electrical delay line oscillator 100mhz

[robowaffe]

So here is an attempt to design an electrical delay line oscillator,  I couldn't label the emitter and collector properly, but its some way around I'm not sure yet.

You close the switch and it will inject the capacitor pulse into it, and the transistors will continue to repeat it like its a little particle accelerator on your desk.  The cap pulse will go around in a loop indefinitely, keep getting repeated at each side of it.

This cannot have a speaker output, because its too high frequency, you need to have an average oscilloscope to see if it oscillates,  I'm not sure if it does, and I don't have an oscilloscope to test it.

So I'm not sure how I'm going to do this, without an oscilloscope.
At 4m wire length its under 100 mhz in oscillation speed at 4 metres im just saying you need 100mhz rated oscillator to read the oscillation of it.

If you decrease the length of the wire to 1mm, but the transistor legs are too long, it should be approximately 100 gigahertz.

[gjvdheiden]

100 mhz you should be able to follow with a multimeter on DC voltage, almost to slow for an analog oscilloscope.

[robowaffe]

oh, thats a bit strange, I was thinking most of the time people read kilohertz off their oscilloscopes.

what do your analogue oscilloscopes like frequency wise?

[jasonRF]

oh, thats a bit strange, I was thinking most of the time people read kilohertz off their oscilloscopes.

what do your analogue oscilloscopes like frequency wise?

gjvdheiden is mostly pointing out that you are asking about building a 100 mhz oscillator.  Note that 100 mHz is a low frequency that has a 10 second period.  Many multimeters update more than once per second, so you can just watch the numbers on the screen to see if there is oscillation.

Of course, we all suspect you are trying to build a 100 MHz oscillator, which has a 10 nanosecond period, but that is not what you asked about...

jasn

[gjvdheiden]

Yes, please, on a EE forum use the right units. Most of us can't sleep at night after seeing such horror. ok ok, I'm exaggerating a bit.

But your question is how to check the oscillator to see it works?

Maybe a band pass filter, really fast diode followed by a cap (peak detect). Crude way to detect the frequency is present. Just an idea.

[Ian.M]

That's not an oscillator, its a SCR!

See: https://www.electrical4u.com/two-transistor-model-of-scr/

When you close the switch, the capacitor (if initially uncharged) dumps a current pulse into the base of the PNP, turning it on, which turns on the NPN, holding the PNP on thereafter. Once both transistors are on, they stay on till power is removed.  No oscillation.

N.B. in practice, if you try to build this from small signal transistors their very high gain + tiny leakage currents often result in the SCR circuit latching on at powerup.  To avoid this add high value resistors from base to emitter of each transistor to divert the leakage currents and reduce their sensitivity. 

[Zero999]

That's not an oscillator, its a SCR!

See: https://www.electrical4u.com/two-transistor-model-of-scr/

When you close the switch, the capacitor (if initially uncharged) dumps a current pulse into the base of the PNP, turning it on, which turns on the NPN, holding the PNP on thereafter. Once both transistors are on, they stay on till power is removed.  No oscillation.

N.B. in practice, if you try to build this from small signal transistors their very high gain + tiny leakage currents often result in the SCRv circuit latching on at powerup.  To avoid this add high value resistors from base to emitter of each transistor to divert the leakage currents and reduce their sensitivity.

You beat me to it. Note it's unsuitable for high currents because all of the current flows through the bases, which will have a much lower rating, than the collectors.

[robowaffe]

When you close the switch, the capacitor (if initially uncharged) dumps a current pulse into the base of the PNP, turning it on, which turns on the NPN, holding the PNP on thereafter. Once both transistors are on, they stay on till power is removed.  No oscillation.

But there is a 4 metre wire between the transistors, and this is supposed to introduce a 100 millionth of a second delay, and ur supposed to make the cap small enough so it stops the current before its hit the base of the transistor its going to.  (so the pulse is shorter than the journey to the transistor.)

[Grandchuck]

Capacitor discharge commutation?

[robowaffe]

Capacitor discharge commutation?

Im guessing its going to end up more of a square wave once it gets stuck in the loop, cause of the way the transistors amp it each time, sorta flattens it out.

[wasedadoc]

Transistors tend to turn on faster than they turn off. Even if you get the correct value capacitor which gives a suitably short initial pulse that avoids the two transistors being on simultaneously the two on times will increase with each cycle until there is overlap. At that point, both transistors are on and no more oscillation.

Also the frequency would be much lower than 100 MHz. Each cycle has the delay of two 4 metre wires. Furthermore, velocity in wire is lower than the speed of light.

[robowaffe]

Transistors tend to turn on faster than they turn off. Even if you get the correct value capacitor which gives a suitably short initial pulse that avoids the two transistors being on simultaneously the two on times will increase with each cycle until there is overlap. At that point, both transistors are on and no more oscillation.

Ah really,  what would you do about this if u were making this? 

<edit> wouldnt just a longer delay time fix that?

[wasedadoc]

Transistors tend to turn on faster than they turn off. Even if you get the correct value capacitor which gives a suitably short initial pulse that avoids the two transistors being on simultaneously the two on times will increase with each cycle until there is overlap. At that point, both transistors are on and no more oscillation.

Ah really,  what would you do about this if u were making this? 

<edit> wouldnt just a longer delay time fix that?

I would not try to fix that because I would not start with such a design.

[robowaffe]

Transistors tend to turn on faster than they turn off. Even if you get the correct value capacitor which gives a suitably short initial pulse that avoids the two transistors being on simultaneously the two on times will increase with each cycle until there is overlap. At that point, both transistors are on and no more oscillation.

Ah really,  what would you do about this if u were making this? 

<edit> wouldnt just a longer delay time fix that?

I would not try to fix that because I would not start with such a design.

So u have another way to make an electrical delay oscillator?


And I've thought of a solution for that,  make it emit to invert a power which charges a cap into the other transistor (so the emitter discharges the cap, then it recharges when its off, but it puts another delay on it.), and then thatll keep the pulse width down.
Thanks for warning me about the problem,  very much helped me if u are right.  thanks.

But maybe that screws with the delay of it,  so its no good!!!  i need the delay to be perfect to the electrical delay time.  damn need to keep thinking.

[wasedadoc]

Maybe you should think less and use the gained time to learn how to use a search engine.

[robowaffe]

Maybe you should think less and use the gained time to learn how to use a search engine.

Could you please show me the beautiful link to it?

[MK14]

So u have another way to make an electrical delay oscillator?

https://en.wikipedia.org/wiki/Ring_oscillator