Challenge Thread: The fastest breadboard oscillator on the mudball

[joeqsmith]

Mounting the connector to the breadboard has been a problem.  The cables will put a fair bit of stress on it.   One solution would be to bolt a bracket to the breadboard with a panel mount connector.   I made up some Teflon cables today to connect to the breadboard and DSO.  These cables are very flexible, like a wet noodle.    I also have added a couple of copper wire loops around the coax to help secure it.   

[TheUnnamedNewbie]

But if "conventional" EM waves are the subject, the line between microwaves and IR is 300GHz. How close have hobbyists been able to get to that line?

I don't fully agree that the line between microwaves (well, technically millimeter-waves at that point) and IR is 300 GHz, generally people would say that you have sub-THz first (300 GHz -> a few THz).
There are at least a few HAMs (WA1ZMS, W4WWQ, W2SZ, WA4RTS) that have done some stuff at those frequencies, 300-400 GHz. This was all with 'conventional' RF techniques (IE, no lasers or optics, but transistors and diodes and mixers).

While the numbers are impressive, it isn't that fundamentally different from RF at a few GHz. The main issue is just that *everything* matters, and you need to use waveguides or expensive very small PCB technologies.

[joeqsmith]

Demonstrating LeCroy's Digital Bandwidth Interleaving along with my simple down converter, limitations with nail polish and my more recent attempt at making a fast oscillator on a breadboard.     

[herc]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components. details can be found at this link:

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/oscill.htm

[joeqsmith]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components. details can be found at this link:

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/oscill.htm

(Attachment Link)

I have not heard of the "twist" oscillator but the idea of using feedback has been around since the oscillator and using wire to form capacitors, inductors and delay lines is certainly nothing novel.   While it's more commonly known as a phase shift oscillator,  I guess I could call mine the double summersault oscillator with twist technology feedback.    Add a few spark plugs and claim it has something to do with free energy.     

[SiliconWizard]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components. details can be found at this link:

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/oscill.htm


(Attachment Link)

It's interesting! Getting ahold of the MSA08-86, which is apparently an Agilent part, may not be that easy though.
This shop seems to have it: https://dutchrfshop.nl/en/drfs-parts/73-msa08-86.html

(The whole article, in french: http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/twist.htm )

[SilverSolder]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components. details can be found at this link:

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/oscill.htm

(Attachment Link)

I have not heard of the "twist" oscillator but the idea of using feedback has been around since the oscillator and using wire to form capacitors, inductors and delay lines is certainly nothing novel.   While it's more commonly known as a phase shift oscillator,  I guess I could call mine the double summersault oscillator with twist technology feedback.    Add a few spark plugs and claim it has something to do with free energy. 

I'm guessing we are dealing with a literal translation from French, somewhere along the line -  what's the betting that the French word for "twist" is used similar to the English word "phase", right?

[joeqsmith]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components. details can be found at this link:

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/oscill.htm

(Attachment Link)

I have not heard of the "twist" oscillator but the idea of using feedback has been around since the oscillator and using wire to form capacitors, inductors and delay lines is certainly nothing novel.   While it's more commonly known as a phase shift oscillator,  I guess I could call mine the double summersault oscillator with twist technology feedback.    Add a few spark plugs and claim it has something to do with free energy. 

I'm guessing we are dealing with a literal translation from French, somewhere along the line -  what's the betting that the French word for "twist" is used similar to the English word "phase", right?

I assumed it was from their picture of the twisted wires.   I looked it up using a French dictionary and it appears to have the same use.   We would need a French EE to chime in.   

[SiliconWizard]

Indeed, it's related to the resonant circuit which is basically a twisted pair. "Le circuit résonant  est constitué d'une simple torsade, le "twist", dont la longueur conditionne la fréquence de fonctionnement de l'oscillateur. "

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/torsade.htm

[SilverSolder]

Indeed, it's related to the resonant circuit which is basically a twisted pair. "Le circuit résonant  est constitué d'une simple torsade, le "twist", dont la longueur conditionne la fréquence de fonctionnement de l'oscillateur. "

http://jf.fourcadier.pagesperso-orange.fr/hyperfrequences/twist/torsade.htm

So the "twist" is the whole LC circuit - both inductor and capacitor - in this design?   Or does it act as a "gimmick capacitor"?

[T3sl4co1l]

It's an open transmission line stub.  The first mode is a series resonance; it's a Pierce oscillator I believe.

Tim

[joeqsmith]

Playing around with the new transistors..

[trash]

do you know the "twist oscillator" ? it operates in the GHz domain and  it is very intriguing, because it just needs 5 components.

I'm guessing you're just referring to the BFR93 section and not the MSA08 MMIC. Does MMIC count as an "IC"?
(Monolithic Microwave Integrated Circuit)  Because with a careful piece of what man magic, I can make those critters oscillate very easily.
In some cases they have issues with them oscillating a bit too often.

Looks like I'm going to have to get in on this gig because I've got some ideas that really need to be put to the bench and now I have an excuse.
 

[LaserSteve]

So if I stick a 10 mW X-band Gunn diode on a soldered breadboard does that count? I do not have any solderless ones.  I know for a fact I will see a Vhf fundamental. Then big spikes in X-BAND.

IN other words does it have to be clean? Because with a spark gap you could make oscillations out to say 60 Ghz.

Steve

[SilverSolder]

So if I stick a 10 mW X-band Gunn diode on a soldered breadboard does that count? I do not have any solderless ones.  I know for a fact I will see a Vhf fundamental. Then big spikes in X-BAND.

IN other words does it have to be clean? Because with a spark gap you could make oscillations out to say 60 Ghz.

Steve

I was thinking the same thing (spark gap) but the challenge of filtering out and amplifying a signal up to 5V is still there...  and not exactly easy for high frequencies!

[joeqsmith]

22AWG solid TFE insulation swept from 2.2-2.6GHz.

[SiliconWizard]

It's an open transmission line stub.  The first mode is a series resonance; it's a Pierce oscillator I believe.

Absolutely. And it's mentioned in the article: "L'oscillateur est de type Pierce"

[SilverSolder]

22AWG solid TFE insulation swept from 2.2-2.6GHz.

I'm not good at interpreting Smith charts...  What does it mean, that the plot follows the circle?

[joeqsmith]

22AWG solid TFE insulation swept from 2.2-2.6GHz.

I'm not good at interpreting Smith charts...  What does it mean, that the plot follows the circle?

That's fine.  There are several good videos and text materials available for free on-line if you are interested.   For now, let me just show you a simple test.   In the attached graph looking at S11, I have swept the twisted wires shown above from 2350 to 2450 MHz.   I then swept a small trimmer capacitor over that same range.   

[SilverSolder]

22AWG solid TFE insulation swept from 2.2-2.6GHz.

I'm not good at interpreting Smith charts...  What does it mean, that the plot follows the circle?

That's fine.  There are several good videos and text materials available for free on-line if you are interested.   For now, let me just show you a simple test.   In the attached graph looking at S11, I have swept the twisted wires shown above from 2350 to 2450 MHz.   I then swept a small trimmer capacitor over that same range.

so is the twisted pair adding more capacitance, effectively, than a capacitor of the same nominal value - due to the time it takes for the wave to propagate to the end of it?

[herc]

no, it's called "twist oscillator", because it twists my shortwave-noob brain, trying to understand why a pair of twisted wires can act as a resonator ;-)

-----

https://en.wikipedia.org/wiki/Stub_(electronics)#Resonant_stub

Stubs are often used as resonant circuits in oscillators and distributed element filters. An open circuit stub of length l will have a capacitive impedance at low frequency when β * l < π / 2 . Above this frequency the impedance is inductive. At precisely β * l = π / 2  the stub presents a short circuit. This is qualitatively the same behaviour as a series resonant circuit.

[trash]

My first attempt came up with ~1.6GHz, I was able to push it to 1.8GHz and keep the stability.

[joeqsmith]

My first attempt came up with ~1.6GHz, I was able to push it to 1.8GHz and keep the stability.

   Nice job. 

[trash]

Hitting a bit of a wall at the moment. 2GHz seems to be where I can get a stable carrier, above that it gets a bit sloppy.
I haven't resorted to any white man magic just yet and I've been sticking with a fundamental frequency.
I'll get to multiplication trickery later

[trash]

A few people mentioned using a tunnel diode but I didn't see anybody actually try.
So I thought I'd give it a go since I have a few of them sitting in a drawer.
Compared to the transistor based oscillator I'm able to get similar frequencies but there is a lot of spurious noise that comes with them.
You can see on the rectum paralyser there are carriers around 500-700MHz, with a lot of spurious noise at ~1.8GHz and a carrier at 1.895GHz.

The circuit is a simple voltage divider to give 700mV and I've used the wire lead of the diode as an inductor and a home made capacitor using some double sided circuit board sandwiched into the inductor loop. I tried to completely remove the capacitor and just rely on parasitic inductance but the result was more a white noise source around 3GHz rather than what I consider to be an actual carrier. I'll come back to this idea in the future when I've had a bit more time to think about it.

And a still have a few other dirty tricks up my sleeve to get up into C band.