the problem of data being received by the slave - the infomation is going into the receive FIFO, however, it is not triggering the interupts. this problem is mostly a coding error - working to solve this.
11/8/07: Ruina comments
Dear lab: Nov 7, 2007
Here are comments to all of you on your google blog entries. All of you should
glance at all of this
Generally most of you need a few more words and phrases so that someone reading your blogs
can know what you are talking about. Especially if this is supposed to serve, possibly,
as long term documentation.
Most especially, your entries from Tuesday night before a pay period ends should
be especially clear. Those are what I read first.
%%%%%%
About Carlos Arango:
I played with the IMU also. Fun. I recommend playing with it to everyone.
I am not convinced that the excursions you see on the angular velocity plots
are not real signal. I think you might be seeing real features of the leg motions.
Some suggestions:
1) Plot pitch angle (not angular rate) to get a better sense of the actual motion.
2) Compare with output from the motion capture equipment (in the lab
directly under the machine shop, contact Madhu
(madhusudhan Venkadesan) about using it.
On 11/5/07 you wrote about CAN bus code. CAN bus on what device? Whose buffer read
function? Where is this work documented?
On 11/2/07 you refer to “the code”. What code? On what sensor? Where is that code
stored? Where are your changes documented?
%%%%%%%%
About John Buzzi:
The Gant chart is great. Some things we need to fix/notate:
a) one person can’t do two one day tasks in one day.
b) What are the blue check marks at left column, “done”?
c) How to make clear non-critical paths (tasks that are on-going
like making the mass distribution symmetric)
d) Need a separate sheet with a list of names, for each name we need
the tasks that person is responsible for or working on.
You should get people to give you the data for that sheet.
About the new electical boxes, when you say “right side” is that the robot’s
right or your right when looking at the front of the robot? Please be clear.
%%%%%%%%
About Stephan Constantin:
Much more detail would be nice. For example, what looks promising, or not,
about mplay? Does it cost money? How much work might it take to get it
to do what? What are the alternatives that you know about? How did you
find it?
You might look into borrowing a camera from the motion capture system
in the IGERT LAB (see Madhu contact information above).
%%%%%%%%
About James Doehring:
Max and Rohit are also doing experiments on the legs and so on? Do you
use each other’s knowledge and experience about how to hold the robot in
such experiments?
When you say “match theoretical predictions” you need to be clear that you
mean that your torque data now well matches your predictions for torque based
on measurements you made of spring constants, lengths, and rest lengths.
What was your previous error that prevented such agreement? How did you find that?
For sure you should be talking with Rohit or Pranav about predicitons of
leg swing motion.
YOu should give about a 5 minute clear presentation at our coming meeting.
You should see if anyone in the lab can read your item 1 from 10/26/07 and
understand what you did. I think you could have made it understandable with
just a few more well-chosen words.
%%%%%%%%
About Greg Falco:
We still need the data sheet www pages for the lab chemicals
When you say “a specific account” why not just say which one?
When you say “Andy” do you mean Andy Ruina or Andy Vail?
I’d like to see that publicity www site you now have, and install it on
the lab www site.
What “wooden structures”? Unscrewed what? Why?
About cameras: I will buy one. Right now I am bidding for a Canon A560.
%%%%%%%
About Hajime Furukawa:
Overall I couldn’t tell if your boards are totally ready to plug on to the plug a pods
and use, or not.
%%%%%%%%%%%%%%
About Rohit Hippalgaonkar:
Please update more frequently (approximately each day you work).
When you say “experimental results” you need to say what was measured.
When you say “error” it is not clear what two things you are comparing.
%%%%%%
About Avtar Khalsa:
Should be clear about what “code”. Do we have a system set up or planned for
the sattelite processors?
Also, have you thought about how we keep track of the legs and state of the
machine? Do we keep track of the stance leg and swing leg or of the inner leg
and outer leg? I think we actually need to be able to do things either way
depending on instantaneous desires and needs.
There is no good text search for finding declarations?
Also, Sam and Jason seem to have ideas about how to deal with the declarations.
Somehow we should figure out how not to disrupt the whole modularity
by this kind of thing.
%%%%%%
About Sam Lee:
About once a week or so could you put in a summary in language that
the rest of us (or me, anyway) can understand? I see you in the lab working hard,
but I’d like some kind of general sense of what you are stuck on and
what you are making progress on.
%%%%%%
About YIngyi Tan:
Overall could you report on the status of the hip sensor. I think you
have it all working, right? Great. Can you give a presentation on
it on Friday (5-10 min).
%%%%%%
About Kevin Ullmann:
Please say what program rather than “the program”. In this case I
know what you mean, but some context is needed for the quick readers.
Your data-graphing program is getting to a point that I think it will be
great to use. Especially with video.
For example, if we had it working with video it would have been good to use
just in the past day or so to help figure out if the new IMU (Inertial
Measurement unit) was giving sensible data on the old (chain powered) robot.
%%%%%%
About Max Wasserman:
Seems like you guys are pretty much done on the analysis side of this
project. You should write up, with good illustrations of your
methods (and photos of various experimental details, that which you have.
Here are comments to all of you on your google blog entries. All of you should
glance at all of this
Generally most of you need a few more words and phrases so that someone reading your blogs
can know what you are talking about. Especially if this is supposed to serve, possibly,
as long term documentation.
Most especially, your entries from Tuesday night before a pay period ends should
be especially clear. Those are what I read first.
%%%%%%
About Carlos Arango:
I played with the IMU also. Fun. I recommend playing with it to everyone.
I am not convinced that the excursions you see on the angular velocity plots
are not real signal. I think you might be seeing real features of the leg motions.
Some suggestions:
1) Plot pitch angle (not angular rate) to get a better sense of the actual motion.
2) Compare with output from the motion capture equipment (in the lab
directly under the machine shop, contact Madhu
(madhusudhan Venkadesan
On 11/5/07 you wrote about CAN bus code. CAN bus on what device? Whose buffer read
function? Where is this work documented?
On 11/2/07 you refer to “the code”. What code? On what sensor? Where is that code
stored? Where are your changes documented?
%%%%%%%%
About John Buzzi:
The Gant chart is great. Some things we need to fix/notate:
a) one person can’t do two one day tasks in one day.
b) What are the blue check marks at left column, “done”?
c) How to make clear non-critical paths (tasks that are on-going
like making the mass distribution symmetric)
d) Need a separate sheet with a list of names, for each name we need
the tasks that person is responsible for or working on.
You should get people to give you the data for that sheet.
About the new electical boxes, when you say “right side” is that the robot’s
right or your right when looking at the front of the robot? Please be clear.
%%%%%%%%
About Stephan Constantin:
Much more detail would be nice. For example, what looks promising, or not,
about mplay? Does it cost money? How much work might it take to get it
to do what? What are the alternatives that you know about? How did you
find it?
You might look into borrowing a camera from the motion capture system
in the IGERT LAB (see Madhu contact information above).
%%%%%%%%
About James Doehring:
Max and Rohit are also doing experiments on the legs and so on? Do you
use each other’s knowledge and experience about how to hold the robot in
such experiments?
When you say “match theoretical predictions” you need to be clear that you
mean that your torque data now well matches your predictions for torque based
on measurements you made of spring constants, lengths, and rest lengths.
What was your previous error that prevented such agreement? How did you find that?
For sure you should be talking with Rohit or Pranav about predicitons of
leg swing motion.
YOu should give about a 5 minute clear presentation at our coming meeting.
You should see if anyone in the lab can read your item 1 from 10/26/07 and
understand what you did. I think you could have made it understandable with
just a few more well-chosen words.
%%%%%%%%
About Greg Falco:
We still need the data sheet www pages for the lab chemicals
When you say “a specific account” why not just say which one?
When you say “Andy” do you mean Andy Ruina or Andy Vail?
I’d like to see that publicity www site you now have, and install it on
the lab www site.
What “wooden structures”? Unscrewed what? Why?
About cameras: I will buy one. Right now I am bidding for a Canon A560.
%%%%%%%
About Hajime Furukawa:
Overall I couldn’t tell if your boards are totally ready to plug on to the plug a pods
and use, or not.
%%%%%%%%%%%%%%
About Rohit Hippalgaonkar:
Please update more frequently (approximately each day you work).
When you say “experimental results” you need to say what was measured.
When you say “error” it is not clear what two things you are comparing.
%%%%%%
About Avtar Khalsa:
Should be clear about what “code”. Do we have a system set up or planned for
the sattelite processors?
Also, have you thought about how we keep track of the legs and state of the
machine? Do we keep track of the stance leg and swing leg or of the inner leg
and outer leg? I think we actually need to be able to do things either way
depending on instantaneous desires and needs.
There is no good text search for finding declarations?
Also, Sam and Jason seem to have ideas about how to deal with the declarations.
Somehow we should figure out how not to disrupt the whole modularity
by this kind of thing.
%%%%%%
About Sam Lee:
About once a week or so could you put in a summary in language that
the rest of us (or me, anyway) can understand? I see you in the lab working hard,
but I’d like some kind of general sense of what you are stuck on and
what you are making progress on.
%%%%%%
About YIngyi Tan:
Overall could you report on the status of the hip sensor. I think you
have it all working, right? Great. Can you give a presentation on
it on Friday (5-10 min).
%%%%%%
About Kevin Ullmann:
Please say what program rather than “the program”. In this case I
know what you mean, but some context is needed for the quick readers.
Your data-graphing program is getting to a point that I think it will be
great to use. Especially with video.
For example, if we had it working with video it would have been good to use
just in the past day or so to help figure out if the new IMU (Inertial
Measurement unit) was giving sensible data on the old (chain powered) robot.
%%%%%%
About Max Wasserman:
Seems like you guys are pretty much done on the analysis side of this
project. You should write up, with good illustrations of your
methods (and photos of various experimental details, that which you have.
November 7, 2007
11/6/07 Avtar Khalsa
7:30 - 11:30
1. Continued working on the code. -4 hrs
The issue that I'm presently stuck on has to do with declarations. I have to move all of them to the header file to make this work correctly. Unfortunately, tracking them all down in several thousand lines of code is somewhat tedious, especially when they are in multiple files themselves. I also spent some time looking into linker order issues. CodeWarrior allows you to set the order that the files are linked in. Most linkers handle this automatically. I'm still trying to figure out what the order should be in this case
1. Continued working on the code. -4 hrs
The issue that I'm presently stuck on has to do with declarations. I have to move all of them to the header file to make this work correctly. Unfortunately, tracking them all down in several thousand lines of code is somewhat tedious, especially when they are in multiple files themselves. I also spent some time looking into linker order issues. CodeWarrior allows you to set the order that the files are linked in. Most linkers handle this automatically. I'm still trying to figure out what the order should be in this case
11/6/07 Avtar Khalsa
3:30 - 5:30
1. Worked on correcting errors in the split up code -2 hrs
The initial splitting of the code turned out to generate over 100 errors. The problem was that not all the declarations had been moved. Additionally there was code in a second file that I didn't realize initially that needed to be both modified and moved to make it compile.
1. Worked on correcting errors in the split up code -2 hrs
The initial splitting of the code turned out to generate over 100 errors. The problem was that not all the declarations had been moved. Additionally there was code in a second file that I didn't realize initially that needed to be both modified and moved to make it compile.
11/7/07 Stephane Constantin
10 to 11 am
1) Finished drawing the schematic for the CAN bus transceiver adaptor
1) Finished drawing the schematic for the CAN bus transceiver adaptor
11/6/07 Stephane Constantin
8pm - 9:30pm (at home)
1) Video capture
I found and explored a matlab application called 'mplay' that could be adapted for our purposes.
1) Video capture
I found and explored a matlab application called 'mplay' that could be adapted for our purposes.
11/3/2007 Max Wasserman
12:00 noon-1:00 pm (Saturday)
1. Inertial Property Testing
----------------------------------------------------------
1. Rohit and I re-tested the inertial properties of the outer legs to see what had changed since the new masses were added.
1. Inertial Property Testing
----------------------------------------------------------
1. Rohit and I re-tested the inertial properties of the outer legs to see what had changed since the new masses were added.
November 6, 2007
11/05/07 Hajime Furukawa
3:00PM - 4:00 PM
1) Tested Boards (1h)
--------------------------------------------------------------
1)Tested the second the third sensor boards' voltage regulators. Measured the output voltage and the power efficency of both to make sure they were acceptable (and they were!).
1) Tested Boards (1h)
--------------------------------------------------------------
1)Tested the second the third sensor boards' voltage regulators. Measured the output voltage and the power efficency of both to make sure they were acceptable (and they were!).
11/6/07 Ullmann Kevin
11:00AM - 1:00PM (2 hrs, from home)
1. Debugging (45 minutes)
2. Fixed resize problem (45 minutes)
3. Commented code (30 minutes)
1. I ran the program many times and tested it under various conditions to try to iron out some bugs.
2. I fixed the resize problem; previously if a user chose a new data file to work with, the figure window would revert to its original size and position, now when a user resizes the window, it will stay that way regardless of whether they choose a new data file.
3. I added comments to many lines of code so that it is easier to understand.
1. Debugging (45 minutes)
2. Fixed resize problem (45 minutes)
3. Commented code (30 minutes)
1. I ran the program many times and tested it under various conditions to try to iron out some bugs.
2. I fixed the resize problem; previously if a user chose a new data file to work with, the figure window would revert to its original size and position, now when a user resizes the window, it will stay that way regardless of whether they choose a new data file.
3. I added comments to many lines of code so that it is easier to understand.
11/06/07 Yingyi Tan
5.00pm - 6.30pm
1. Connected the wires to the battery charger connector clip (0.25h)
2. Drilled holes in the robot to attach the satellite box to bottom of robot (1.25h)
----------------
1. Jason attached the wire that had come off its header to a new header. Had to insert the wires according to previous diagram I made into a new battery charging connector.
2. John prepared the boxes to hold the satellite boards. I drilled the holes in the bottom of the robot's center box to attach these boxes to the robot.
1. Connected the wires to the battery charger connector clip (0.25h)
2. Drilled holes in the robot to attach the satellite box to bottom of robot (1.25h)
----------------
1. Jason attached the wire that had come off its header to a new header. Had to insert the wires according to previous diagram I made into a new battery charging connector.
2. John prepared the boxes to hold the satellite boards. I drilled the holes in the bottom of the robot's center box to attach these boxes to the robot.
11/6/07 Carlos Arango
3:10-4:40pm
1. Acquired data from IMU and then discussed it with Jason
-----------------------------------
1. I attached the IMU to the old robot again and showed the data to Jason. He was disappointed because it shows the same amount of noise we got from our previous sensors. Under his suggestion, I padded the IMU with memory foam and got new measurements. I then plotted both sets in Excel; the foam does not seem to help much. Lastly, I tried holding the IMU to my thigh as I walked and looked at the data: there was a lot of noise because it was not attached rigidly, but the data exhibited the same characteristics as that from the robot (mainly heel-strike spikes, even when I walked "smoothly")
1. Acquired data from IMU and then discussed it with Jason
-----------------------------------
1. I attached the IMU to the old robot again and showed the data to Jason. He was disappointed because it shows the same amount of noise we got from our previous sensors. Under his suggestion, I padded the IMU with memory foam and got new measurements. I then plotted both sets in Excel; the foam does not seem to help much. Lastly, I tried holding the IMU to my thigh as I walked and looked at the data: there was a lot of noise because it was not attached rigidly, but the data exhibited the same characteristics as that from the robot (mainly heel-strike spikes, even when I walked "smoothly")
11/6/07 Gregory Falco
3:30-4:00
1. MSDS (30 min)
-------------------------
Assigned names to each number corresponding with chemicals on the to-be website. Still needs some fixing up before publishing.
1. MSDS (30 min)
-------------------------
Assigned names to each number corresponding with chemicals on the to-be website. Still needs some fixing up before publishing.
11/6/07 Gregory Falco
12:00-2:00PM
1. Accounts (2 hrs)
-------------------------------------
Today I found a discrepency of $2,423.75 dollars in our accounts. We do not have receipts for these charges so I have reason to believe they are mistakes by accounting. Also, there were three overhead charges applied to a specific account that should not be there. Finally, we still do not have an account for the NSF grant which was promised to us earlier in the semester. My next steps include discussing this with accounting after confirming charges with Andy.
1. Accounts (2 hrs)
-------------------------------------
Today I found a discrepency of $2,423.75 dollars in our accounts. We do not have receipts for these charges so I have reason to believe they are mistakes by accounting. Also, there were three overhead charges applied to a specific account that should not be there. Finally, we still do not have an account for the NSF grant which was promised to us earlier in the semester. My next steps include discussing this with accounting after confirming charges with Andy.
11/6/07 Avtar Khalsa
12:30AM - 2:00AM
1. Finished splitting up code ~2 Hrs
---------------------------------------------
1. I finished the first attempt at splitting the code up. If this Kobrain code works, the next step is to try to split it into many more files, instead of just being in 3 files like it is now. If that works, the next step is to do the same thing with the actual ranger code.
1. Finished splitting up code ~2 Hrs
---------------------------------------------
1. I finished the first attempt at splitting the code up. If this Kobrain code works, the next step is to try to split it into many more files, instead of just being in 3 files like it is now. If that works, the next step is to do the same thing with the actual ranger code.
11/05/07 Avtar Khalsa
4:30PM to 6:30PM
1. Worked on Splitting up the code ~2 hrs
---------------------------------------
I started splitting up the kobrain code into multiple files. The idea being that its much easier to edit and handle code in general with it split into many files. In this case, we can test if its working very easily by splitting up what is effectively working code, and seeing if anything changes when we run it. If nothing does, then splitting it up was successful. If it starts working any differently, then something went wrong.
1. Worked on Splitting up the code ~2 hrs
---------------------------------------
I started splitting up the kobrain code into multiple files. The idea being that its much easier to edit and handle code in general with it split into many files. In this case, we can test if its working very easily by splitting up what is effectively working code, and seeing if anything changes when we run it. If nothing does, then splitting it up was successful. If it starts working any differently, then something went wrong.
11/05/07 Lee Sam
Time in 10am
Time out 4pm
1) Manged to get SPI to communicate from the Master(3180) to Slave(MAC7111) - data sent from the master to the slave
2) data from the slave to the master is still not working - nothing is being received or the interupts are just not triggering
3) External Circuit seem to be working well under the testing conditions
4) SPI_B for the MAC7111 low cost evaluation board seem to be probably wired to other pins that is not listed in the schematic, hence we would only us SPI_A for testing purposes.
Time out 4pm
1) Manged to get SPI to communicate from the Master(3180) to Slave(MAC7111) - data sent from the master to the slave
2) data from the slave to the master is still not working - nothing is being received or the interupts are just not triggering
3) External Circuit seem to be working well under the testing conditions
4) SPI_B for the MAC7111 low cost evaluation board seem to be probably wired to other pins that is not listed in the schematic, hence we would only us SPI_A for testing purposes.
November 5, 2007
11/05/07 Yingyi Tan
5.00pm - 7.30pm
1. updated CAD drawings of hip insert part based on actual parts in robot (0.5h)
2. Removed switch and battery plug from back of robot and took out the connector (2.5h)
-----------------------
1. The actual parts in the robot are slightly different from those in the drawings. Updated the CAD drawings so the dimensions are of the same size as the actual parts. Made an assembly in CAD of the entire hip encolder assembly.
2. In order to place the satellite boxes, the switch and battery charging connector had to be removed from the back of the robot. However, the connector was too large to fit through the holes in the bottom of the box. I had to cut apart the connector. This involved using a dremel tool to remove the plastic covers, cutting with a wire cutter and finally pulling the wires out of the holes in the connector.
1. updated CAD drawings of hip insert part based on actual parts in robot (0.5h)
2. Removed switch and battery plug from back of robot and took out the connector (2.5h)
-----------------------
1. The actual parts in the robot are slightly different from those in the drawings. Updated the CAD drawings so the dimensions are of the same size as the actual parts. Made an assembly in CAD of the entire hip encolder assembly.
2. In order to place the satellite boxes, the switch and battery charging connector had to be removed from the back of the robot. However, the connector was too large to fit through the holes in the bottom of the box. I had to cut apart the connector. This involved using a dremel tool to remove the plastic covers, cutting with a wire cutter and finally pulling the wires out of the holes in the connector.
11/5/07 Gregory Falco
1:00-5:00PM
1. Publicity website (3 hrs)
2. Lab Accounts (1 hr)
-----------------------------------------
1. I found that the website I had created had many broken links on it. In order to fix them I had to remake each file in order to connect to its predecessor and store it in a new folder. Now, there is a website for each general topic that leads to a list of sources (ie. magazines) which contain the articles.
2. I worked on all the financial account info for the month of October and found a large discrepancy between the Sept. accounts and October ones. Also, i cannot find out NSF super account. I plan on speaking to you or Jason about this and then work with accounting to see where our money is going.
1. Publicity website (3 hrs)
2. Lab Accounts (1 hr)
-----------------------------------------
1. I found that the website I had created had many broken links on it. In order to fix them I had to remake each file in order to connect to its predecessor and store it in a new folder. Now, there is a website for each general topic that leads to a list of sources (ie. magazines) which contain the articles.
2. I worked on all the financial account info for the month of October and found a large discrepancy between the Sept. accounts and October ones. Also, i cannot find out NSF super account. I plan on speaking to you or Jason about this and then work with accounting to see where our money is going.
11/5/07 Ullmann Kevin
3:30 - 4:30PM (1 hr)
1. Spoke to Andy about current version of graphics program (20 minutes)
2. Minor changes to program (10 minutes)
3. Fixed data path issue (30 minutes)
1. I showed the current version of the graphics program to Andy, he gave me some suggestions.
2. I fixed small issues that he noticed, such as mislabeling "t" as "x" in the max/min select window.
3. I changed the program so that when the user selects a data file, the path will be saved so that when they want to select another data file, the default directory will be the one that the previous file was saved in. This saves the user the trouble of searching through their directories every time they want to select a file.
1. Spoke to Andy about current version of graphics program (20 minutes)
2. Minor changes to program (10 minutes)
3. Fixed data path issue (30 minutes)
1. I showed the current version of the graphics program to Andy, he gave me some suggestions.
2. I fixed small issues that he noticed, such as mislabeling "t" as "x" in the max/min select window.
3. I changed the program so that when the user selects a data file, the path will be saved so that when they want to select another data file, the default directory will be the one that the previous file was saved in. This saves the user the trouble of searching through their directories every time they want to select a file.
11/05/07 John Buzzi
10:30am-12:20pm
1) Updated GANTT Chart (1hr 30 min)
2:15pm-4:00pm
2) Printed and assembled GANTT (20 min)
3) Developed a tentative plan for satellite processor box mounting (1 hr 15 min)
------------------------
1) After our meeting this friday i reviewed and made sense of the necessary updates to the GANTT chart. I went to upson computer lab in order to acess the microsoft project software. I spent my time filling in the various catagories of information for each of the project tasks.
2) In the lab i printed the newly updated GANTT chart and combined the six pages into one large chart. The chart is currently poseted on the wall to the right as you enter the lab. If you are reading this please review this chart and make necessary changes regarding your portion of the project.
3) The satellite processors are to be mounted in these gray plastic boxes that will hang under each of the 3 parts of the body of the robot. I spent my time planning how to make these boxes actually fit there since there are currently multiple components obstructing this placement. For the right side box, all that needs to be done for mounting is to move the electrical components temporarily while the box is placed. For the middle box the battery wires can be re-routed through the back of the bot so the box will fit. The left side is most difficult because the beeper, wireless antenna, and the serial port adapter will need to be moved.
1) Updated GANTT Chart (1hr 30 min)
2:15pm-4:00pm
2) Printed and assembled GANTT (20 min)
3) Developed a tentative plan for satellite processor box mounting (1 hr 15 min)
------------------------
1) After our meeting this friday i reviewed and made sense of the necessary updates to the GANTT chart. I went to upson computer lab in order to acess the microsoft project software. I spent my time filling in the various catagories of information for each of the project tasks.
2) In the lab i printed the newly updated GANTT chart and combined the six pages into one large chart. The chart is currently poseted on the wall to the right as you enter the lab. If you are reading this please review this chart and make necessary changes regarding your portion of the project.
3) The satellite processors are to be mounted in these gray plastic boxes that will hang under each of the 3 parts of the body of the robot. I spent my time planning how to make these boxes actually fit there since there are currently multiple components obstructing this placement. For the right side box, all that needs to be done for mounting is to move the electrical components temporarily while the box is placed. For the middle box the battery wires can be re-routed through the back of the bot so the box will fit. The left side is most difficult because the beeper, wireless antenna, and the serial port adapter will need to be moved.
11/5/07 Carlos Arango
12:45-3:15 pm
1. Debugged part of the CAN bus code (1.5 hr)
2. Installed and tested comercial IMU software on old robot (1 hr)
--------------------------------------------------
1. The buffer read function had some copy-paste errors. I went through and, enabling one buffer at a time, found and fixed their individual bugs. This took so long because there are 15 working buffers
2. The IMU comes with a LabView program that reads its output and displays it in useful plots. Sam and I installed this software on the programming laptop. We then mounted the IMU on the older 2D biped and got some data to see the effects of foot impact. Sam believes the noise is not so great that we can't compensate for it with software "tricks"
1. Debugged part of the CAN bus code (1.5 hr)
2. Installed and tested comercial IMU software on old robot (1 hr)
--------------------------------------------------
1. The buffer read function had some copy-paste errors. I went through and, enabling one buffer at a time, found and fixed their individual bugs. This took so long because there are 15 working buffers
2. The IMU comes with a LabView program that reads its output and displays it in useful plots. Sam and I installed this software on the programming laptop. We then mounted the IMU on the older 2D biped and got some data to see the effects of foot impact. Sam believes the noise is not so great that we can't compensate for it with software "tricks"
November 4, 2007
10/31 thru' 11/05 Rohit Hippalgaonkar
Hours:
-------
Wednesday 10/31 - 3 pm to 4.30 pm (worked from home)
Thursday 1/11 - 4 pm to 5.30 pm
Friday 2/11 - 9.45 am to 12.15 pm, 1 pm to 4.30 pm
Saturday 3/11 - 11 am to 3 pm
Sunday 4/11 - 4.30 pm to 7pm
1. Experiments to account for added masses (3 hours)
2. Testing on code with new results (5.5 hours)
3. Accounting for effect of friction in experiments (3.5 hours)
4. Meetings (2 hours)
----------------------------------------------------------------------------------------------------
1. Measured the inertia and mass-moment values of both legs once again - to account for change in values due to James' springs and John Buzzi's replacement bar on the inner leg.
Surprisingly the values of the 3 relevant mass properties of the inner leg are now all greater than the corresponding values of the outer leg. The m*y values being very close together while the m*z and inertia values show a big difference. Extrapolations using data from the old experiments (done without the added masses) and the mass and location values supplied by James and John predict the experimental results to a satisfactory degree.
2. Tested once more with the new experimental values - with the consideration that we might want to add another battery.
3. Finally got MATLAB to solve the solid friction problem - used smoothing function from Manoj Srinivasan's thesis to model solid friction at omega = 0. Error due to simplifying approximations is less than 2% - that is, error in natural frequency values, with and without having a friction term in the governing ODE. It seems one must neglect the periodic force applied, and say it doesn't make much difference as we cannot model that force.
-------
Wednesday 10/31 - 3 pm to 4.30 pm (worked from home)
Thursday 1/11 - 4 pm to 5.30 pm
Friday 2/11 - 9.45 am to 12.15 pm, 1 pm to 4.30 pm
Saturday 3/11 - 11 am to 3 pm
Sunday 4/11 - 4.30 pm to 7pm
1. Experiments to account for added masses (3 hours)
2. Testing on code with new results (5.5 hours)
3. Accounting for effect of friction in experiments (3.5 hours)
4. Meetings (2 hours)
----------------------------------------------------------------------------------------------------
1. Measured the inertia and mass-moment values of both legs once again - to account for change in values due to James' springs and John Buzzi's replacement bar on the inner leg.
Surprisingly the values of the 3 relevant mass properties of the inner leg are now all greater than the corresponding values of the outer leg. The m*y values being very close together while the m*z and inertia values show a big difference. Extrapolations using data from the old experiments (done without the added masses) and the mass and location values supplied by James and John predict the experimental results to a satisfactory degree.
2. Tested once more with the new experimental values - with the consideration that we might want to add another battery.
3. Finally got MATLAB to solve the solid friction problem - used smoothing function from Manoj Srinivasan's thesis to model solid friction at omega = 0. Error due to simplifying approximations is less than 2% - that is, error in natural frequency values, with and without having a friction term in the governing ODE. It seems one must neglect the periodic force applied, and say it doesn't make much difference as we cannot model that force.
11/02/07 Yingyi Tan
1:00 - 6:00pm
1. Soldered ribbon cable to sensor board (1.5h)
2. lab meeting (1.5h)
3. Installed magnet into hip insert and zeroed it (2h)
-------------------------
1. Soldered a ribbon cable to the sensor board based on the data sheet and the wiring position given by Jason.
2. Lab meeting.
3. Had to use silicon to glue the magnet into to the hip insert. Connected the sensor board to a satellite board and connected it to the computer. There was no response. Used a multimeter to test the wires on the sensor board. They were all working fine. Then took the sensor board out and placed a magnet in front of it. The sensor produced a response. Concluded that the magnet was too far away from the sensor in the robot. Took the magnet out and tested with the sensor to find the right angle the magnet had to be at in order to produce a good range of values when the robot walked. Made a marking on the robot and magnet. Used a piece of cable tie to fit into the hip insert. Then used silicon to put the magnet over the cable tie. Retested the whole set-up which produced a reading.
1. Soldered ribbon cable to sensor board (1.5h)
2. lab meeting (1.5h)
3. Installed magnet into hip insert and zeroed it (2h)
-------------------------
1. Soldered a ribbon cable to the sensor board based on the data sheet and the wiring position given by Jason.
2. Lab meeting.
3. Had to use silicon to glue the magnet into to the hip insert. Connected the sensor board to a satellite board and connected it to the computer. There was no response. Used a multimeter to test the wires on the sensor board. They were all working fine. Then took the sensor board out and placed a magnet in front of it. The sensor produced a response. Concluded that the magnet was too far away from the sensor in the robot. Took the magnet out and tested with the sensor to find the right angle the magnet had to be at in order to produce a good range of values when the robot walked. Made a marking on the robot and magnet. Used a piece of cable tie to fit into the hip insert. Then used silicon to put the magnet over the cable tie. Retested the whole set-up which produced a reading.
Subscribe to:
Posts (Atom)
