Challenge Thread: The fastest breadboard oscillator on the mudball

[BreakingOhmsLaw]

As far as i am concerned the entry is valid. Joe does hold the breadboard into the camera in his video.
I do not require to name parts. It is a competition after all, so some trade secrets should be allowed.

[0culus]

As far as i am concerned the entry is valid. Joe does hold the breadboard into the camera in his video.
I do not require to name parts. It is a competition after all, so some trade secrets should be allowed.

OK, it's been spoken. I disagree, but it doesn't really matter. I was hoping to perhaps learn a few things from other's designs on this thread. If Joe wants to keep secrets, he's free to do so but I do not wish to partake any further. Any remaining desire of mine to work on this has been mostly tamped down anyway by all the unpleasantness. I've got other, more interesting projects to work in my lab that don't involve forum politics. Cya!

[herc]

does someone have a suggestion for a fool proof, easy to built power amplifier (centered around 2.4 GHz, lets say with a target power of 1 Watt ?)
thanks !

[avlijas.sladjan]

Yesterday I found a YT video about this funny contest. Thanks to Joe's video I want to make my own oscillator. In the first try I made my record (4.9GHZ) and after a small tuning (cutting) it was 14.4GHz. With this transistor maybe I can go until 20-21G (transistor will have some gain), but it will be maximum. For me it is enough

[tim_]

Yesterday I found a YT video about this funny contest. Thanks to Joe's video I want to make my own oscillator. In the first try I made my record (4.9GHZ) and after a small tuning (cutting) it was 14.4GHz. With this transistor maybe I can go until 20-21G (transistor will have some gain), but it will be maximum. For me it is enough

Nice, so it's possible to break the 10 GHz barrier without drama.

Can you share anything about the circuit and the transistor you used? This looks like a common base collpits oscillator, but why is the base only biased to the ground?

[avlijas.sladjan]

No problem. This is GaAs FET and needs negative bias, like JFET. With this transistor it is hard to make amp without oscillation

[avlijas.sladjan]

Tonight I tried some small tuning, and got 18.7GHz. Now I was done, for more I will need better transistor and some reconfiguration on board

[joeqsmith]

Looks good so far.   I see 20GHz+ in your future. Good luck!

[avlijas.sladjan]

Thank you Joe .
Maybe it is good to stop here. I'm afraid that my wife will kick me out from house  because I make 5G radiation in children room (my current lab).

[avlijas.sladjan]

[joeqsmith]

I would like to see the "or more".    I had bought some other parts back then to hedge my bets in case the ones I used wouldn't get the job done.  I tried one of them out a while back using a delay line by just shorting the base to the collector.   Fairly decent performance.  After a bit of trimming, eventually damaged the part.   That part is a bit better than what I used for that 25GHz but I can't do anything more with it with the equipment I have available. 

Your so close to 20, may as well keep going until you reach what that SA can show. 

[avlijas.sladjan]

I changed the YIG filter/preselector on this SA and only adjusted to 22GHz. Also I have an Advantest R3271A (100Hz-26.5GHz, not at home right now), and maybe next week I will have it. Also can bypass the yig filter and use signal ID (to have 26GHZ instrument), but my main problem is the transistor. I will try to find a better transistor but it will be a hard task in Bosnia, also in covid period. I have already waited for some components and parts since april.
Before 4-5 month I wanted to buy an external mixer (26.5-40GHz) for HP, but I did not find a reason for this. Now I see that will be good to have it

[joeqsmith]

The mixer would be the choice if the goal is to shatter the record.  If you have easy access to the other SA, it's more than enough to take the lead.

We are fortunate to have distributors here like DigiKey who will sell small numbers of parts and carry a wide selection.   

Hopefully your parts will show up soon enough that you don't loose interest.   Personally I am glad to see you putting your equipment to use.  Here, it's very easy to get your hands on older equipment, but it's rare to see anyone actually use it. 

I gave away my old 22GHz HP SA from lack of use.   That 20GHz mixer and DRO was about the best I could come up with without buying a new SA.   My old counter isn't very sensitive and getting a strong enough signal to drive it was a problem.  You're in a much better position.     

26.5GHz or bust!   

[avlijas.sladjan]

Today I found some transistors (for +20GHz) and ordered them. But I will need to wait, maybe more than 20 days for shipping.

To be honest, I was very happy to have instruments to play it. It was very hard to find and buy here, but I saved every money to buy these instruments. And now after regular work I have a small lab for playing.

[RJSV]

I like this! Kind of mad scramble, for what? A, la, the classic movie, "It's a Mad mad, mad mad World", you gotta dial-in some GREED, there: How about one Philly Steak, per day, free. Now, I'm in!
  I'm thinking Opto-Microwave, seriously; an optical gate or modulator, actuated by microwave, and that responds / reacts to light. Then, your OSC will flash a light, very fast, 'ruining' the gate coherence. Even tho, that gate was really supposed to be microwave actuated. Wrong way, wrong way, but possible to OSC way fast. (?)

[SilverSolder]

I like this! Kind of mad scramble, for what? A, la, the classic movie, "It's a Mad mad, mad mad World", you gotta dial-in some GREED, there: How about one Philly Steak, per day, free. Now, I'm in!
  I'm thinking Opto-Microwave, seriously; an optical gate or modulator, actuated by microwave, and that responds / reacts to light. Then, your OSC will flash a light, very fast, 'ruining' the gate coherence. Even tho, that gate was really supposed to be microwave actuated. Wrong way, wrong way, but possible to OSC way fast. (?)

Make it run on a breadboard and show us!

[joeqsmith]

Today I found some transistors (for +20GHz) and ordered them. But I will need to wait, maybe more than 20 days for shipping.

To be honest, I was very happy to have instruments to play it. It was very hard to find and buy here, but I saved every money to buy these instruments. And now after regular work I have a small lab for playing.

Maybe while you wait, you can bring home the faster SA. 

After playing around with different parts, it seemed as far as making them oscillate and what frequencies could be achieved, I could push them far past the data sheets.  I think the part I used at 6GHz, was only characterized to a couple of GHz.   With it just needing to oscillate, I took the just try it and see what happens approach.  Most of the transistors I tried cost under 50 cents so I wasn't out anything if they get damaged.   Shipping was more than the parts so it made sense to buy a few different ones. 

Good luck.  I'll polish up my crown for you! 

[RJSV]

A single component, I'm loosely suggesting, would properly be interfaced by a 'micro-wave' input (port) but the test would be to ignore that, and start using a pulse of light. It's in the realm of parasitic response, utilized (instead of avoiding the light being there at input region).
  But I gotta read the rules better, as I meant a MEMs device, having a visible light port in and out, plus a micro-wave actuated control input. The control causes disrupted or 'ruined' coherence, eliminating light output emission. Suggestion is try go outside usual device interface, but I guess a try could be made, still within (this) contest.
  A public contest can be where some really smart Humans will shine their best!
  I am wondering, is there a region where a waveguide / active MEMs and RF transistor start losing 'trace' wiring and become wholly independent (200 GHz?).
And, of course that's called a 'Radio'.  Duhooh

[ogden]

I gotta read the rules better, as I meant a MEMs device, having a visible light port in and out, plus a micro-wave actuated control input.

Don't worry, rules are flexible here. As long as breadboard is part of DC suppy circuit for oscillator - your entry most likely will be accepted. Could you elaborate more about your MEMS device with visible light input and output plus a micro-wave actuated control input?

[RJSV]

Ogden, the idea bears similarity with the 'garden' light analogy, pointing an LED light, at the photo-panel. So since the little solar cell drives the switch, any increased light (i.e. from unusual pointing set-up) will cause lamp shut-off.
  Then, such LED shut-off allowing the garden light to come back 'on' thus a circular situation (perpetuates.)
  This is the basis for further speculation, with additional thoughts regarding hysterisis, and time delays, often extending timing range / stabilizing function.
  So, you have bought a few surplus micro-optical 'modulators' and the two ports for optical / infrared laser beams are in / out, looking reasonable. But then, a short instruction / spec. sheet lists: "Novel control port uses 110 nM pulsed maser"...
   What did that mean ?
  Well, apparently, without that micro-wave arriving, the modulator distorts, and output goes mostly internal reflecting uselessly.
  Key to all this, apparently, MFG. Sez the unit must be light shielded, as light will also 'mess up' your coherence, (whatever that is).
  So that is your key: Just take a bit of that infra-red, feed it back around to your 'RotoPhotator' or whatever it's called; The light-pulse, disrupts the means that sent it.
  It's all the simple paradox of any OSC, really.
  Trick and skill really gets going, when Tech person has a handle on timing dynamics, and hysterisis.
  Kept wondering, why the data sheet note insistent on 'shield any nearby source of visible light'

[RJSV]

Ok, starting to hit my stride, (or stagger), but here is continued direction, of last post:
  I would, first, create a 'Contest Goal', that being first one to reach 200... That's 200 GHz.
  Now, addressing the hypothetical 'optical' switch / modulator, I believe this was, still, a switched signal, at say, 2 Thz (terra-herz), being switched at 200 GHz. That means you get a few cycles (about 5 cycles) of the faster signal, per each, that's typical modulated, but maybe needs (lots) of wiggle room, to call it a Sine or nearly so...
  Thanks to T3s14co1l (Tim) for the discussion re: optical signal speeds / voltages / probes reality, a contortion act to comprehend those issues...
How about X-rays, next.  There, maybe have to keep particle physics, not wave.
I make any sense, here ? Some

[ogden]

Ok, starting to hit my stride, (or stagger), but here is continued direction, of last post:
  I would, first, create a 'Contest Goal', that being first one to reach 200... That's 200 GHz.
  Now, addressing the hypothetical 'optical' switch / modulator, I believe this was, still, a switched signal, at say, 2 Thz (terra-herz), being switched at 200 GHz.

Why don't you take that 200GHz oscillator, throw your 2THz source with modulator in the bin and win 200GHz contest?

[RJSV]

Attached photo features a MEMs block, that sits above bread-board, uses 10 VDC supply, but I see my faulty plan: That laser, for switching, was to provide the 'substance' being switched, kind of was supposed to be more of a 'DC' potential, for more common OSC.
That is, a comparator or OP amp switches DC, to create the waveforms, so right there, FAIL... on my effort.
  But, idea there, is a MEMs device, requiring the side control port be shielded, from visable light. Which I immediately violate. See the run of optical fiber, to bring some (light) into the CTL port, in disregard of Spec. So correct, you have an oscillator, but that feedback fiber created that. However, not done there, as the 'ruby' laser violates the contest, being an assumed 300 Terra-hz.
  Keeps me thinking.

I searched on 'Optical Modulator has micro-wave control's and got some good results. Plz see Keyang Photonics.
Research gate had some papers like; 'RF signal port, to phase-modulate Optical 633 nM output'.
A good article, Q.A.Duong, has 2018, called 'Suppression of residual A.M. in microwave controllable field-ready module.'

  So a little un-focussed, but the OSC basics still apply, that is to use a bit of output, for feedback. I'm using feedback dynamics, but 'negative resistance's, got to be there,... just not explained (here).
  Now waiting, while study Tim's comments here.

[ogden]

I see problem in your plan: you connect optical output to microwave input.

[RJSV]

Yes, and that is why I try make a schematic, I can see my own goofs.  Right now, I've an interest, in making a non-linear spectrum scale:

  For EM Spectrum base lines,
  10 T
    9 T
    8 T
    7 T
...
    1 T
      900 GHz
      800 GHz
...
      100 GHz
         90
         80
...
         10 GHz
            9 GHz
 
  Looks boring, but I need self-educate. The diagram should include 10 khz, a low low Rf
More edu, less speculation.