CC3200 Network Processor Information Station


#54

I just switch to another AP (freedompop) and everything works fine from connecting to the AP and pinging external host, so we can eliminate possibility of project0 issue. Looks like this is related to the previous linksys router settings, i posted this issue on TI forum as well and some gentlemen is looking at it, hopefully I can get some hint from there. I will keep it posted here.


#55

Happy to hear its working :smile: . I’ll be interested to hear what the problem was with the LinkSys


#56

@Droidifi, any updates on posting some of that stuff on cc3200.io?


#57

Pra, just an update. issue resolved after some debugging with TI support. refer to this thread from TI forum http://e2e.ti.com/support/wireless_connectivity/f/968/t/351524.aspx?pi307168=2 on the details.


#58

Makes sense. That was great teamwork between you and TI to get that resolved


#59

Great to see some more work has been done on this. Has anyone been able to port over the spark and have programs up and running yet? i know it was the plan and if so i may have missed it.

With so many pins that can all be used for virtually anything i could see this chip be connected to separate base boards that all for more of the pins to be used, say a 20pin board or a 100pin board

no need to buy new processors and wireless that way :wink:


#60

Just FYI for anyone on a 3.12 kernel or later the modprobe arguments for VID/PID are no longer used.

From: https://bbs.archlinux.org/viewtopic.php?id=175499

From kernel 3.12, the ftdi-sio module no longer accepts the “vendor” and “product” parameters(commit link).
 I think it was assumed that this was a debugging feature that should
not be used by end users, and so the change was not at all publicised.The
 new way to add a new VID/PID pair for the ftdi-sio driver is to write
them both, one after the other, to /sys/bus/usb-serial/drivers/ftdi_sio/new_id

# modprobe ftdi-sio
# echo 0451 c32a > /sys/bus/usb-serial/drivers/ftdi_sio/new_id

#61

Good to know. I’ve added that to a Hackpad that I’ve been using to collect everything that I know about getting it to run: https://hackpad.com/Using-the-CC3200-Launchpad-Under-Linux-Rrol11xo7NQ

(Anyone is free to add to it if you see something missing or wrong.)


#62

We uploaded a patch file for the latest version of Nuttx (V7.5) to our server at http://www.droidifi.com/nuttx_droidifi_cc3200_2014-08-19.patch

Instructions -

  1. Use git to retrieve nuttx

git clone http://git.code.sf.net/p/nuttx/git nuttx-code

  1. Apply patch

cd nuttx-code/nuttx
patch -p2 < nuttx_droidifi_cc3200_2014-08-19.patch

  1. Use the nuttx buildroot script to build an arm compiler and tools (see the Nuttx documentation) or install via apt-get.

  2. Run -

cd nuttx-code/nuttx/tools;./configure.sh cc3200-launchpad/nsh

  1. Run -

cd nuttx-code/nuttx; make

The openocd config file is in nuttx-code/nuttx/configs/cc3200-launchpad/tools.

See Azdle’s Hackpad page or the Nuttx documentation for more info.

We will have a firmware flashing tool for Linux available shortly. Check back for updates.


#63

Just wondering how it is going with the CC3200? :slight_smile:


#64

Nuttx has been patched directly to support the CC3200. You no longer need step 2) above.


#65

I’m ready to buy the Spark3200! :wink:


#66

I’m interrested too. Any GitHub account?
Thanks,
NB: I’m porting the CCS/IAR IDE files to Eclipse one with OpenOCD debugging. As soon as it’s done I’ll push the code to github. Unfortunately not all the TI exemples have the makefiles


#67

guys, one quick question - is the NW subsystem, including the http server, entirely in hardware? I mean in full silicon, not in any form of internal sram, microcodes, ROMs etc.,


#68

For me, it doesn’t matter since all these tasks are performed in a separate hardware on the chip (another processor). But of course some high layers are done by firmware on this processor. Fortunately, this can permits firmware update from TI (their firmware are usually well buggy :slight_smile: )


#69

There are 2 ARM processors inside the CC3200, one handling low level wifi stuff, the other handling protocol work (including http & TLS in the case of the CC3200.) So it is upgradable.


#70

u mean 2 more processors dedicated to just wifi, apart from the main/application cortex-4 MCU?

As to the other reply to my question which states “it does’nt matter” – well, from a logical functional unit perspective, fine. But the details like whether the whole HTTP-server could have been evolved like fpga -> asic/asip/asoc…are needed to judge the speed and the possibility of tcp/ip stack executing concurrently.

again, just curious, nothing serious here.
Alex


#71

TI’s datasheet describes it well enough.


#72

CC3200:
Hi, I’ve transposed some of the TI cc3200sdk code to be used and debugged with Eclipse and OpenOCD. I’ve made also a small tuto in my website ( http://cc32xx-4-linux.io ) dedicated to the open-source development (build/debug) of the CC3200 chip on Eclipse with OpenOCD (on Linux). I’ve put the projects also in my Git repos: https://github.com/monnoliv/CC3200
Olivier


#73

https://www.kickstarter.com/projects/wipy/the-wipy-the-internet-of-things-taken-to-the-next - Spark 2.0 on Py :).
850uA in suspend node with connection alive - awesome.
Hmm, why this link off-top on this topic??? WiPy base on cc3200 chip.