Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Problem with my HA17 555 timer

  1. Feb 14, 2010 #1
    i recently bought my 555 timer and i simply can't get it to pulse. I flick on the switch and i get my buzzer to work that is connected to the output but i don't hear it turn off and on. I set my configuration as astable 6.5 Hz. i checked all my connections and everything seems right. However during my first run, i connected my timer circuit parallel to 1k that was instead of two 1k connected with source along with 30 ohms the circuit (that gave off 9v 8mA). And during my first 2 runs my timer was heating up really fast. After adding another 1k to the source (as the pic) i was able to get 5V and then it didn't heat up. I even checked my simulation with multisim and despite of that i don't get any pulse (the buzzer jus keeps going and no beep). For some reason on measuring the output with reference to ground my analog multimeter spikes upto more than 250mA, not sure if that is supposed to happen.

    I am thinking its either i burned my timer or i got the pinning wrong. Check both my image files to see i am doing anything wrong. Btw 9V with 30 ohms in series depicts my adaptor that gives off 9V with 300 mA

    Download file from link below.


    http://squadinvenire.uuuq.com/555_timer.rar [Broken]
     
    Last edited by a moderator: May 4, 2017
  2. jcsd
  3. Feb 14, 2010 #2

    vk6kro

    User Avatar
    Science Advisor

    You should measure the output voltage, not the current, on pin 3. The output should vary between zero volts and almost the supply voltage.
    Try doing this with the buzzer disconnected. You may be able to see the meter moving rapidly up and down or it may settle on about half the supply voltage.

    If it is doing this, the 555 is still working.
    You may need an oscilloscope to see the true output waveform.

    Your connection at the power source seems odd. You could put in a LM7805 regulator if you wanted to drop the outout to 5 volts. You don't need to put resistors in parallel with the supply.
    If you do put in series resistors to drop the supply voltage, you may need to put a bypass capacitor at pin 8 of the 555 chip. About 1 uF should be OK.
     
  4. Feb 14, 2010 #3
    thx for ur fast reply. 1uF at Vcc didn't solve my problem. Assuming u checked my pinning, is 2 mA enuf to be applied at Vcc and and reset? ( i just measured it) And since i don't see see any voltage going up any down (the voltage is barely up by 1v) do i assume my timer is still functional and not burned?
     
  5. Feb 14, 2010 #4
    Also if my reset isn't plugged in with Vcc is it supposed to work? cuz my in multisim it stops working.
     
  6. Feb 14, 2010 #5
    And also like u said i my my connection at the power source seems odd. I think i shud simply reduce my amps by adding appropriate resistors but can u suggest how much amps do i need?
     
  7. Feb 15, 2010 #6

    vk6kro

    User Avatar
    Science Advisor

    Pin 4 and pin 8 on a 555 should both go to the + supply voltage.

    You don't have to reduce the current. Only the voltage matters because the 555 will only draw the current it needs, not everything the supply can manage.

    Do you have another capacitor you could put across the 10 uF? Another 50 uF would be good. then you could see the output varying, if it is. Or replace the 10 K with 100 K
    6.5 Hz may be too quick for the meter.

    If you have a CMOS 555 then 2 mA is probably OK. The chip should not draw more than 0.5 mA but the 1 K resistors and the timing resistors would take a bit.

    You should be able to just put your full 9 volts across this CMOS 555 without any series or parallel resistors. The maximum voltage for these devices is 15 volts.
    Do this without the buzzer in place, just for the test.

    If you still can't get it going, try the following circuit:
    http://dl.dropbox.com/u/4222062/CMOS%20555%20Osc.JPG [Broken]

    eg in this circuit, 10 uF and 11.7 K should give 6.5 Hz output.
     
    Last edited by a moderator: May 4, 2017
  8. Feb 15, 2010 #7
    I will defintely try it inthe morning btw my concern is again regarding the amps. i know that max rating for voltage is 15v but since ur sayin it needs 8mA and if i connect it directly to 9 v as u say i think it might burn up since my adapter provides 300mA with 9V. so i jus wanna confirm it again if i shud go ahead and plug right in or add resistors to bring it down to 8mA around. Also i have various capacitors and resistors so don't wrry about being reluctant suggesting me the exact kind of components. i have tried 1 - 80Hz on my multimeter so yea it will pick up 6.8Hz. Last concern, do u want me to use 11.7K like u said alone or with 1K since i need 2 resistors for timing?
     
  9. Feb 15, 2010 #8

    vk6kro

    User Avatar
    Science Advisor

    If you put a 1K resistor across a 12 volt car battery, it will only draw 12 mA even though the car battery could probably deliver 300 AMPS.

    The 300 mA on your power supply is just the maximum rating. It won't supply that unless you put 30 ohms across the power supply. If you put less than this, the power source will be overloaded.

    There is another meaning of this 300 mA which is used on really cheap power supplies. The output voltage ONLY applies at that load because they have such poor regulation. For this reason, you need to check the voltage out on no load. It might be higher than the voltage shown for small current loads.
    Also you need to know that this is a filtered power supply. It might be pulsing DC which would play havoc with a 555.

    No, the exact value of that 11 K resistor doesn't matter unless there is some reason for getting exactly that frequency.
     
  10. Feb 15, 2010 #9
    I gotta thank you for clearing that out for me. I used standard 9V battery and yup its working and lol i can even see 0.7 Hz in my multimeter. My question now is how come i don't see 200mA output like it says in datasheet? or like u pointed out with my power supply that 200mA is just the maximum rating. In any case how can i get 200mA output?

    Also by connecting my timer directly to 9V, doesn't there have to be any significant amps with it to make my timer work? And does 1mA to 8mA input doesn't need to be applied to Vcc or was it just for pins 7,6 and 2?


    You really got me concerned about my power source. I really think my power source cud be pulsing now which has to be the only reason y my 555 timer isn't pulsing. Is there anything i can do about it?
     
  11. Feb 15, 2010 #10

    vk6kro

    User Avatar
    Science Advisor

    My question now is how come i don't see 200mA output like it says in datasheet? or like u pointed out with my power supply that 200mA is just the maximum rating. In any case how can i get 200mA output?

    This is the same story as the power supply. The current depends on the load.
    The main output is the voltage and as you put smaller resistors across this voltage, more current goes through the resistor.

    To test if your power supply is pulsing, measure just the power supply voltage then put a large capacitor (10 uF ?) across it and see if the voltage goes up. If it does, the supply is pulsing.

    To fix it, get an electrolytic 1000 uF 40 V capacitor and put it across the output permanently.
    Measure the voltage before you use it, though or it may be too high for your 555.

    Also by connecting my timer directly to 9V, doesn't there have to be any significant amps with it to make my timer work? And does 1mA to 8mA input doesn't need to be applied to Vcc or was it just for pins 7,6 and 2?

    CMOS chips work on quite low currents, so a small current is normal.
    The 9 volt supply should only be applied to pins 4 and 8.

    yup its working
    the only reason y my 555 timer isn't pulsing


    is it working or not?
     
  12. Feb 16, 2010 #11
    Oh srry for not making myself clear, the timer is working right and it is pulsing to the "hertz of my desire" heh which happened as soon as i changed to my normal 9V battery.

    Btw wen u replied as below

    To test if your power supply is pulsing, measure just the power supply voltage then put a large capacitor (10 uF ?) across it and see if the voltage goes up. If it does, the supply is pulsing.

    To fix it, get an electrolytic 1000 uF 40 V capacitor and put it across the output permanently.
    Measure the voltage before you use it, though or it may be too high for your 555.


    i jus want to know the idea behind it. does the 10uF and 1000uF serve as a bypass capacitor? i was jus learning about high pass/low pass filter so does it follow that concept?

    For this question of mine
    Also by connecting my timer directly to 9V, doesn't there have to be any significant amps with it to make my timer work? And does 1mA to 8mA input doesn't need to be applied to Vcc or was it just for pins 7,6 and 2?

    CMOS chips work on quite low currents, so a small current is normal.
    The 9 volt supply should only be applied to pins 4 and 8.


    What i really meant was does 1mA and 8mA meant only for pins 7,6 and 2 instead of Vcc and Reset pins. Now i understand that my question didn't make sense. Before clearing my concept about my power supply, i thought voltage and current go along together to make various specifications. A stupid mistake for a second year student but all this because of that power supply i have been using since gr12.

    Anyway thx for ur help in gettin my timer to work. I have moved on to JK flip flops and i need ur help there
     
  13. Feb 16, 2010 #12
    For my Hef4027 Flip flop i am following this circuit

    http://squadinvenire.uuuq.com/flip_flop_circuit.JPG [Broken]

    i don't see where i am making a mistake this time and yes i am using my 9V battery this time. The isolated AC power source represents the 555 timer output and even that didn't work. But then again i wanna get the basic connection to work first i.e the circuit in the link.

    After my connection i somehow get 10 V (cud be the fault of multimeter) and neither current or voltage drops wen i flick the switch to make it on/off
     
    Last edited by a moderator: May 4, 2017
  14. Feb 16, 2010 #13
    Nvm i get output of 3V (ignore 10V) even wen i have it connected directly to 9V battery but now wat i notice is very weird, the needle of multimeter goes to 3 to 5V and back randomly even wen i am not touching anything. Also wen i hold down the switch for pins J,K and clock i the voltage again rises to 5V and goes back to 3V as i release it. I have my connected to the same board but not connected to the source (if that is responsible in anyway).
     
  15. Feb 16, 2010 #14
    Just realized the battery is down so it showing 5V instead of 9V
     
  16. Feb 16, 2010 #15
    another wierd part is that i am getting 5V from both true output and output compliment and yea the voltage is still dilly dalying
     
  17. Feb 16, 2010 #16

    vk6kro

    User Avatar
    Science Advisor

    If you tie J and K to the + rail (ie straight to 9 volts not via the switch), and clock the clock input, you should get output at half the clocking rate.

    The clock pulses should have the same amplitude as the power supply of the JK flipflop.

    You really need access to an oscilloscope to understand what is happening here.

    You might like to try using a 4020 binary divider. One of those would divide your 555 output frequency by up to 16384. It does this in binary steps, although there is no divide by 4 or 8.
    Your 6.5 Hz would divided to produce a period of about 41 minutes.

    i jus want to know the idea behind it. does the 10uF and 1000uF serve as a bypass capacitor? i was jus learning about high pass/low pass filter so does it follow that concept?

    If the output of the power supply was not filtered, the meter would read the average value of the pulses which may be about 0.7 times the peak value.
    Putting a capacitor in would make this voltage rise to the peak value and this would be an obvious change on the meter.
    If the supply was filtered, adding more capacitance would have little effect off load.
     
  18. Feb 16, 2010 #17
    Thx for info on binary divider. But I don't understand y i wud need one to divide my frequency. I simply want my flip-flop to switch on/off with the clock. Right about now, i simply want my flip flop to work properly and you aren't telling me how to get that to work.(don't take it the wrong way) I mean wat does it have to do with clock wen i simply want it to turn it on/off ?
     
  19. Feb 16, 2010 #18

    vk6kro

    User Avatar
    Science Advisor

    I told you how to get it going.

    Connect the J and K inputs to +9 Volts.

    Then you have to get the clock input to go from 0 volts to 9 volts and back.

    Whether this is the output from the 555 (which is what you said you wanted to do) or if you do it with the switch doesn't matter.


    To use the switch, you would probably need a 10 K pulldown resistor from the clock input to ground since the IC is a CMOS device.
     
    Last edited: Feb 16, 2010
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook