Connect with us

Digital time zone clock/temprature sensor project

Discussion in 'Sensors and Actuators' started by Iain, Aug 1, 2012.

Scroll to continue with content
  1. Iain

    Iain

    10
    0
    Jun 23, 2012
    Hello there everyone,

    I have an ambitious project and would like some guidance.
    my experience is limited to the 555 timer lols 10 years ago. soo I believe its best to be honest.

    Soo here we go- I must be certified nuts that I am sure :p
    Functions:
    1: So I would like to make a 10 time zone clock with 4 digit 7 segment + 6 DP display with RGB LEDs.

    2: an addition of a 10 digit name underneath each as well as a alphanumeric date with
    Julian date.

    3: Finally There is a display for 2x humidity and temperature sensors. (another 28 segments :eek: )

    4: USB and WIFI for programming and control.
    5: outputs for relays to a motor controller.

    The simplest solution I believe is a PIC controller, but there isn't one that will do all this so...its easier to break it down to the individual clocks, and have a master PIC I believe.
    In this case, a maxim will do the job I believe of driving the LEDs for each clock, and name.

    This results in 11 outputs for LEDs 4 inputs from sensors and one feedback loop from a mains sensor for manual override- is there a way I can reduce the led inputs on a SPI or other bus or even how to go about this on block diagrams?

    I am seriously over complicating this I guess.

    the xtals are also confusing me as to a refresh rate for displays, and one for the correct timing? the devices are designed to be standalone after being set by an international clock though usb.

    Is an fpga approach better?

    thanks

    Iain
     
  2. GreenGiant

    GreenGiant

    830
    6
    Feb 9, 2012
    What you are trying to do is EXTREMELY complicated, most off the shelf microcontrollers wont be able to handle this amount of stuff simultaneously.

    I would say before jumping into something this big, get a smaller controller, build yourself the clock that you want, get it working properly (just the clock not the setting of it from a computer or another clock or anything)

    Next move on to making the name work correctly (Im guessing LCD? or will you use alphanumerical displays?) and any effects you want to use with it

    Then move on to temp/humidity read & display

    Finally (the hard part) get the wifi connections to work for timing (outputs for relays is easy)

    Each of these should be a separate project, so you can focus on one thing then once you have all of ironed out you can move on to combining everything together in one project.

    I am not too familiar with the PIC hardware/software but from experience with Arduino there isnt a single chip/board that will do all that, and FPGA while it may be able to do that you are looking at THOUSANDS of lines of code

    I know this isn't the answer you want but take it from experience, you want to start small and work in parts, and get familiar with it all before jumping into something that complex
     
  3. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,361
    2,756
    Jan 21, 2010
    There are pic chips that will do this while standing on their heads. Look at some of the Pic24, dsPic and 32 bit PIC chips. The word "microcontroller" doesn't mean what it used to.
     
  4. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    Yeah, the new stuff is pretty much a full self contained computer now...

    I also agree this is a big task, especially for a newbie... There is lots to do here, if it was me I would personally go to a small form factor single board computer with video out and use an LCD display and skip all the LEDs, it makes it a whole lot easier...
     
  5. BobK

    BobK

    7,680
    1,685
    Jan 5, 2010
    Except for the number of outputs. One PIC + port expanders or shift registers, yes.

    Bob
     
  6. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    Yeah you would gobble up the I/O pins real quick trying to drive all the LEDs, but Charlieplexing might get you there with a high pin count chip (FYI I didn't actually do the math)... But, I'm sure you would suffer some flicker and refresh rate issues... Best to use some dedicated LED driver chips, you can easily get LED driver chips that will run 100s of LEDs off a couple of serial lines without an issue...

    I still say an LCD display is the best option vs 100s upon 100s of LEDs, but hey they make full color LED billboards now so it's all possible ;)
     
  7. Iain

    Iain

    10
    0
    Jun 23, 2012
    thankyou for helps, 7 segment madness

    Hello all :)

    I have decided to sort of turn this into a journal of the project and maybe someone will gain info from my mistakes :)


    So, what I have learnt,
    1: There is a lot more to segments than the 555 timer that I used at school :O
    2: I am sure the ardiuno has not enough ports to test all this..
    3: There is a bewildering array of chips that can drive LEDs, some simple, like a 7449, to a PIC controller that seems to do everything! and unbelievable complicated.
    4: I am still confused over what is going on after looking at all these diagrams..so I would like to ask for some help in the block diagram first of all.


    Segment > BCD > COUNTER/TIMER > XTAL where does shift register fit in?

    Segments need to be driven by a Binary Counter Decoder, the numbers are generated by flipflop/counter circuit I believe? - which im still not sure entirely how they work, which is regulated by a xtal.

    so Q.

    1: Is an additional xtal required?

    2: Now I have trouble working out the correct frequency. - used to believe that I would need 60 hz/s for the cycle, then /10 for hours /6 for mins and then /10 /6 for seconds,
    However this is not how it seems to be done..sighs.

    3: What do shift registers do, are they like ram? and how do they operate,

    4: There is also a design out there with transistors for each segment that boosts the power somehow allowing them to be brighter, is this necessarily if they can be directly driven?

    Notes:
    Segments can be wired up differently depending on the task or can even be synchronous over a SPI link using 2 cables.. impressive, if you know how

    thanks

    Iain
     
  8. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    This statement is 180 deg out of phase. uCs simplify circuitry, not complicate them.
     
  9. Iain

    Iain

    10
    0
    Jun 23, 2012
    pic update,

    Modern PICs are very complicated and comprehensive inside, for the uninitiated, programming one can be very daunting. which is why I was hoping to stay away from programming, hard with something this complex, so I am focusing on making a 4 digit clock so far as the first mini project
     
  10. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    If you think PICs are daunting wait until you see the board layout and the number of connections you are going to have to make trying to make even a single clock display do what you want with logic chips... Let alone a 10 time zone clock...

    [​IMG][​IMG]

    Modern PICs with a good compiler is very simplistic vs some other controllers and programming environments... Yes, there is a learning curve, there always is...

    Near equivalent PIC version of the above...

    [​IMG]
     
    Last edited: Aug 3, 2012
  11. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    You don't have to know what's inside. You only have to know how to deal with inputs and outputs. That's why they put it all in a chip in the first place.:D
     
  12. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    And that is a GOOD thing, I fully give my hats off to the guys that layout dies for these micros... I was following a thread on the Parallax forums a while back where their layout guy would post pictures of his work in progress for the die on their new Propeller chip... Just mind boggling, crazy, stupid and beyond involved...
     
  13. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    CC, got a link to it? I'm sure many of us would find it fascinating.

    Chris
     
  14. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
  15. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
  16. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    It's a cool read/listen/watch thread, just seeing a stat on the first page like "Almost 7 thousand transistors per I/O with 92 IO's." that's about 644,000 transistors to route on the die just for the I/O lines!!! Micro have come a long way...
     
  17. Iain

    Iain

    10
    0
    Jun 23, 2012
    7 segment display basics

    Perhaps I should be asking peoples guidance on the actual stages of driving a led segment from the battery to the display circuitry.

    I was trying to understand the basics before going into pic controllers

    Linked images are indeed impressive, I don't understand how people wont go mad with all those transistors, or loose track of what they are doing, then again, GPUS have billions of transistors, and intel cpus have millions..

    but even the most powerful pic is not enough channels

    thanks

    Iain
     
    Last edited: Aug 4, 2012
  18. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    Using Charlieplexing there probably is enough pins, but the refresh rate would almost certainly get out of hand...

    The good news is that there are serial controlled 7 segment drivers that will run 100s of LEDs with only a few i/o pins... You can chain together multiples and thus you would only need a handful of pins to control dozens upon dozens of displays... The Maxim MAX7219/MAX7221 is one such chip, it's a little costly and although I sampled a few recently I have not acutaly used them, but it will do 64 segments (eight 7 segment displays)... The MAX6955 will do 128 segments (sixteen 7 segment displays)... There are others out there as well from other manufactures that will do more segments, but I don't have the chips numbers handy...

    But, again I asked why the need to stick to 7 segment displays?

    Also you can can possibly cheat on this by hard wiring the location names and thus not needing any control over them...
     
  19. Iain

    Iain

    10
    0
    Jun 23, 2012
    I have decided to go with an arduino uno, and experiment with led displays,

    I will keep you posted as to the results

    I am certainly realizing tri leds are very impractical and maybe even lcds screen may work better, even though there would be a lot of them.

    For some reason I like them, also really like the nixie tubes, but of course they only come in one colour :)

    whist charlie plexing may work, I don't believe the ardunio has the pins even with expansion boards hence lcd screens may be a necessity in the long run.

    Generally I wished to learn how to make them as i had never made one at school

    thanks

    Iain
     
  20. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    The Arduino is a developer board, and IMO should be used as such... If you are actually designing a specific product you should tailor you micro and circuit choice to the design, not force the design to 'work' with a limited development platform...
     
    Last edited: Aug 4, 2012
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-