I want to build a simple and discrete circuit (without any integrated circuits) which puts out a square wave (50:50 duty cycle) with N pulses (rising edges)?

For example for N=4 the output should look like this: enter image description here

The time of one period (from rising edge to rising edge) might be in the range of 10 - 100 ms.

Any ideas how to achieve this?

I thought on using a capacitor, which charges and puts out a square wave and another capacitor which "enables" (AND gate?) the signal and is sized the way that it enables as long as 4 pulses take. But I've no idea if this is even possible ...

EDIT: here are some more details:

  • The circuit should be discrete: using only transistors, resistors, capacitors (no "complex" devices/ICs)
  • The N is fixed, meaning that there is a N=4 circuit which generates 4 pulses on button press
  • "... good old dialphone's mechanical approach" - I've looked it up and this is what I want but not mechanically but electronically
  • "What is the maximum number of pulses you want to create, and how do you want to set the count?" N ranges from 1 to 9 and the circuit is built for N, so it is not variable.
  • 1
    \$\begingroup\$ it's possible, that's how computers were built pre-moonflight. Anyway, it might be good if you could define why you're trying this (there's different levels of awkwardness you can apply to this!) and what components you actually want to use, instead of which not. \$\endgroup\$ – Marcus Müller Oct 4 '19 at 18:13
  • 1
    \$\begingroup\$ It's pretty interesting to know how you want your input "N" to come into your system, by the way. (by the way, a very easy way of doing this would involve a good old dialphone's mechanical approach to pulses) \$\endgroup\$ – Marcus Müller Oct 4 '19 at 18:14
  • 1
    \$\begingroup\$ P.S., maybe a chain of "one-shot" pulse generators, each of which triggers the next? Though honestly, if you asked me to design it, I'd specify some itty-bitty microcontroller in maybe an 8-pin package. \$\endgroup\$ – Solomon Slow Oct 4 '19 at 18:51
  • 2
    \$\begingroup\$ "but I'm challenging myself on how to do it in a simpler way." The microprocessor is the simpler way. Anything else will be a pile of complicated, finicky stuff that looks like a robot puked its lunch. \$\endgroup\$ – JRE Oct 4 '19 at 20:10
  • 2
    \$\begingroup\$ No, you're challenging us. Sorry, but I wrote my answer before I saw your comment. \$\endgroup\$ – Dave Tweed Oct 4 '19 at 20:53

Thinking outside the box:

An astable multivibrator that has a diode and cap on the output. When the pulse is high it charges the cap a little bit. After N pulses the voltage will reach a threshold which is detected by a comparator, and that disables the multivibrator.

You'll need a method to reset it by discharging C1 as well.

Here is a conceptual diagram:


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Output should be connected to the other side of the diode. I have considered a similar method that allows for independent control of Fo and N. \$\endgroup\$ – EinarA Oct 4 '19 at 23:52
  • \$\begingroup\$ @EinarA Thanks for that. I was thinking it would just be a diode drop incursion, but depending on the load, the load could discharge the cap. \$\endgroup\$ – Aaron Oct 5 '19 at 0:08

Use a monostable multivibrator to "gate" an astable multivibrator:


simulate this circuit – Schematic created using CircuitLab

Adjust R5 and R6 for the frequency and duty cycle of the pulses. Adjust R7 for the number of pulses.

Here's a simulation result showing 4 pulses at 10 Hz each time the trigger fires.

simulation result

  • \$\begingroup\$ Impressive speed in drafting what is the OP's ( vaguely expressed ) idea; but it doesn't work. When Q5 is on both Q3 and Q4 will turn on and the multivibrator cannot get out of this state. To have it work, connect Q5 to the base of Q3. \$\endgroup\$ – EinarA Oct 4 '19 at 23:39
  • \$\begingroup\$ @EinarA: Yes, I see what you're saying, and yet CircuitLab says it works. I guess that shows how valuable their simulator is (NOT!). I actually tried what you're suggesting, but then it was giving me some "reverse active" action when the collector of Q5 was driven far enough negative, which was messing up the timing for Q3 and Q4. I'll put some more work in on this using LTspice and get back to you. \$\endgroup\$ – Dave Tweed Oct 5 '19 at 0:22
  • \$\begingroup\$ OK, I think I got it now. \$\endgroup\$ – Dave Tweed Oct 5 '19 at 0:37
  • \$\begingroup\$ I didn't think of it two hours ago, but putting a low voltage diode ( Ge suggested since we are being old school ) in series with Q3's collector would block the reverse voltage. What you have now should work too. \$\endgroup\$ – EinarA Oct 5 '19 at 2:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.