Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by silverado

  1. I think this is the same thing I've been looking for - a way of connecting to the WiFi without having to code any connection information. Surely this is essential for a viable IoT application. I'd assumed there would be a WiFi.begin() that would connect using info previously configured by SmartConfig, but there doesn't seem to be one.
  2. Retuning of the library code is probably beyond me. I did reduce the number of sockets to 1 which buys back 2K, but is still short and that's without 99% of my code in place. I see that the compiler already has size opt switched on (-Os) and the linker is already throwing away unused code (--gc-sections) so I think I have to admit defeat and just go for a CC3200 instead. Simpler anyway. But thanks for the advice.
  3. My intention in developing the code for sending an email was to transfer it from F5529 to G2xxx (not that I yet know how to drive the CC3100 from a G2xxx!) so that ultimately I could build a circuit on stripboard with the G2xxx and a breakout CC3100. For this to be possible for home assembly the G2xxx has to be in a dip package, and I've discovered that the largest memory variant of G2xxx that comes in this format is the G2553, with a memory of just 16K. I find that this isn't enough to do even a bare minimum WiFi.begin() - the compiler reports the ram already overflows by 2294 bytes. So, am
  4. I'm really grateful for your SSL update! I've been trying to figure out how to send an email directly from my MSP430F5529/CC3100 boosterpack using a gmail SMTP server, and had come to the conclusion that it's not possible without an SSL connection. I was about to give up until I saw your post here. Having seen it, I updated my library (/msp430 instead of /cc3200) with the latest WiFiClient (cpp and h) from github and was able to write some code that successfully connected securely to the gmail server. Woohoo! Whatever may still be missing from your update doesn't appear to prevent my scen
  5. Thanks guys. I think I need to try some of this before asking more questions - learn to walk before trying to run so to speak. I will try with the TLC5940 as suggested and make use of the examples RobG points to.
  6. Sorry - I failed to say that each LED needs to be individually controlled. I know I could use a 2 port MSP430 variant to address them all directly but I still have the problem of a total of around 50ma to play with - so with say 10 LEDs lit that's only 5ma each. How bright would they be? If I understand correctly, PWM would make them appear brighter. I believe I could do PWM from the 430 itself but I'd have to code it myself - using shift registers is easier.
  7. Thanks - this does look exactly like what I asked for, but... This is supposed to be a low voltage, low power application, portable, battery driven - so the sort of power consumption I was suggesting is probably out of the question. As I said, I'm a electronics beginner, so perhaps someone can put me straight. Is a shift register solution (in other words PWM) the only low power option I have? How much brighness do I sacrifice by using this method? Can someone possibly point me at a diagram that shows me how to hook up, for example, 2 x 74HC164 to drive the LEDs? You also menti
  8. I want to drive up to 16 LEDs (not in a matrix form) from an MSP430. I've played with an G2211 on a breadboard/Launchpad combination to successfully do something with 6 LEDs, but I need to step up to 16 of them. My background is software development, but I'm new to microcontrollers and my knowledge of electronics is sadly no more than basic. My requirements are that although normally only a few LEDs will be on, occasionally most of them could be, and they need to be bright enough to seen in daylight, even sunlight. Also the brightness should be relative to the ambient light level so they'r
  • Create New...