paranormal activity of CMOS IC

Discussion in 'General Electronics Chat' started by vinodquilon, Jun 14, 2010.

  1. vinodquilon

    Thread Starter Member

    Dec 24, 2009
    234
    0
    Look at two attachments...

    In the circuit 4017 doesn't like its home place.

    When I changed its place from circuit IC socket to a breadboard and fed its input &
    supply directly from corresponding IC socket in PCB (by tapping upper socket pins through
    simple wire strips) ,I got output at pin 1.

    But when I return IC back to PCB socket from breadboard, I didn't get output at pin1!!!


    I am sure there would be no problems with IC and socket. Also all inputs to 4017 was correctly
    there.

    Does there exist any interferences from neighboring components as 4017 is a CMOS IC ?
     
  2. bertus

    Administrator

    Apr 5, 2008
    15,646
    2,345
    Hello,

    Do you have decoupling capacitors at EACH IC, directly at the power pins?
    A capacitor of 0.1 uF will do.

    Bertus
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    No, he does not.

    Like Bertus says, you need 0.1uF (100nF) decoupling capacitors (ceramic or metal poly) across the power/ground pins of each IC (except the optocoupler). Since you didn't plan for them, you can solder them directly across the power/ground pins on the bottom side of the board. Keep the cap leads as short as possible.

    In addition to a 0.1uF/100nF ceramic or metal poly bypass capacitor, the 555 timer specification requires another 1uF or larger cap across it's supply pins; otherwise it will cause instability on your whole board.

    When you are laying out your PCB, you should use wider traces and place the components more closely together. Narrow, long traces add lots of parasitic inductance. That in itself will cause problems with signals that have fast rise/fall times; you will get lots of high-frequency noise on the board.
     
    Last edited: Jun 14, 2010
    vinodquilon likes this.
  4. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,535
    In my applications I use a 0.1µF for the 555/556, just like all the rest. A 555 is a combination of digital and analog, but where it counts it is digital.

    Any device that switches between the power supply rails, be it CMOS, 555, or any other digital logic family needs those decoupling caps. Pulses on the power supply can be interpreted as pulse in a gate, much like the problems you're having.

    Most cases electrolytics won't work well for this application, so you want a smaller better caps with really low ESRs.
     
  5. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    An old fashioned 555 "shorts" the power supply with a shoot-through current of 400mA each time its output switches. That is why the datasheet for the LM555 recommends two supply bypass capacitors for it on their datasheet.
    Maybe you should use a Cmos 555 that does not cause this problem.
     
  6. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,535
    Which companies datasheet?

    Are you counting the bypass cap on pin 5?

    ********************************

    OK, Found it on the NE555 data sheet.

     
    Last edited: Jun 14, 2010
  7. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    On the datasheet for Intersil's ICM7555 Cmos 555 they describe the very high shoot-through current of an old fashioned ordinary 555.
     
  8. vinodquilon

    Thread Starter Member

    Dec 24, 2009
    234
    0
    Someone says it is due to 'cold solder joints'.
    Some others care about 'missing of bypassing of supplies'.
    Others are interested in 'floating inputs'.
    I think by clearing all the above three cases, my problem would be solved. Thanks...
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It's really easy to forget the necessity of using sufficient bypass capacitors. You can then spend days or more trying to troubleshoot all the problems you're finding that just don't seem to make sense.

    Long runs of digital signals will cause big problems; the signal traces will "ring" like the dickens when the levels change states. Using small resistors (say 50 Ohms) near the end of the trace can help a great deal to "snub" the ringing.
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Connectors are more likely to fail than components or solder joints - particularly when they have been used improperly (pushing a wire into it). I know you said that the socket is not a problem, but it would still be my biggest suspect.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Like Ron_H says, I've experienced broken sockets.

    Unless there is a really good reason to do so, avoid using sockets. They are usually more trouble than they produce benefits.

    One of the few exceptions are wire wrap sockets. However, it's so easy to make PCB's and use SMT IC's that those days are basically over.
     
Loading...