Shadow signal

[KrisBlueNZ]

You can add 2 flip flop ic and an inverter to create that shadow waveform.

Are you sure you understand the requirements? Can you draw up a schematic?

Of course, you need to define what happens if there is a long input pulse and then a shorter input pulse which ends *during* the shadow pulse.

That won't happen unless the frequency suddenly increases by a factor of 5 or more within the space of one cycle. Can that even happen?

I'm having a "bit of bother" with my circuit. My oscillators are amplifying and my amplifiers are oscillating If anyone else wants to jump in with a design, please do!

 

[abuhafss]

Can anyone explain me the reason of overlapping of the shadow pulse over the original pulse, in my design posted at #8?

 

[KrisBlueNZ]

Yes. The whole approach you've used isn't going to work. Switching power to the second oscillator ON and OFF won't synchronise it to the first one; in any case, how would the second oscillator track the frequency of the first oscillator? The whole idea isn't workable I'm afraid.

The sensible way is to charge a capacitor at a constant rate during the input pulse, then discharge it at the same rate after the pulse; the "shadow" pulse is active while there's any charge on the capacitor. An op-amp integrator is the simplest way to do the capacitor charge/discharge, and I've got that part doing its thing. I'm having trouble defining the output pulse.

 

[abuhafss]

The whole approach you've used isn't going to work. Switching power to the second oscillator ON and OFF won't synchronise it to the first one

The second oscillator is not being switch on and off. In fact, it is powered on at the fall of the first original pulse and then that state is retained using SCR as latch.

in any case, how would the second oscillator track the frequency of the first oscillator? The whole idea isn't workable I'm afraid.

The second oscillator is a copy of the first one and the frequency varying POTs of both oscillators, R2 & R7 are ganged won't the frequencies of both the oscillators be the same?

The sensible way is to charge a capacitor at a constant rate during the input pulse, then discharge it at the same rate after the pulse; the "shadow" pulse is active while there's any charge on the capacitor. An op-amp integrator is the simplest way to do the capacitor charge/discharge, and I've got that part doing its thing. I'm having trouble defining the output pulse.

I am anxious to see it completed.

 

[KrisBlueNZ]

So the duty cycle is always 20%? How about running the oscillator five times faster than what's required and clocking a 4017 set up to divide by 5. Then you will get a nice clean pulse for each 20% of the total cycle. You can pick off two adjacent ones for your main pulse and your shadow pulse.

 

[abuhafss]

So the duty cycle is always 20%? How about running the oscillator five times faster than what's required and clocking a 4017 set up to divide by 5. Then you will get a nice clean pulse for each 20% of the total cycle. You can pick off two adjacent ones for your main pulse and your shadow pulse.

Yes the duty cycle is fixed 20%.

Hmm, need to try with 4017. Do you mean the first oscillator (running 5 times faster) would provide the clock to 4017?

 

[KrisBlueNZ]

Yes.

 

[KrisBlueNZ]

 

[abuhafss]

Yes.

Yeah, that did it.

 

[KrisBlueNZ]

LOL! I guess that's why it's so important to explain your requirements clearly. I assumed you had some uncontrollable signal that you wanted to echo. That's why I was trying to measure it with a capacitor. But it turns out you're generating the signal yourself so the solution is very simple.

 

[abuhafss]

LOL! I guess that's why it's so important to explain your requirements clearly. I assumed you had some uncontrollable signal that you wanted to echo. That's why I was trying to measure it with a capacitor. But it turns out you're generating the signal yourself so the solution is very simple.

With due respect, I had clearly mentioned and illustrated that it is a square wave with variable frequency. Besides, my design clearly indicated that the original signal is generated using an op-amp oscillator. Anyway, I'll be more careful in future

And last but not least, thanks for your help and opening doors for me to the decade counters.............I have never played with them, this would be my first project with 4017. And definitely, it will broaden my skills.

 

[KrisBlueNZ]

With due respect, I had clearly mentioned and illustrated that it is a square wave with variable frequency.

Yes, I knew that.

Besides, my design clearly indicated that the original signal is generated using an op-amp oscillator.

Yes I saw that but simulations often include placeholder circuits to simulate a separate part of the system, so the fact that an oscillator exists in the simulation doesn't necessary mean it's part of your overall project.

And last but not least, thanks for your help and opening doors for me to the decade counters.............I have never played with them, this would be my first project with 4017. And definitely, it will broaden my skills.

No problem

 

[abuhafss]

Well, the 4017 worked perfectly if the original pulse is square-wave.
If the original pulse is a triangular, can a triangular shadow pulse be obtained with some analog circuit?

 

[Laplace]

If you are generating the triangular wave from the first pulse, could you not then generate the shadow triangle from the shadow pulse in the same way?

 

[abuhafss]

If you are generating the triangular wave from the first pulse, could you not then generate the shadow triangle from the shadow pulse in the same way?

I am not generating the triangular pulse from the first pulse. In fact, I am replacing the first square wave generator with a triangular wave generator. The 4017 output are high or low, so the triangle gets converted to square.

 

[hevans1944]

I don't want to intrude too much on this fascinating discussion, but what exactly are you trying to accomplish with variable-frequency square or triangular pulses followed by so-called "shadow" waveforms? BTW, the two oscillators with ganged frequency controls are not coherent to each other, hence the eventual overlap of their outputs. Would you please provide some background information concerning the eventual application of whatever circuitry someone eventually comes up with? My first thought on this was a digital delay line (shift registers) clocked at some suitable multiple of the maximum pulse frequency, but I like @KrisBlueNZ approach of an analog integrator charging at the leading edge and discharging at the trailing edge, the output pulse being the duration of the discharge part of the cycle. A similar approach could work for triangle wave inputs, but more attention to when the triangle wave ends, in order to define when the output triangle wave starts, would require some more thought.

Please "fill in the blanks" for us! What will this thing DO? Is it a plan to rule der verld or a plan for world peace?

 

[abuhafss]

Actually, I am trying to build a pulse train generator with variable frequency, something like this:

The originating generator could be a square wave or triangular wave. The duty cycle to be about 20-30% fixed hence the pulse width would change according to the selected frequency between 10 and 250 Hz.

My idea is to generate a shadow pulse immediately when the original pulse gets to zero, then invert it and merge it with the original pulse to get a pulse train. I have accomplished the square wave pulse train using 4017 but stuck with triangular version.

 

[hevans1944]

Let me think about this. I have seen such waveforms from biological muscle stimulators, both as you describe as well as polarized. We used these on dissected frog legs back in the day, part of a university heart research program. Don't even ask why frog legs should be applicable to this type of research.

 

[abuhafss]

Introducing an RC integrator at the two outputs of 4017 resolves the issue.

Thanks all of you for your inputs.