4 Jailbreak

I need to follow up on the jailbreak post I had last year.  I got my hands on a few new iPhone 4’s now that we upgraded to iPhone 5.  These have now been jailbroken on iOS 6.2 and I’m continuing the configuration based on the http://www.ohscope.com/2009/02/24/serial-port-on-iphone-with-minicom/ instructions.

> ssh -l root
root@'s password: 
> cd ../../usr
> mkdir etc
> minicom -s
# follow instructions from ohscope on setting up the serial port
> cd /usr/etc
> ls

Good to go!

New Machine

So I decided to treat myself for my birthday and upgraded my MacBook Pro to the new 15″ retina.  Very fun.  Migrating over has been mostly trouble free but there were a couple of issues.  I prioritized getting the Teensy and Arduino build environments up and running.  Nothing strenuous but there was a small code change necessary.  Download and install:

  • https://github.com/adafruit/Adafruit-GFX-Library [Graphics Library]
  • https://github.com/adafruit/Adafruit_SSD1306 [OLED code]
    • Adafruit_SSD1306.cpp:425:26: error: 'TWBR' was not declared in this scope
    • Comment out references to TWBR
  • https://github.com/adafruit/Adafruit-WS2801-Library [RGB LED strand]
  • http://www.pjrc.com/teensy/td_download.html – This includes the XBee library and other libraries modified for used by the ARM platform

And that was it!  The makefiles needed a little modification to refer to the Arduino / Teensyduino installation but that was it.  Compiling 10 times faster!



Tonight was a good night. XBee / Teensy configuration problem solved.  Multiple coordinator problem…kind of figured out.  Code convergence for vest / hat / tower solved.

I’m looking at a vest and a tower running in front of me and changing the patterns through the iPhone.  Pretty cool.  Here is a picture of the Tower breadboard.  Bright red LED is the XBee power and you can see the LED’s dangling down at the bottom.  The Teensy is the dark board on the upper right, you can see the white micro USB plugged into it.


Teensy has been working great.  I’ve had to work around the makefile build / upload issue but I think we’re through that.  PJRC says that they will have a 3.0 command line interface available soon.  I’ll wait for now.  The upload GUI tool is sufficient for now (rebuild and press the button).  The issue with the XBee configuration code was that I was getting a MODEM_STATUS_RESPONSE from the network association.  I had to add an extra getResponse() call to get it off the queue.  My thought is that it’s due to the extra hardware serial on the Teensy and that it’s listening to everything from the radio on power up which is not what the Arduino is able to do.

As for the XBee extra coordinator, I’m having some trouble.  I have a working coordinator.  I want a couple of more for the other test beds.  The problem is that the 16bit PAN ID is being set differently as is the 802. channel.  I can reset things through AT commands but it’s not ideal and I really like the idea of not having to configure so much manually.  The PANID was very convenient as a network identifier and it worked flawlessly until now.  I’ll have to research this bit a bit more.


I’ve been having discussions with Randy about how to do better at synchronizing the computers to have better pattern coordination.  It’s a difficult problem because there are messaging latencies in the XBee network that are not stable themselves.  We came up with a couple of mechanisms to sync clocks but they all involve walking around to the members and performing either a tethered sync or a rapid fire messaging challenge.  I think it’s going to be OK but we’ll need to write up the code.

One thing that I needed to do in order to help keep things in sync is to get everyone working at the same frame rate.  This was easy enough to do in the Arduino code but I needed to read up on the Teensy ARM timer interrupt setups.

Teensy and XBee

As part of the mock up process I needed to get the XBee hooked up to the Teensy.  The latest release of the Arduino IDE from PJRC has an updated XBee library that I needed to use.  I keep forgetting to update the firmware of the XBee (I lent Elan a radio and configured it to his needs, I needed to reconfigure it to mine).  The big issue was API=2 and the Router API firmware.  It’s working great now except that my XBee configuration code is not working.  The AT commands aren’t able to be processed correctly so I can’t determine the BAUD rate or the PANID for the radio programmatically.  That’s the next issue to address.  The good news is that I’m able to control the tower LEDs from the iPhone through the Teensy!

Base designs

I spent an hour today looking at designs for bases.  I have the 4 post base physically mocked up and thought that a 3 post base would be interesting too.  I’m concerned with the centering ability for the 3 post but that will have to be mocked up to see.


seaGrass base version 1

Started mocking up the first version of the base today (very, very wet day in Oakland).  I reused some of the wood from our camp from 2 years ago (still has the bubble-bubble stickers on it).  We have some spring loaded restraints on 4 corners that will hopefully give us some slow period swaying.  It will take some tuning but I think our first mock up is giving us some good results.  Without a high center of gravity we’re still able to get some interesting reaction from the mechanism.  Also, the system has a good self centering reaction.  It’s not only the main axis interaction but the orthogonal springs that add to the stability.

One issue is the stiffness of the structure.  That will need to be dealt with using better design but I think we’re starting off on a good base.

A couple of pics.  Elan helped a lot with getting things organized and up in the air.

photo (1) photo

seaGrass mock-up video

Google is really incredible.

I decided to mock up one of the towers in SketchUp this morning to get some spacing I needed for the spring supports.  Easy enough with a little reading.

But I thought, why not use the model to see what it would look like in Nevada?  I geolocated the model to the Black Rock desert to get the right ground texture.  Then I duplicated the towers to get a sample arrangement.  Then I found a model of burningman someone built a few years ago.  Then I animated a fly through.  Then I published it to YouTube.

In total about an hour’s worth of work.  All with free software and very little training.  Geez.

So it’s a starting place for the fundraising video…