We were restricted yet again, but this time even more so. We were not able to have Internet access nor were we able to broadcast a network within the stadium. But we were still able to record everything and it all went as smoothly as at the Pittsburgh Regional.

This time however we were able to provide anyone that came up to us with custom printouts with their own selected custom sorts. That’s right! We have a print feature now!

I’m very sorry that the information has not found its way to the Internet as of yet, but hopefully within the next couple of days the data we collected will get there and the sort and match pages will be opened up.

Since we were unable to have Internet access at the venue, we instead had to rely on the GHX laptop that I installed an Apache server on. We were able to broadcast our own wireless signal though, and it worked perfectly. GHX was able to reach the pits and the rest of the stands as well. Everything was moved smoothly until several glitches were found, none of which were cataclysmic but still they needed fixing. They weren’t too hard to fix, and so the website past the first test.

The new data we collected still needs to be uploaded onto the Internet. As of now, the lap-top’s information has been backed-up twice since all information was entered. I expect that this weekend will hold enough free time to allow me to upload the information and fix/adjust anything on the internet version. I will also open the sort-feature and Match-details to the public.

To test out the scouting system for this year, several of our members went to the rookie Kettering competition. They came back with some bad news! The 4 person scouting system was way too hard to pull off, so we are going to have to pull it back up to a 6 person scouting system. The new scouting sheets can be found on the new Gearhead Crossing Web site’s FAQs page. 

The new download is provided in a Zip format, and the forms and the Sorting System are separated into two files.

Finally, the website is up and running on the internet. It took awhile because of the way I set up the database queries. On the system I was using, it didn’t care if I used capital letters when making table calls. Well the server thats running it now does!

But after awhile, Josh and I got it to work. We even uploaded some test data, check out team 1189! There still seems to be some problems with how the website handles the links since there are 2 URLs that point to the same website. Heres what I mean:

• http://gpgearheads.org/ghx/2008
• http://gearheadcrossing.org/2008

Both URLs point to the same page and this causes some problems, so I fixed it so the page wont show up if you try and visit the first one, it will just give you a link to go to the new site.

Well, every page has been completed! I’m still testing to make sure everything is running correctly, and that the database is being edited correctly for every operation. I’m also working on a README for anyone that wants to learn how the code works, the public interface works, and anything else that I can think of.

The mobile edition has also been completed and is fully functional. It even has a sort function that is just as easy to use as the regular version. Overall, I’m pleased at how well things are proceeding and I hope I can get things up on the internet soon.

This is my first post in a while, but I have an update on what’s going on with the shirt, hoodie, and swag orders we placed in late December.

Dr. D emailed me and confirmed that the shirts, hoodies, and swag are in production as of this past week. I should have the week the orders will be arriving in the next week or so.

Afer some delay (ok, a lot of delay), Mallory S. and I have managed to update the old content to create the 2008 Gearhead’s website! There’s not much left to do content-wise, so I’m planning on finishing (checking for spelling, grammar, accuracy, etc) everything tonight. Group leaders, thank you for your updates–the miracle that we performed this afternoon couldn’t have happened without them!

Yesterday, work on the BOM, or bill of Materials, got a Nitrous injection. The list of parts and materials used in this year’s robot started Saturday, but most of the important work was done Sunday. A mentor familiar with the most interesting capabilities of Excel worked with me and another student in order to create an easily managed and well organized list that can answer almost any question concerning what is on the 2008 Overdrive Robot.

Other than work on the BOM, design has reached a standstill, all work has been handed off to the Build and Prototype group in order to finish the robot in time.

Sorry about not posting for the past week, I have been busy designing the robot and handing off prints to the build team.  Hopefully I’ll be able to help with building sometime, but designing is always fun.  Anyways, I’ll get to it.

Over the last week the design and build groups have been working even harder in order in order to finish the robot on time.  The design group has handed off chain specifications as well as elevator and mast designs to the build group for construction.

The axle assembly has been going well and the new half inch hex axle design, which I am proud to say I helped create, looks highly promising.  The half inch hexagonal axles, which are about 5.6 inches long, and the hexagonal sleeves, which allow the axle to rotate in the IGUS Iglide bearings, have been fabricated.  Welding the sprockets to the aluminum hubs, or “sandwiches,” as well as final assembly have yet to take place.

The chain sizes involved look very well calculated.  The procession of chain goes like so:  First, a sprocket “sandwich,” or two sprockets on an aluminum hub, mounted on the transmission axle transfers rotational force through two chains to the middle and back wheels.  The transmission sandwich is unique-it is the only sandwich with a key way. All other sandwiches rest on half inch diameter hexagonal axles.  So, now we have four wheels driven-remember there is one of these systems on each side.  A second sprocket “sandwich” attached to the middle hexagonal axle then transfers the same force to the front wheel.  The middle and transmission axles are the only hubs that need to hold two sprockets.  So far, the chain lengths match up perfectly.  (Except for one chain length, the axle position for the transmission on the left rail is off just enough to upset us, but it is being taken care of)

The mast has seen its main components fabricated and the elevator, or trolley for the arm whichever you prefer,  has seen most of its fabrication as well.  The mast and elevator operate like so:  The mast is comprised of three parts on both sides, a 2×2 aluminum tube with a .75 inch slot on the inside surface allows an elevator to move up and down inside it.  These lengths of aluminum are at a 80 degree angle, which allows the capturing of track balls when the mast is shifted all the way to the back.  Vertical supports all the way in the back support the system and aluminum stringers keep the mast from falling over.  The last part is a plate, which holds both together at the top with 4×1/4-20 bolts.  The elevator is quite simple as well, two vertical 2×1 aluminum tubes and two crosspieces of the same size comprise the frame.  The arm will attach to this elevator via four .25 inch thick plates, I personally used stress analysis software on Autodesk Inventor and compared different configurations for lightness and durability.  They look pretty cool as well.

Well, that’s all for now, check back in about three days everybody!

On Saturday we were lucky enough to have Clemenson Hall, for the second Saturday in a row, for driving practice. We started a little late but because of the team working together and listening to each other, we made up for lost time and ran more efficient trials! A+ guys! This goes to show that practice does help and that the team is growing together!

Keep it up! We are going to have a strong showing this year at competitions!

If you’ve noticed in the forums under Imagery Managment, there’s new posts for the bot name.

Actually, I was given the choice to select the name as Imagery Group Leader or have the whole team take a vote on a name…….I’ve selected the name of our bot after going through the forums.

The 2008 GP Gearheads bot name is Atlas, though common, many fellow gearheads along with me feel it’s appropriate for our bot this season.

On Saturday the school allowed us use Clemenson Hall from 12:30-3:00 as a field to drive the robot on. It worked out nice and we ended up burning up all six batteries we brought! All of the drivers that came dove very well, but a few stood out from the rest. Over all, it was a great day for practice and will prevent problems from occurring later on. All ready we have practice more than last year and we have reserved Clemenson again for next weekend.

Practicing at Clemenson is bringing us more than just driving experience, we are working together to setup and take down; and we are going through the motions for the competitions. Also, since we are putting the robot through such strenuous tests (roughly 2 hours of constant running), we are finding little things, things like increased wheel wear on certain wheels, are going to be useful to know before we go into the first competition.

OK, a lot has been done, revised, and tested for the new site.

• The ‘Add/Edit Game year’ feature has now been added and successfully tested. With this done, the rest of the site is coming into place a lot easier.  (All the global variables now can get their values from the database)
• The mobile feature has been tested, and it works. The site just needs a little revision to fit on the smaller screens.
• I’ve now begun work on the ‘Add/Edit Venue’ and ‘Add/Edit Matches’  pages simultaneously. They have similar features and will both impact each other.

Over all, it is beginning to look like the site i had planed out in the beginning. So I’m pleased.

Our work sessions are scheduled as follows for the duration of the 2008 FRC build season ( starting 01/07/08 through 02/19/08 ) -

- Mondays through Fridays, 6:00-9:00 pm
- Saturdays, 12:00-9:00 pm, with the weekly team meeting starting at 12:00, and dinner at 5:00 pm in the 2nd floor Art Classroom for those who are working at this time
- Sundays, 1:00-6:00 pm

This schedule may change as circumstances dictate, so please check the forums and team calendar daily for updates! There’s always something for you to do, and many hands make for light work!

Today the bot’s frame rails were completely finished on the mill and now the chassis assembly has begun!

On Thursday 1/23 after school, I met up with Kara, Josh, and Mr. C to start brainstorming for the 2008 Website’s new style/image.

I’ll be trying different things with color and logos using HTML code; the GP Gearheads website is going to be “pimped!”

The chassis is beginning to come together.  Insets for retainment tabs and holes for bolts were milled into the frame rails yesterday.  The insets make space for circular tabs on the corner plates and strengthen the joints between the frame rails.  Helio coils will be anchored into the frame rails in order to create stronger threads, and eighth of an inch of aluminum is not enough to hold a bolt.

The lightening holes and transmission slots will be milled Saturday and Sunday, the 26th and 27th.  These are the finishing touches and will allow us to install our new hexagonal shaft drive train and super shifters.  I know I’m excited about seeing our new design perform in driving tests.

The next step is to begin final construction of  the arm and elevator.  To do that we’ll start creating the base for the structure–parallel  rails on top of the frame that also serve as transmissions mounts.  A group of students have been experimenting with different manipulator configurations and will be ready to explain their findings to the design and build team.

In the past couple days the Build and Prototype group has been working in full swing. The Design and Engineering group lead by Charles W. has been pounding out final part designs and NC paths for the mill. Peter S. as making parts on the Lathe, Matt S. and other Gearheads have been work on the prototype arm and manipulator with great success, I have been taking the new NC mill paths and cutting the new bot rails and other bot parts on the mill.

A couple days ago we received the data logger that I mentioned before. We had to run out and get an SD card and a reader, but those are easy enough to come across. I did a quick hookup with a breadboard to the TTL port of the RC. I didn’t fully explore the configuration, but I made the RC write to that port instead and I’ve confirmed that it does indeed log. I’ll need to configure it more to make it make a log which can be read by my LabVIEW diagnostics program.

We also received the ultrasonic range finder. We’ve yet to play with it. As I said before, that’s a lower priority. A more important task is to use the encoder to get wheel speed so PID can be used to get both sides to drive the same speed. Which I played with today. I copied the PID routines used last year, but that was originally for powering a motor until it gets to a position. This doesn’t work to well with the drive train because after it gets to the speed we don’t want it to slow down. So basically before I had to leave, it was increasing the power to get to the speed, then decreasing it as it got closer to the correct speed. When it decreases, the speed goes down too much. I’m not sure. I’ll just have to play with it some more.

Today in the shop we offered the drivers test. The following took it:

Al, Mark , Kara , Ian , Mr. C

Overall, the test scores were good; but we will see tomorrow because Charlie, Mike and Peter are going to take it then. The Coaches test will also be available to take also. Remember, the coaches test will be taken in addition too the drivers test.

On Saturday, (tomorrow), we have Clemenson Hall to stage the driving test using Charlotte.

Hey everyone, I’m coming along!

Members of the design team have been hard at work making parts for Mike P. to mill.  Mike has not only been kicking out parts, but training Andy W.  Looks like the Gearheads may have found someone to take Mike’s spot at the mill next year already.

Lately, Josh C. has been setting up his laptop in the workshop to tweak my drive controls.  It looks good so far!  Josh has also been meeting with Strategy to figure out how to control me out in the field.  It’s important that controls has a clear understanding of what Strategy wants out of me in order to create an effective control system.

As most of you already know, this year’s game includes the movement of the big, red or blue trackballs.  So, in order to manipulate trackballs, I’m going to have an arm!  The arm implements pneumatics in order to open and close my gripping pincer.  The Gearheads have been working hard to get it to perform correctly, but it requires the balance of gripping power and the time needed to build up pressure. If the arm can’t build up enough pressure, the grip on the ball will be weak.  If the grip on the ball is really strong, it takes to long for me to build up air pressure.  So the build team needs to find a good balance.

Overall I’d say the team has bounced back from a slightly hindered week of finals and everyone’s back in the swing of things.  I’m being worked on everyday, and everyday in the shop the Gearheads make great progress.  Thanks guys, and keep up the good work!

We had more issues with the LabVIEW diagnostics program. Apparently the LabVIEW serial thing I was using had a termination character of ASCII 10 (a newline) on by default. Whenever a newline was sent over the port, wham. Everything else after that was marked as 0. Whoops, can’t believe I didn’t notice that earlier. Also, the two 255s in a row being used as the header flag didn’t pan out as much as I wanted. I’m not sure why, but when two 255s showed up naturally (I didn’t think that’d happen much, but when sending a negative long variable… yeah.) So now the header is four 128s in a row. If that happens in nature, it deserves to crash. Seriously.

Another thing about sending negatives. A negative is actually just a really being positive number. I can’t seem to find a LabVIEW function which converts this to a negative. When I do it using what I thought would work, by giving a conversion function a really big number, it just returns the upper bound of the variable. Boo. I think I’m going to have to convert it myself. From memory, this is done by checking the most significant bit, if it’s a 1, then NOT the entire thing and add 1. I wonder if I can make a function which can do this for multiple variable types.. I bet I can. I did something quickly like that for just Words (16 bits), need to figure out how to do it for multiple variable types.. That can be done later.

We ordered a steering wheel for possible use with the chicklet. Not sure if we’ll be able to use it, but we’ll have it in case we can. We also ordered one or two (surprisingly expensive) string pots for use on our bot. I hope they work out, it’d be a shame to have spent that kind of money one something we’ll just end up hacking together.

The driver’s selection process is now available for the team to see. The Coach Selection process is on the way. Both will involve a written test.The driver’s written test is 20 questions long and covers section 7 about the game. You should be able to score a 90% or higher on the test if you want to be a driver, so read the rules.To be a coach you are responsible to know section 9 as well as section 7. This test will be longer and more difficult than the drivers test considering the Coach needs to know the rules AND how to play the game.

Milling is going forward at full bore, with plenty of parts to make and much more math on the way; the milling crew is going to be really busy.  Mike P. started it off with frame corner plates a few days ago and is now moving on to adjuster blocks.  The blocks will allow us to make finite tweaks to chain tension and robot maneuverability.  Andy W. joined him and is learning how to use the mill quickly–it looks like we’ll have a capable miller next year.
The mill uses .nc files, the first of which made the corner plates.  More are on their way for the frame rails themselves, they make up most of the robot chassis.  I am currently working on the frame rails and will have them done soon.

We fried a Victor today. After the new transmissions were put in I was there with the controller to give them power. After we figured out we put in all the PWMs backwards, we were set to test. Well, they moved. One clicks a bit and will have to be taken apart to see what’s going on. The other had issues moving. And then it started smelling funny… then it started smoking… It’s never nice when a Victor starts smoking. At first we couldn’t tell what it was, so we tried it again. It started smoking, and the PWMs weren’t even plugged in. That’s when we found the metal shaving touching the input and a part of the Victor it really shouldn’t have been touching.

Ouch. There goes a Victor.

Yesterday, the 19th, Controls had a session with Strategy. We used a nice giant whiteboard to figure out Strategy’s requirements.

• OI
  • Steering
    • Single Joystick
    • Dual Joystick
    • Steering Wheel
    • Xbox Controller
  • Feedback, LEDs.
• Hybrid
  • IR, maybe even sound, less interference?
  • Drive in straight line
  • Get ball off overpass
    • Rotary limit switch?
  • Actions for the RoboCoach

Strategy also provided a list of hybrid options in order of feasibility.

1. Drive straight
2. Drive straight, turn left
3. Make laps
4. Etc.

We both have more to work out so we can figure out what is possible or not.

Finally got around to playing with the gyro. It needs to be mounted on something so it can be of more use than me just turning it in place. But when set to tenths of degree, I was able to rotate it 180 degrees and see a nice “1800″. The hardest part today would probably be trying to debug integrating Kevin’s Gyro code with the Test Chassis (which was already a combination of his Serial and his Encoder code.) I need to remember that I need to modify the Makefile when I add files to it. At least it works now.

The Test Chassis has the transmissions on it! Woo, that means I have real Encoder data now. Didn’t have much time to play with them though, I had an issue with my LabVIEW Diagnostic program. I had set it up to send the number of variables so the user didn’t need to set it in the LabVIEW panel. I had also decided to have the program count the number of variables, then send it when it sends the starter flag for the packet. Well, this ended up being a Bad Idea®. The starter flag was two 0xFFs. I figured this wouldn’t happen in nature very often. Well, I was wrong. I completely forgot how negative numbers were stored, and I didn’t have the foresight to see that I’d be sending Longs over the span of two variables, making one variable subsequently a 65535.. the packet starter flag. When it saw this it decided to send the variable count again, in the middle of a packet. As you might be able to tell, this completely messed everything up. There were 2 bytes in the middle of the packet which weren’t part of the real data… so I removed that and it should work now. =)

I also fixed up the LabVIEW Diagnostic program. Part of why I made it was so I could set the Program Mode remotely. I explained this in an earlier post, you basically hook up a digital IO port to the programming pin and send a command over the serial to set it. It was also supposed to release control of the serial port so the IFI Loader could upload the program. It did neither of these very well. I’m not sure why though. I think because I was compensating incorrectly for closing the port too quickly after sending the command to set the programming mode. After thinking about it some more, I rearranged how I was sending the close commands, and it started to work. Wooo. I was also using a very ugly way of getting the data out of the stream and onto the front panel. It involved combining the two bytes back into one number. For some reason I was converting the string to a byte array to a cluster and then unbundling the cluster and combining every two slots together. Now I’m just taking the byte array, passing it to its own VI which takes a variable number and outputs the combination of the two bytes.

Yay.

Well, I’ve gotten a date for when the t-shirt, hoodies, and swag for the team will be in. March 1st, or before (if they come earlier) no later then that.

I’ll be notified by Dr. D when the orders arrive, because I know that everyone wants their t-shirts and hoodies. Of course, I know I do.

I really should have published this in the same post as last, but it’s just too important. By using Autodesk Inventor Professional 2008, the design group was able to estimate the weight of the new chassis. We updated the materials of all parts in the CAD and found the weight to be about 28 lbs.! A design mentor orignally ball-parked the weight to be about 40 lbs. So this is good news indeed.

We still need to update a few parts but we have a very good feeling about this year’s chassis.

Robot chassis construction began two days ago with the milling of frame corner plates. The plates hold 2×4 aluminum tubing together, which provides structure for the entire chassis. The design team and the build team are pleased that the mill worked perfectly off of CAM paths created by brand new students to the programs, the design group was able to quickly pick up the Keycreator CAM path program and use it effectively.

Milling for the frame rails is not far off and the complete CAD for the chassis is nearly finished. All the design team needs to do is double check clearances on adjuster blocks-which help create finite tweaks to chain tensioning and wheel positioning-and create the CAM paths for the frame rails.