I haven't fully implemented Josh's feedback yet, but one other thing to consider, the shield runs HOT, i.e., it uses a lot of power. I haven't even begun looking at power management options or sleep modes yet, but it ran down 4-AA cells in a few days. Not sure if that is good or bad, but I'd like to look at strategies to lower its energy usage once I figure out the stability issue (and assuming I don't ragequit). I suspect this is why there are literally over a dozen different types of phy links in the IoT space...
On Wed, Nov 26, 2014 at 1:38 PM, seth wolf <[email protected]> wrote: > Thanks for the notes. I'm adding a wired Ethernet shield this afternoon to > my project in preparation of swapping it out for the CC3000 once all my > connectivity works. Is the CC3000 so cantankerous that I should stick with > the wired Ethernet. I was really planning on making this device more > portable and not needing a wired connection. > > On Tue, Nov 25, 2014 at 10:19 AM, Joshua Noble <[email protected]> > wrote: > >> >> >wouldn't a 2n2222 be insufficient to switch power? >> >> Ah, in retrospect that schematic is not correct. I was using a 2n2222 >> and it very well might not have been the best choice but it did work for >> me. Anyhow, let us know how the reflash goes, I've still got my CC3000s >> sitting around gathering dust so if there's a way to get them more reliable >> I'd love to know. >> >> On Tue, Nov 25, 2014 at 9:42 AM, Peter Torelli <[email protected] >> > wrote: >> >>> Joshua, >>> >>> Thanks for the great summary, extremely similar to my path. The ARP >>> cache issue from the spark.io forum was fascinating. I was headed down >>> the path you took with a hardware reset, although I don't follow the >>> schematic (e.g., the npn Vce is ground to ground!). Also, wouldn't a 2n2222 >>> be insufficient to switch power, wouldn't you use a NFET power transistor >>> like an IRF530 and drive the Vin to the breakout regulator? Bah, doesn't >>> matter, I'll try the firmware flash and then order a wifly. >>> >>> Thanks! >>> P >>> >>> >>> On Tue, Nov 25, 2014 at 5:02 AM, Joshua Noble <[email protected]> >>> wrote: >>> >>>> >>>> Ah CC3000, how I don't miss you. I had a deeply unfun time with exactly >>>> the same problems you had and wound up ultimately making a watchdog timer >>>> that completely rebooted the arduino. My story is here: >>>> http://thefactoryfactory.com/wordpress/?p=1140 and it ends with me >>>> basically saying "don't use this thing". I actually had better luck with >>>> the WiFly for long term connectivity *however* reliable friends of mine >>>> have told me that the device firmware update released by TI fixes the >>>> problems I ran into. Instructions on updgrading the firmware for the >>>> internal 8051 are here >>>> https://learn.adafruit.com/adafruit-cc3000-wifi/firmware-upgrades >>>> >>>> On Mon, Nov 24, 2014 at 10:03 PM, Robby <[email protected]> wrote: >>>> >>>>> You mentioned its an Uno shield, are you using an Uno? I remember >>>>> deciding it wasn't worth it on a Leonardo. (kind of worked on its own, if >>>>> I >>>>> remember, I think it was breaking all the other sensors) >>>>> >>>>> How's your sketch size? I remember their library being quite large, >>>>> only enough space left for one additional sensor library. Might be worth >>>>> some time trying to shrink while you're in there. Wouldn't mind a copy if >>>>> you end up working on it! >>>>> >>>>> Robby >>>>> >>>>> On Mon Nov 24 2014 at 11:41:10 PM Peter Torelli < >>>>> [email protected]> wrote: >>>>> >>>>>> I've been playing with a CC3000 (Adafruit shield) for a few weeks >>>>>> now, and my web client code is naively based on the sample code from >>>>>> their >>>>>> github, with a few tweaks based on discussions I've read on the adafruit >>>>>> blog and lots and lots of debug messages. >>>>>> >>>>>> I've seen some weird behavior with the device loosing the internet >>>>>> connection, or failing to connect to the server where it just hangs and >>>>>> needs a reboot (.connect() seems to be notorious for this), or having >>>>>> innocuous lib calls hang so that my timeouts don't fire. I haven't added >>>>>> any wacky hardware, just the UNO shield as sold by adafruit. >>>>>> >>>>>> I'm [a] assuming the hardware is fine (lolz), and [b] I'm preparing >>>>>> to write some kind of CTC based interrupt monitor and state-machine to >>>>>> keep >>>>>> tabs if the network device is stuck somewhere for an inordinate amount of >>>>>> time, and reset it accordingly (disconnect from AP or reconnect to >>>>>> server, >>>>>> or full on reset, etc.). >>>>>> >>>>>> Has anyone written a more robust (perhaps self-aware) network lib for >>>>>> the CC3000? Googling turns up the adafruit lib and lots of discussion >>>>>> about >>>>>> similar errors with non-answers like "put a delay(1000) after every class >>>>>> method call...". Feh. >>>>>> >>>>>> If not, I'll have whack at it and solicit input... >>>>>> >>>>>> Thanks for listenin', y'all. >>>>>> >>>>>> Peter >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> dorkbotpdx-blabber mailing list >>>>>> [email protected] >>>>>> http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >>>>> >>>>> >>>>> _______________________________________________ >>>>> dorkbotpdx-blabber mailing list >>>>> [email protected] >>>>> http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >>>>> >>>> >>>> >>>> >>>> -- >>>> joshua noble >>>> http://thefactoryfactory.com >>>> >>>> _______________________________________________ >>>> dorkbotpdx-blabber mailing list >>>> [email protected] >>>> http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >>>> >>> >>> >>> _______________________________________________ >>> dorkbotpdx-blabber mailing list >>> [email protected] >>> http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >>> >> >> >> >> -- >> joshua noble >> http://thefactoryfactory.com >> >> _______________________________________________ >> dorkbotpdx-blabber mailing list >> [email protected] >> http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >> > > > _______________________________________________ > dorkbotpdx-blabber mailing list > [email protected] > http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber >
_______________________________________________ dorkbotpdx-blabber mailing list [email protected] http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber
