#969

[@johnwest7993]

Lower a big powdered iron cylinder down into the center of your coil and watch what happens. I think that delay effect is almost entirely due to the inductance of the spool of wire as opposed to the length.

[@HerrHeisenheim]

Are you sure the change in frequency is due to the length of wire, not the added resistance and inductance from this spool of wire?

[@joop2295]

IMHO it's the inductance of the coiled wire! Would be nice if IMSAI Guy would measure the inductance of the reel of wire.

[@ploegmma]

That was my thought too. It would be more interesting to analyse and explore what effect is at play most here. Measure the inductance of the spool and replace it with something similar (other then a long wire).

[@marko.692]

Assuming that the capacitor has value of 180pF, I'm guessing that inductance of the spool is about 166mH.

[@JoseSilveira-newhandleforYT]

@@joop2295 Yep!

[@pbaemedan]

This circuit is a variation of a relaxation oscillator. This circuit was the start of digital flip flops. The interesting change with the wire was the change in duty cycle.

[@IMSAIGuy]

if it ever works

[@RexxSchneider]

The duty cycle is difficult to calculate because the 2N2222, like most NPN small signal transistors, has a breakdown voltage between base and emitter of around -6V. So if the supply voltage was for example 12V, when the second transistor switches low, the base of the first transistor will be driven below ground and the transistor EB junction will break down, discharging the capacitor much more rapidly than expected from a naïve analysis.

[@RexxSchneider]

The basic collector bias circuit provides negative feedback from collector to base and stabilises the bias point. The description starting at 1:04 is wrong. The transistor is indeed "on", but it's in its linear region. The voltage at the node isn't the "transistor drop" (i.e. Vcesat), but it can be calculated if we have the supply voltage and the Hfe of the transistor. For a 12 volt supply and a Hfe around 100, the node will be at about 2.8V. If the Hfe is 50, then the node will be around 4.3V. You'd need a current gain of more than 800 to drive the voltage there down to 1V and that's extremely unlikely with a 2N2222.

[@jstro-hobbytech]

Very very cool. Almost as cool as that scope. Great video man.

[@Enigma758]

It would be interesting to repeat the same experiment but with the wire extending straight, as opposed to being in a coil.

[@dri50]

another way to look at this circuit is: The first transistor inverts and the next transistor inverts. So the input signal appears at the output of the second transistor with no inversion. Wrap that signal back to the input and that is POSITIVE feedback, which results in an unstable circuit. And they usually oscillate.

[@IMSAIGuy]

it is a self flipping flop

[@R2AUK]

A fun experiment indeed. Thanks for sharing!

[@Ali6x944]

Isn't the delay caused by the additional series inductance/resistance in the circuit? If not, then where is it coming from and how can we calculate/measure the propagation delay of the wire? Really intrigued and eager to know 👀

[@IMSAIGuy]

1nS per foot

[@Ali6x944]

thanks @IMSAI Guy ❤️❤️

[@RexxSchneider]

@@IMSAIGuy Says who? That's the speed of light in a vacuum. The speed of signal propagation in a wire is necessarily less than that, and you don't have a good way of calculating it from first principles. If it were coaxial cable with known capacitance and inductance, then we could calculate the propagation delay as √(LC). For example using RG-58, you get 1.5ns per foot (or 30cm), but you don't have any such information about a random coil of wire on a spool.

[@absurdengineering]

@@RexxSchneider Propagation velocity in a wire in free space can be calculated from first principles :)

[@RexxSchneider]

@@absurdengineering But the circuit isn't in free space. Take a look at the circuit that was built at 2:25. Does that look like free space to you? There are lots of parasitic capacitances and inductances that will make a nonsense of any attempt to calculate the actual propagation delay from first principles.

[@Proud2bmodest]

This is a simple relaxation oscillator and is controlled by the charge and discharge time of the capacitor. The increased series inductance reduces the charge and discharge currents and slows the oscillator down. The coil is a very poor transmission line and it's unlikely that a pristine pulse is travelling through the coil to switch the transistors.

[@abhijitrashinkar]

I liked your experiment

[@jjoeygold]

What happens if you put a large magnet or metal in the coil and move it up and down :)

[@kirkkohnen5050]

That spool is best modeled as an inductor.

[@fer_fdi]

Very very nice! The delay introduced by the long wire is fascinating...! I wonder what's the resistance of that wire... Just thinking on how to control frequency of it other than changing the capacitor... making the two 50k resistors variable maybe?

[@jonka1]

Please make a follow-up with a variable resistor in place of the long wire. Also try component values to affect the ratio of the square wave.

[@byronwatkins2565]

Was the crystal's frequency less than 137 kHz? Higher frequencies require faster amplifiers (lower collector resistors), so you will need to choose the resistor values with far more care. You will also need to choose values that yield more of a square wave than a pulse train. Connect your probe to the left collector or your scope will short out the crystal.

[@SeAfasia]

Hi....replace the capasitor with LC tank circuit...?

[@fabiotrevisan8922]

I'm sorry to say, but you got the whole concepts wrong... Like Dri50 mentioned, this oscillator oscillates because it's got positive feedback, and positively back-fed amplifiers are just prone to oscillating. Whatever reactive elements you put in the feedback path just makes the oscillation to happen at some frequency. Lengths of wire... Is not exactly what is more likely at play when you add a long coil of wire... You're actually creating an inductor of a relatively high inductance value. This amplifier, being non inverting as it is and having very little delay itself, it's already set at the most perfect condition for oscillation (at pretty much any frequency from DC to a few tens of megahertz). When you add reactive elements (e.g. capacitor and or inductors) the phase shift they introduce actually makes the circuit harder to oscillate than it is without them, creating conditions for oscillation to be more likely at specific frequencies.

[@inspectormills3290]

It would be great if you could show the test points where the oscilloscope is connected

[@IMSAIGuy]

scope connected to ground and collector of second transistor

[@AugustineAriola]

I love your video but I am somehow unclear how the second transistor will turn off after turning on. I knew the negative accumulation on the capacitor plate will turn off the first transistor but the second transistor turning off beats my imagination.

[@krish2nasa]

Interesting oscillator circuit. Do you know any simple oscillator circuit with just one transistor and RC components? Thanks

[@IMSAIGuy]

yes, here is one: kzbin.info/www/bejne/rJbJgoqghLGmfMk

[@krish2nasa]

@@IMSAIGuy Thanks

[@bobdoritique7347]

Merci. Does the oscillator diagram (without the long wire) have a name please?

[@IMSAIGuy]

while not exact, it is similar to phase shift oscillators

[@bobdoritique7347]

@@IMSAIGuy Merci.

[@mr1enrollment]

it would be interesting to see the freq change with change in V+

[@Pyromonkey360]

what about a piezo?

[@rfdave3980]

What happened with the FLL (Frequency Locked Loop) thingy. You still work on it?

[@IMSAIGuy]

yes

[@RixtronixLAB]

Nice video clip, keep it up, thank you :)

[@pufango4059]

Yes but what use is it ?

[@IMSAIGuy]

use it like a 555 chip. it generates a clock. You could use it to measure the inductance or length of a wire. or a capacitors value. Flash an LED.

[@DAVIDGREGORYKERR]

Maybe that meant a Butler Oscillator.

[@mountvernon5267]

It's 11.8" per nanosecond. I still have the 'nanosecond' that Grace Hopper gave me...along with a packet of picoseconds (ground black pepper).

[@IMSAIGuy]

I find it amazing how slow the speed of light really is.

[@MikeFikes]

Hah. I tried the circuit with a 10,000 µF capacitor and got a nice leisurly 200 s period.

[@IMSAIGuy]

nice, how much wire is needed for 200 seconds? 1/ns per ft. so, 200 x 10^9 nS or 200000000000 feet

[@MikeFikes]

@@IMSAIGuy Hah. Then we are in the world of the latest Veritasium video-trasmission line theory involving long wires-but replacing lunar with Martian distances. :)

[@irgski]

Maybe make a metal detector out of this?

[@__--JY-Moe--__]

So you have made a switch , with no switch ! So how will you discipline the child?

[@stephenamy9879]

you did not mention it being a resonant circuit....

[@IMSAIGuy]

that's because it is not. it is a relaxation circuit.

[@azav8raa]

Cool little oscillator circuit. An additional experiment would be to see how much the frequency changes with temperature. Hmmm... Negative or Positive coefficient? I'm guessing it would oscillate faster as the transistors get warmer.