How does a 555 timer work?

Discussion in 'General Electronics Chat' started by RichardO, Nov 8, 2014.

  1. RichardO

    Thread Starter Well-Known Member

    May 4, 2013
    1,231
    382
    I am more showing than asking...

    I got to wondering how fast a 555 timer could be after seeing information on the MIC1555.

    So, entered the schematic of the NE555 from application note AN170 in the 1988 Signetics Linear Data Manual Volume 2, Industrial. I tried to format the schematic the same as the original; including using the same device reference designators. All of the resistor values are taken from the schematic.

    Changes I made:
    I used high speed transistors: the 2N2369 for the NPN and the 2N5771 for the PNP.
    I substituted 1N4148 diodes for the diode-connected transistors.
    I added the timing resistors and capacitor to the 555 schematic so that I could simulate an operating circuit.

    With just those changes, the circuits works!

    BJT_555_schematic.png BJT_555_Waverforms.png
     
  2. WBahn

    Moderator

    Mar 31, 2012
    17,718
    4,788
    Okay....

    Your circuit is running at about 14 kHz. How is that addressing your query into just how fast a 555 timer could be? I guarantee you they can run a LOT faster than that!
     
  3. MrChips

    Moderator

    Oct 2, 2009
    12,429
    3,360
    drbenne and Bernard like this.
  4. RichardO

    Thread Starter Well-Known Member

    May 4, 2013
    1,231
    382
    You got me. I wanted to show what the waveforms looked like normally so I purposely chose a low frequency.

    With a 33 pF timing cap, the simulation gives 750 KHz. With a 33 pF cap and 2 each 1K ohm timing resistors the frequency is 3 MHz. So far I have refrained from modifying the resistor values or the circuit topology. Just reducing the resistor values should give much higher frequencies.
     
  5. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    I have long wondered if there was a curret limit on pin 7. In tony van Roon's 555 blog, believe I saw " pin 7 current limited by design". Back in the early 1970's thought there was a data sheet listing pin 7 as 200 mA max? Some day I'll sacrifice a 555 with Vcc on pin 7 to see if it smokes.
     
  6. tindel

    Active Member

    Sep 16, 2012
    568
    193
    Rich - have you built it? Seems that you made your own discrete 555. There may be a 555 or variant out there with much faster transistors - I haven't looked - that can obtain these frequencies. Of course they probably just call them converter chips! :p
     
  7. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Although I never bothered to verify it - I was told that the practical upper limit for a bipolar 555 is about 200kHz.

    Many years ago I bought a PAL standard TV colour bar generator in kit form, both timebase generators used 555 timers, the 15.625kHz line timebase proved waaaaay too unstable for setting up TVs. In the end I replaced both 555 chips with a board of TTL counters dividing down from a 10MHz crystal.

    Allegedly, the CMOS 555 can do about 2MHz - I bought a packet of them, but never got around to playing with them.
     
  8. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    I went searching in vain for an answer to that. The consensus? It depends.

    If you do the experiment, the result may only be good for that manufacturer's version. For the cost of the sacrificed 555 I'd want to know the current that caused it to fail, but also the voltage. It's likely the heat dissipation that is the limit, not the current per se.

    The rule-of-thumb I've been going with is the 200mA value. So I never aim for more than ~100mA continuous load, and I'm usually in the 5-12V range. So far so good.
     
    Bernard likes this.
  9. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    In the standard astable mode, your timing resistor is directly above the timing capacitor - above that, and from pin 7 to Vcc you have a current limiting resistor to protect the discharge transistor.

    In a feature article published by ETI magazine on 555 applications - it was suggested this resistor should be at least 1k, but I don't remember whether it stated any particular value of Vcc that applied to.
     
  10. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    Will stick with my rule, any C over 10 uF will limit pin 7 current to 200 mA, & with wayneh, 100 mA on continuous discharge.
     
  11. RichardO

    Thread Starter Well-Known Member

    May 4, 2013
    1,231
    382
    I have not built it. I am a masochist but not to the tune of 25 transistors... :D
     
  12. RichardO

    Thread Starter Well-Known Member

    May 4, 2013
    1,231
    382
    How embarrassing. I forgot to put the bypass capacitor on the Control Voltage pin. Even in simulation it matters! :(

    Now the frequency with 10 K's and 33 pF is more like 770 KHz. The frequency with 1 K's and 33 PF is still 3 MHz.

    With 33 pF and 10 K's timing values, delays are 117ns at the positive peak and 130ns at the negative peak. With 33 pF and 1 K's delays are: 86ns at positive peak, 48ns at negative peak.

    Not sure why the difference in delays when the timing resistors are changed. Maybe due to the finite voltage gain of the comparators???
     
  13. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    You'd be surprised how many people just don't bother with that.

    Once or twice I've had projects that didn't work properly without that capacitor.
     
  14. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    An awful lot of 555 projects seem to get away with leaving the control pin open. Some fail. I wonder what determines who wins and who loses.
     
  15. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    That capacitor only decouples the top 1/3 of the internal divider for the window comparator.

    Seems reasonable to assume it might get a bit more tetchy as the frequency goes up.
     
  16. RichardO

    Thread Starter Well-Known Member

    May 4, 2013
    1,231
    382
    I the simulation, I changed the "discharge" timing resistor to a very small value and measured the peak discharge current in the Discharge pin. I got a peak of about 31 mA for about 50 nS and about 26 mA for another 375 nS.

    Interesting. I expected a higher current with the 2N2369 used in the simulation. I would expect that the transistor in the real 555 to have lower drive current.
     
  17. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    Back on pin 7 discharge current: Connected 6 & 7 together with 37000 uF to ground, Vcc 12 V ; discharge took 3 sec, nearily linear except for last sec tapered off. NE 555 mfg ??
    Longest time was about 6 hours, C slightly leaky, 3 M +-, with 4.7k & C = 3 min.
     
    Last edited: Nov 11, 2014
    wayneh likes this.
  18. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    Assuming discharge required about 3 x RC, then the "R" is roughly 27Ω. At 12V, this would be a starting current of 440mA and it would take over one second to fall below 200mA. Risky for the 555? I have no idea.
     
  19. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Some designers just ignore pin 7 and take the top of the timing resistor to pin 3 - the two are in the same phase, so either will work. Driving the timing network from pin 3 of course saves the current limiting resistor required by the shunt switch at pin 7.

    Driving from pin 3 produces less mark space ratio error, so you get closer to 50/50. There's nothing to stop you having separate resistors for charging and discharge with a steering diode for each resistor.
     
Loading...