24 hour clock display

Discussion in 'The Projects Forum' started by erich_7719, May 29, 2010.

  1. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    I am not a student. I took a basic electronics class in high school (about 14 years ago). I put down electronics about 8 years ago. Now I am starting to get back into it.

    I've been reading quite a few threads here and learning that I am not as much a total noob as I though I may have been. With that said.

    The topics of building a 24 clock struck my fancy to see if I could do it; and I did, it works great (in binary). Now I am ready to do the display, the display will consist of Hours, Minutes, and Seconds, constructed out of about 200 LEDs. For simplicity the attached schematic shows 7 segment displays.

    The question is...

    Is the way I have everything wired going to work (in theory anyways) to "Multiplex" the digits.

    I have read through a couple datasheets of multiplexers and can not seem to rap my head around it, so that is why I went the route I did.

    I left out the working clock portion since it works just fine needing no modifications.

    Thank you in advance for any input.

    Also, with the final clock everything below the red line in the attachment will be operating on 5Vdc with a battery back up and everything above will be either 10 or 12Vdc (haven't carver that in stone yet).
     
    Last edited: May 29, 2010
  2. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    OK, I have built the circuit and I realized that the colons were wired wrong. I fixed that problem and attached the updated schematic.

    I do still have one problem still All the digits show the same digit at the same time. Does any one have any ideas how to fix this problem?
     
  3. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    You do not have to multiplex the digits. In design terms it is simpler not to, but in parts count it takes more chips, so it is cheaper to multiplex.

    With multiplex you need one decoder chip, and a simple binary counter and gates to select which digit is chosen and displayed. This makes the logic a bit more complex, but overall it doesn't use as many parts.

    With no multiplex you need a decoder chip for each display, and very few additional gates. If you include seconds this is 6 decoder chips all wired pretty much the same way. I would go with the latter myself.
     
  4. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    Thank you Bill for your reply. I got it situated by adding pull down resistors between the diodes connected to the 4081's and the 4543.
     
  5. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    I have this clock built and I let it run for 18 hours. After the 18 hours the clock was about 20 seconds off (averaging .000308 seconds loss every second).

    When I use my multi-meter to measure the frequency I get 2Hz at pin 3 of U4 (4060), it also reads 1HZ at pin 1 of U3 (14027), but it does read 32.77Hz <--(the accuracy of my multi-meter) at pin 9 of U4 (4060).

    Is there a way without a O-scope to determine if it is the Oscillator or some other part of the circuit?
     
  6. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    Measuring 1 or 2 Hz on a frequency counter is usually not very accurate as the last digit has some inherit timing problems. If you go to pin 7 of the 4060 you should measure 8192 Hz. Adjust the trimmer in the oscillator circuit for the exact frequency. If the timing circuit of you counter is accurate then the clock will be right on.
    I have built a number of digital clocks with the 4060 as an oscillator and divider and adjusted the crystal ckt with my frequency counter, and usually get accuracy near one or two seconds off per month, as compared to WWV
     
  7. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    A 4060 also has a pretty good test point as part of the IC package. Tweak the oscillator to 32767 and you'll be very close to on frequency.
     
  8. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    k7elp60: Thank you for your reply. I will test this after I get home from work today.

    Bill: Thank you for your reply, although I can not test your suggestion for my DMM only has a 4 digit accuracy. If there is another way to adjust that accurate with what I have, I will be happy to do so.
     
    Last edited: Jun 8, 2010
  9. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    I have uploaded the complete schematic.

    k7elp60: At pin 7 I get 2048, and no matter how much I adjust the trimmer Cap I still get the same value. I looked at the data sheet and the 4040 at pin 7 should be 8192 the 4060 skips that step.

    For my seconds I changed the way I enable the PE. The original position of the diode in the red circle (D4) was where the blue dot is. Could that have caused the delay?
     
  10. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    You could also add a flip flop and tie it to the 16384Hz output. It will drop it to 8192 Hz, which your counter should be able to handle.
     
  11. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    I don't believe the diode(D4) position is affecting the clock. I think you DVM is giving you a false reading. If you have a watch or a clock with second hand you could compare the two for a period like 8 hours, readjust the trimmer and then compare again. This will take a lot longer time, or you could try pin 14 of the 4060 which should be 128 hz.
     
  12. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    Bill: I would have to use two flip flops connected to pin 9 of the 4060 to go that route either after work tomorrow or Thursday, since Pin 7 is the fasted speed on the 4060 at 2048 Hz.

    k7elp60: All the output pins are reading what they should be (before I tried adjusting the trimmer).

    Hows about the fact that this is built on a solder less board(the 4060 portion outlined in red ir soldered to a permanent board). I've read, in many post here, that there is a lot of capacitance in them. I doubt this is happening, but I have to ask. Could this be causing the time being off?
     
  13. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    [QUOTE).

    Hows about the fact that this is built on a solder less board(the 4060 portion outlined in red ir soldered to a permanent board). I've read, in many post here, that there is a lot of capacitance in them. I doubt this is happening, but I have to ask. Could this be causing the time being off?[/QUOTE]
    It is possible. Are you using a insulated screwdriver to adjust the trimmer?
    If the rotor of the variable capacitor is connected to ground an insulated screwdriver isn't necessary. I think your clock is close to the right frequency, but just needs a slight adjustment.
     
  14. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    I just checked my rotor by looking for V drop to ground and found the rotor is Positive.

    Is a insulated screwdriver still necessary (since I do not have 1) if I adjust the cap wait a second or 2 than test?
     
  15. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    With an uninsulated screwdriver just be carful not to touch the blade while adjusting. Or you could unsolder the capacitor and turn it around. The blade of the screwdriver may affect the capacitance of the circuit, so making an adjustment and waiting a short time should work.
     
  16. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    Thank you for all your help, just 1 more question (hopefully).

    The power supply for the LED's. Is there an easy war to incorperate PWM to controle the light intensity, while utalizing the crystal as the trigger clock, and is so could you point me in the right direction of the proper PWM circuit to use?
     
  17. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    I'm having some difficulty following the diagram but I think the displays are multiplexed and U6 controls the segments and U6 controls the digits. If I have it right then possibly connecting a variable duty cycle wave from a 555 timer to pin 1 of U6 could control the intensity of the displays. Let me know if I am right and we will go from there.
     
  18. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    Yes they are multiplexed. You were close, U7 selects the digit (in and out of U6) while U6 controls the segments. I didn't consider using pin 1 of U6 for the control of the led intensity, I was considering the individual segments power source (denoted in blue of the attached).
     
  19. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    If you put pin 1 of U6 to ground the segments would be turned off. So with a square wave on pin 1 it would affect the time the segements are on, and off. It has been my experience that because the eye can see blinking about 20 times a second, the intensity of LED can be controlled if the on time of the LED's is about 3ms and of about 10ms. This is with direct drive. By appling pulses to pin 1 of U6 does not require any power circuits that would be required by pulsing the applied voltage.
     
  20. erich_7719

    Thread Starter Active Member

    Oct 14, 2009
    92
    7
    k7elp60: Thank you for all of your assistance, and knowledge.
     
Loading...