Author Topic: Purely analog decade counter  (Read 4344 times)

0 Members and 1 Guest are viewing this topic.

Offline juanfermedTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: gt
  • Fully characterized performance
Purely analog decade counter
« on: January 18, 2013, 05:25:36 am »
Hi to everyone. I have to make a projec in which 4 sets of 4 LEDs have to turn ON and turn OFF, in sequence. Something like a decade counter but using only basic componentes. The thing is that for one set, each light has to be ON for a specific time, say 25 seconds, then turn off, inmediately the next one  turns on for 10 s, off, and inmediately the next for 5 s, turns off and the last one for 20 s, turn off and repeat everything again, so that all the sequence lasts for 60 seconds (1min +- 1%) , each set must not have the sime timing (no other set can have de 25, 10, 5 and 20 s combination) but they all must have a sequence period of 60 seconds. 

The only components I can use are: capacitors, inductors, resistors, bjt transistors (2n3904, 2n3906) and voltage regulators and that is why I added the name "purely analog", because I can not use discrete gates, but I can make them with the components I mentioned.

I must do it with the less component count possible, so I thought about doing something like a ring oscillator with inverters, because it is simple to build a inverter with transistors, only one is needed, and use some large capacitors and resistors for the delays, but it must have the sequence above, in which one light is ON and AFTER it turns off the next turns ON, which is something a ring oscillator does not meet.

I have not been able to make a ring oscillator with a "variable" time between inverters and to make them turn off automatically after that time and "pass" the "logic High" to the next inverter in the sequence.

 In my mind is a matter of charging a capacitor up to a voltage, use a comparator to trigger the output to HIGH, from that time start to count the seconds, after the capacitor reaches a maximum voltage start to discharge it and when it falls to a voltage specific voltage, compare it again and that would set the output to LOW, and make sure the time between HIGH and LOW is the time I desire, and also send the signal to the next stage to start.

  I would really thank anyone here who could suggest me any ideas on how to make this, the charge and discharge circuit for a capacitor and a voltage comparator (without opamps).  I have done my search for some time right now but I am still confused  :-[ . Thanks a lot!
Batteries are, like any other research area...an area in research.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: Purely analog decade counter
« Reply #1 on: January 18, 2013, 06:10:40 pm »
First have a stable supply voltage, then use thyristors ( C106D) as the switching devices in the ring oscillator. Google a thyristor oscillator and you will see it uses a minimum of components, though you will need to select the RC values on test.

Pity you cannot easily get unijunction transistors, though you can fake them using 2 transistors in a PUT configuration in some cases.
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 5121
  • Country: nl
Re: Purely analog decade counter
« Reply #2 on: January 18, 2013, 06:29:14 pm »
Too bad you can't use neon bulbs:


Keyboard error: Press F1 to continue.
 

Offline jeroen74

  • Frequent Contributor
  • **
  • Posts: 396
  • Country: nl
Re: Purely analog decade counter
« Reply #3 on: January 18, 2013, 11:33:44 pm »
I think a string of monostable multivibrators should do the trick, possibly capacitively coupled.
 

Offline juanfermedTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: gt
  • Fully characterized performance
Re: Purely analog decade counter
« Reply #4 on: January 22, 2013, 12:48:56 am »
Hi everyone, thanks for your ideas.

@ SeanB. Thanks for your idea, I will look for the thyristor oscillator. I know understand what you meant when saying that I would have to select the RC value on test. I didnt know this devices, their I-V relationship is somewhat extrange, it is not a function. If I am not wrong, for one voltage there is a region where there are two posible currents. the V-I relationship is a function anyway. I will try to get one of this devices as soon as I can, I hope I find them as the electronics stores here are not very varied.

I looked at the put and unijunction transistor also, and there is something that confuses  me:  an unijunction transistor, a thyristor and a SCR are the same? are those synonymous words? The C106 Datasheet says " Sensitive Gate, Silicon Controlled Rectifiers, Reverse Blocking Thyristors" . Also, I looked at a some tutorials on thyristors and they treated them as "unijunction transistors", so that confused me a little bit more.

@PA0PBZ , thank you for your suggestion, the video is really interesting. Yeah, I cannot use Neon Bulbs, I need various colors and they are not quiet available here. Anyway, how did you get to do those bulbs light like that? Is it a characteristic of the bulbs or there is a "control" circuit?

@jeroen74  Yeah, I thought of that before, but a string of monoestables is not quiet the "least component count" approach, I was making some numbers befores and I had to do 15 monoestable multivibrators for that. Anyway, there is something very interesting and that I do not understand in your words:  "possibly, capacitively coupled". I know that in almost any way I have come up to do this, I need a trigger signal after one signal has "finished" to light on, what is the capacitive coupling for? does a capacitor in series between stages creates a little pulse that may work as a trigger? this is something I havent had any idea and I would really appreciate if you could lead me on this.


Batteries are, like any other research area...an area in research.
 

Offline jeroen74

  • Frequent Contributor
  • **
  • Posts: 396
  • Country: nl
Re: Purely analog decade counter
« Reply #5 on: January 22, 2013, 12:59:11 am »
Yes, you can use a capacitor and a resistor to create a small pulse from a digital signal that changes state. Basically, it's just a high pass filter, filtering out the DC/static level, leaving only the changes.

edit:
like this:
http://www.seekic.com/circuit_diagram/Measuring_and_Test_Circuit/EDGE_DETECTOR.html
« Last Edit: January 22, 2013, 01:03:03 am by jeroen74 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf