Why are my cmos chips always breaking ?

Discussion in 'The Projects Forum' started by simeonz11, Aug 25, 2009.

  1. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
  2. Mike2545

    Active Member

    Mar 26, 2009
    116
    3
    Check your power supply, try adding capacitance, just a thought.
     
  3. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
    I have added tons of capacitor on my power lines , ceramic .001 and .1 uf bypass capacitors on the lmc555 chip and iother cmos chips , it does the same thing with or without capacitors .

    It will ruin any cmos input in a microsecond .
     
  4. Mike2545

    Active Member

    Mar 26, 2009
    116
    3
    I was thinking more like 250-500uf across the rails...
     
  5. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
    I have more than 5000 UF across the rail , before I had massive oscillations and transcients from these cmos chips on my rail , now it never even moves when I see on on scope .

    Many different varieties , stuck almost all I had there . from tiny to huge , it seems to be working .

    Most of the problrms were caused by this lmc555 and my only cmos chip clock , the lmc555 is only connected to one clock input thx god .

    I tried with other types of cmos cheaps , same thing happens , always breaking ...
     
  6. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
    I did get it working before , for like 2 hours straight , I changed a few capacitors , a few resistors and now its just messed up ...

    Perhaps I momentarely short circuited the lmc555 ?
     
  7. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    Can you post the schematic?
    What is used for load at he moment?

    Greetings,
    Bertus
     
  8. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    There are a few possible cures for ringing / overshoot like that.

    1: Add a low value series resistor directly at the output of the driver IC (the 555 in your case). This could be anything from 10 Ohms to 100 Ohms or more, depending on the circuit. It acts in conjunction with the circuit capacitance as a low pass filter, slowing the edge transitions to reduce ringing.


    2: Add a terminator at the recieving end of the line - the chip the clock goes in to.
    That's trial and error, but start with something like 470 Ohm in series with 100pF, from clock in to 0V/Ground. If the clock line goes to more than one chip, put this farthest from the clock source.

    Edit: I'm thinking CMOS (4000) drive levels.. If the CMOS 555 is a high current drive device like the standard one, just use 470 Ohms from the clock input to +5V (or 1K if you are running it on 12V).


    3: Clamp it - fit a schottky diode from the chip input to 0V , so the clock pin cannot go negative of 0V. (Or actually not enough to do harm).

    With a lot of CMOS chips, if a pin is forced more than 0.7V outside the supply (above V+ or below 0V), the whole silicon die fires like a thyristor, shorting across power.

    With anything like this, you need to design and prototype to suit the frequencies involved - not the clock speed, but the logic edge speed;
    eg. if the high-to-low or low-to-high transition took 1uS (very slow), two transitions - a full cycle - would take 2uS, giving an effective frequency of 500KHz.

    With fast logic you more likely have edge speeds of 1nS = 2nS/Cycle or 500MHz+ effective frequency.

    You need to use RF design techniques for grounding, decoupling and impedance matching with high speed logic, even if the clock *rate* is only 1KHz.

    As an example, if you look at a PC motherboard or memory module, you will see masses of tiny surface mount resistor arrays. These are the series resistors as in (1) above. They actually allow things to work faster, as the logic levels stabilise sooner if ringing is prevented.
     
    Last edited: Aug 26, 2009
  9. jj_alukkas

    Well-Known Member

    Jan 8, 2009
    751
    5
    First of all please provide your circuit diagram.

    Second thing is, this is a CMOS chip, did you take precautions to avoid chip damage due to electrostatic discharge (ESD) from your hand? It will damage the chip even before you connect the wires by just touching it.

    Check your power supply. Add electrolytic capacitors of atleast 1000uF to filter the ripples. Use a supply voltage of minimum 3V when testing.

    If it is not quite a problem, try using the normal version which is very rugged.
     
  10. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
  11. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,175
    397
    Try putting a .1 micro F cap ,pin 5 to ground.
     
  12. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The supply voltage is not shown.
    The Cmos 555 does not have enough output current to drive a feedback resistance that is less than about 10k ohms but yours is less than 1k ohms.
    An ordinary 555 will work in that circuit if its supply voltage is at least 4.5V and the pot is more than 100 ohms.

    The long connecting wires on a breadboard are inductive and cause ringing.
     
  13. simeonz11

    Thread Starter Active Member

    Apr 29, 2008
    98
    0
    Gwaaaa sometimes I feel like the devil is attacking me , I HATE CMOS !!!!!!!!!!

    One of my dividers just started heating up for no reason.

    I tried a regular 555 , brand new ... and it was doing the same .

    I'm using a regular cd4047 and there is no negative whatsoever .

    This 555 has a problem , both of them have problems , it does the same with a single resistor no pot and tiny wires .

    I need a better oscillator , does nyone know of a chip that can go to 3 mhz and always 50/50 .

    .
     
  14. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    What frequency range do you need?
    On what voltage are your circuits working?
    Make an oscillator on the double frequency and divide it by 2.
    The result will always me 50/50 duty cycle.

    Greetings,
    Bertus
     
  15. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Most ordinary Cmos counters do not go as high as a 3MHz clock when the supply is as low as 5V.
     
  16. scaldara

    New Member

    Aug 27, 2009
    1
    0
    This circuit is not the typical way to use a 555. My speculation is that having the charging capacitor connected to the output presents a high capacitive load that results in significant undershoot/ringing, potentially damaging the downstream device (or possibly causing it to not function correctly).

    Here is a circuit I found that can be used to get a 50% duty cycle with a more conventional 555 setup.
     
    • 555.jpg
      555.jpg
      File size:
      120.5 KB
      Views:
      22
Loading...