Bricked 556?

Discussion in 'The Projects Forum' started by Sloucher, Dec 19, 2013.

  1. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    Hi Guys

    Help! I hope someone can come up with an explanation for this:

    I have a 556 circuit, which I've made probably over a dozen examples of. Its all assembled on a pcb (etched professionally). The pcb has been checked over and over and checks out just fine, but......

    After being used for a short time, 50% of the assemblies stopped working, in that one of the timers (usually, but not always the same, slower timer) "locks up" and the led controlled by it stays permanently on. Switching off the supply to the circuit and reapplying it makes no difference. Once it stops working, it stays not working.

    My first suspicion was that one of the electrolytics (all surface mount components btw) may not be soldered on properly as the pads are quite small, but no, they've all been fine. The circuits all seem to be fine initially but fail later.

    I seem to remember reading somewhere (no idea where now) that this problem had been documented and a work round included a diode and a resistor somewhere, but I couldn't find any references.

    Circuit diagram is attached.

    Thanks in advance :)
     
    • 556.png
      556.png
      File size:
      10.9 KB
      Views:
      41
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    759
    Try this.

    Connect a 100uf 16V cap across the 12V and a 100nf ceramic across the 556 positive and negative. solder the 100n across the IC pins. It should be as close as to the IC
     
    Sloucher likes this.
  3. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    Ah, now thats a thought, thanks R!f@@. A spikey power supply could cause problems indeed. I'll try what you suggest, but..... I've also had this problem when using a 9v battery supply though.

    Also I've noticed that when one of these circuits locks up like this, that it seems to have suffered permanent damage in that no matter what supply is re-connected, the symptoms remain:(.
     
  4. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
  5. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    759
    Base resistor is a bit small for a Tr. Try 1KΩ to 4.7K

    Even with batteries spikes are there since the 556 switching causes it
     
  6. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    I've just recalculated Rb for a 2n2222, assuming a minimum Ic of around 40mA (just 2 leds) and using an hfe value of around 80 (extrapolating from the data sheet values), I get 12 x 80/0.04 (Rb = Vr * hfe/Ic) and arrive at a value of 24K.
    Thats around 100 times higher than the resistor I'm using (270R). Does that sound right, it seems awfully high to me?
     
  7. sheldons

    Active Member

    Oct 26, 2011
    616
    101
    that 270 ohm base resistor is a bit too low value wise..........what is your circuit driving......
     
  8. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    Thats what I thought, but my further calculations lead me to believe I should be using a resistor around 24K, and that, to my mind, seems far too high.

    The circuit drives two sets of leds of anything from two up to around 10 leds per set (I make these circuits for modellers who want navigation lights and single strobe lights on their models, so the number of leds is not always known). The transistors were used originally because the original design had to cater for leds that had to be low side switched.
     
  9. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,240
    Do you have resistors in series with your LEDs? If not the high current could blow the 2N2222. In which case you would want the high resistor to beta limit the LED current.
     
  10. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    759
    Try 4.7K. It should be fine
     
  11. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    Whenever I've tested these circuits, yes I have. Though I can't guarantee that the modellers whose boards "lock-up" have also done so, although most maintain that they are using pre-resistored leds. I also list the resistor values to use with different colour leds with the wiring instructions that I include with the boards. All the transistors on the faulty boards seem to be functioning ok as well. I did try replacing a couple but it made no difference.

    I wonder if they've simply reverse voltaged them? How tolerant is the 556 to abuse like that do you think?

    I'll make a board up with the extra capacitance across the supply and extra resistance on the transistor bases then switch it on and leave it to see what, if anything, happens.
     
  12. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,010
    3,784
    You, as the designer should put a diode at the power input to insure that no damage is done if the power is accidentally reversed by the user.

    Also, it would be a good idea to include a handful of resistors for various strings of LEDs and a header in your board for the user to add the resistor to complete the circuit (forcing them to comply.)

    Also, do you expect the user to add one string of series connected LEDs to each 2n2222 (eg 5 red LEDs in series for 10 volts) ?

    Or do you advertise that the user can connect more in series parallel arrangements? If so, do you tell them that each series string needs a resistor?
     
  13. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,006
    1,527
    Wouldn't the addition of a diode, in the supply line input, eliminate that possibility? When ever making something to sell you have to cover all bases.:)
     
  14. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,240
    Hate to mess up your board, but you could add a diode in series with the +12 (anode to the +12) and a 120 ohm or so in series with the output of the 2N2222.
    You will need the 100 Ufd. for sure in that case. The absence of the cap might make it erratic but shouldn't blow the IC.
     
  15. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    To be honest, reverse connection shouldn't happen as the connector is polarised, but I get your point about building in foolproofness.

    I intend using an ATTiny controlled board in future. I've used it for other things before and its proved to be fairly bulletproof (so far!). That does have reverse connection protection on it (diode and cap) uses a switched regulator (LM2705) and outputs via an ULN2308 so that should be fairly foolproof.

    I supplied the 556 circuits with comprehensive wiring instructions, the two outputs were only ever intended to be used for parallel strings and the instructions quite clearly show that. I don't supply the leds or resistors as different modellers have different needs/requirements so the instructions pretty much cover all they need.

    Anyway, thanks for all the advice guys, I now know the extra precautions to be borne in mind when using 555/556 circuits!
     
    Last edited: Dec 19, 2013
  16. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    As a general rule, when using a transistor as a saturated switch, as Audioguru or the Sergent would tell you, you treat the transistor as though its Hfe is only 10. So to switch 40mA, you need 4mA of base current, so Rb= ~ 10V/4mA = 2.5K
     
  17. sheldons

    Active Member

    Oct 26, 2011
    616
    101
    Heres a simple transistor stage and mosfet output stage which you can play with.....
     
    Last edited: Dec 19, 2013
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Use a "forced beta" value of 10 when calculating base resistors for transistors that you want to operate in saturation. Sometimes you can "get away" with higher values, but if you want to be certain of it working reliably, use 10. R!f@@ is gambling that a beta of 20 will work sufficiently well. He may be right, but you should go through the math yourself and be certain.

    The output of a bjt (transistorized) 555 or 556 timer will be about 1.2v-1.3v less than the supply voltage when under light to moderate load. Your Vbe will be around 0.6v-0.8v for light to moderate loads. So, if you're using a 12v regulated supply, you need to calculate 12v - (1.25v + 0.7v) = 10.05v to drop across the base resistor.
    You say your LEDs require 40mA - so, you will need 4mA base current.
    10.05v / 4mA = 2.5k Ohms.

    A table of standard resistance values is here:
    http://www.logwell.com/tech/components/resistor_values.html
    You'll find that 2.5k is not a standard E24 value, however 2.4k and 2.7k ARE standard values. In this case, you can likely use either.

    Don't forget to calculate the power dissipation of the base resistor, and double it to make certain you're using a high enough wattage rating.
    10.05v x 4mA x 2 = 80.4mW. You can safely use a 1/10 Watt or higher rated resistor.

    (edit) Silly me, I didn't realize there was a whole 'nother page to this thread - and MikeML correctly anticipated my beta-rant ;)

    I wouldn't be surprised if many of your failures are due to being operated in an automotive environment. The circuit just won't survive long there; transients will kill it. There are voltage spikes sometimes exceeding 60v during load dump events, like turning off headlamps. So, make certain to tell your customers that automotive use is not supported, and voids any warranty.
     
    Last edited: Dec 19, 2013
  19. Sloucher

    Thread Starter New Member

    Feb 7, 2012
    23
    0
    Again, thanks for all the info guys, it all seems very good advice, particularly about the transistor base resistor calculations, which is where I'm sure my particular problems lie.

    The circuits aren't likely to be used in an automotive environment as they are simply to flash leds in a static model (aircraft, spaceships - that sort of thing), however the power supply used by the modeller is out of my control and could well be noisy, so I take on board the supply suppression and supply reversal advice.

    I'm definitely going to build at least one more of these circuits, if only to try all the recommendations out :)
     
  20. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Now that I think about it some more - the lack of the bypass caps across the Vcc/GND pins, along with the inductance of the wiring, could very well be causing high voltage spikes across the 555 when it toggles states - which would kill it in fairly short order.

    Bypass capacitors aren't optional. This thread: http://forum.allaboutcircuits.com/showthread.php?t=45583 goes into it in some detail.
     
Loading...