What is the best place to find new Gunn diodes?

[Ben321]

I wanted a gunn diode for microwave experiments. However, I can't find them at the normal places I get most of my electronic components (Digikey and Mouser). I did google searches "digikey gunn diode" and "mouser gunn diode", but they don't seem to have any. Are these such specialty components that even major electronic component distributors like Mouser and Digikey don't have them?

I know there are some on eBay, but they are all Russian components. Probably taken from some Soviet era radar equipment, that's decades old, so likely poor quality and not a reliable component. I want something made by well known semiconductor component manufacturers, preferably new, and not used.

[coppercone2]

or a substitute made of whatever

i think there is a peak gunn diode problem. 

[LaserSteve]

You sure about that?

MACS-007800-0M1R00

Is in stock at Digikey.   I'd start at 10 Ghz if I were new to microwave, but MACOM modules are all over the place, easily available.

[coppercone2]

thats like buying a house to get the toilet

[wraper]

Soviet components are not poor quality, especially those with rhombus - military grade mark. I don't think you can easily get any western branded part, especially for sane price. EDIT: also military part numbers start from digit, consumer grade from letter.

Probably taken from some Soviet era radar equipment

Most of them are brand new (old stock), many in factory packaging.

[DC1MC]

Get your Microsemi Microchip diodes from any authorized distributor:

https://www.microsemi.com/product-directory/rf-discretes/3305-diodes-gunn#selection-table

[bdunham7]

If you want someone that will sell single units retail, here is one:

https://www.rf-microwave.com/en/gunn-tunnel-diodes/445/

[Maurizio1957]

Or:
https://www.rf-microwave.com/it/diodi-gunn-tunnel/445/
Maurizio

[LaserSteve]

Um, unless you have waveguide laying around or a milling machine, buy a module or a used police radar or door opener.
Its so much more fun if you have detector and or varacter diodes included.

 Getting a nice horn or better yet ridged waveguide horn from a 3 band radar detector is a bonus.

Many multiband radar detectors have a Gunn LO at 9.3 Ghz, and usually you can re-tune that to 10 Ghz ham band.

GUNNPLEXOR COOKBOOK is worth a read.

As is the older MACOM appnotes

On the other hand, GUNNs are so easy to get going, rolling your own apparatus is easy, .

Don't apply power unless the Diode is in a waveguide or resonator . Some parts have a bad habit of breaking into squegging  VHF or UHF spurious oscillations that can lead to destruction of the diode if not in a resonant cavity.

WR90. Aka Waveguide 16 is cheap and plentiful on Ebay, for 10 Ghz work.

Every time I've ever needed something from MACOM they have been very friendly.

Steve

[LaserSteve]

Coppercone,

It may be buying the house, but it is a beautiful, precise , Mil-spec toilet.

Steve

[radar_macgyver]

You could also get them from Eravant, they have modules from 24 GHz up to 110 GHz (though the latter are quite spendy).

[Ben321]

You sure about that?

MACS-007800-0M1R00

Is in stock at Digikey.   I'd start at 10 Ghz if I were new to microwave, but MACOM modules are all over the place, easily available.

Thanks for that. Unfortunately that unit is 24GHz, much too high for commonly available receivers. If I had a gunn diode that operated in the range of 10.7GHz to 11.7GHz, then it would be easy to receive it with a standard Ku-band LNB intended for satellite TV reception. All I would need then on the receiving side would be a simple bias-T adapter (something that I could probably buy or even make myself) to inject the DC power for the LNB, and extract the down-converted microwave signal (which would be in the range of 950MHz to 1950MHz, well within the receiving range of RTL-SDRs or even some "police scanners"). Such receiving LNBs are very low cost devices, like this one on Amazon https://www.amazon.com/Dual-Output-Band-High-Universal/dp/B0117BHDHS/

It's the transmitting side that's more difficult to get, and also probably much more expensive. A quick Google search for "transmitting LNB" (which would do the opposite of a receiving LNB, up converting a 900MHz band signal to a 10GHz band signal) doesn't return any desired results, instead showing only receiving LNBs. Either Google's algorithm thinks I'm actually looking for a receiving unit (and ignores the word "transmitting"), or it actually genuinely can't find any transmitting LNBs on well known websites.

Whatever the case, I think it might be easier to build a gunn diode transmitter, instead of trying to find a hard-to-find device like a transmitting LNB. I just need to find a modern gunn diode for the 10GHz band (and hopefully not new-old-stock Russian component).

[Ben321]

You could also get them from Eravant, they have modules from 24 GHz up to 110 GHz (though the latter are quite spendy).

24GHz is too high to be receiviable with reasonably priced receiving equipment. I'm thinking more in the 10GHz range, as I can receive that with a simple Ku band free-to-air satellite TV LNB (and yes it must be the free-to-air type, as these are linear polarized, while paid satellite TV service like DirecTV use circular polarized LNBs instead, and a gunn diode will output linear polarized microwaves to the best of my knowledge).

[Ben321]

Um, unless you have waveguide laying around or a milling machine, buy a module or a used police radar or door opener.
Its so much more fun if you have detector and or varacter diodes included.

 Getting a nice horn or better yet ridged waveguide horn from a 3 band radar detector is a bonus.

Many multiband radar detectors have a Gunn LO at 9.3 Ghz, and usually you can re-tune that to 10 Ghz ham band.

GUNNPLEXOR COOKBOOK is worth a read.

As is the older MACOM appnotes

On the other hand, GUNNs are so easy to get going, rolling your own apparatus is easy, .

Don't apply power unless the Diode is in a waveguide or resonator . Some parts have a bad habit of breaking into squegging  VHF or UHF spurious oscillations that can lead to destruction of the diode if not in a resonant cavity.

WR90. Aka Waveguide 16 is cheap and plentiful on Ebay, for 10 Ghz work.

Every time I've ever needed something from MACOM they have been very friendly.

Steve

I've heard that while gunn diodes require a resonant cavity to set the exact oscilating frequency, at the frequencies in question, almost any metal container will be resonant at some frequency within the diode's operating range. So really all I'd need to do is put the gunn diode into an empty tin can, and it should oscillate when powered, at whatever frequency is resonant in that tin can. With a low cost linear-polarized Ku LNB like https://www.amazon.com/Dual-Output-Band-High-Universal/dp/B0117BHDHS/ (when combined with a 12V DC power source and a bias-T adapter) I should be able to connect the output of the LNB (in the 950 to 1950GHz range) to the antenna input of my RTL-SDR and then just tune around until I find the strongest signal, which will then tell me the exact frequency that the gunn diode is transmitting. Once that is set up, I should be able to slightly vary the input voltage to the gunn diode (keeping it within the range it needs to oscillate) in order to amplitude modulate the microwave output of the gunn diode (more voltage in means stronger microwaves out). Then on the receiving end (the RTL-SDR) I should be able to simply amplitude demodulate the signal in SDR software to retrieve the original audio signal that was being used to modulate the beam.

Once that is setup correctly, I should be able to use it as a point-to-point audio or even data communications system. Just point the tin-can (which acts as both a resonant cavity and a directional antenna) containing the gunn diode in the direction I want to transmit, and then take the Ku band LNB and put it at the location I want to receive, and point it back at the transmitter, and I should have a usable microwave communications link.

The biggest problem isn't knowing how I'm going to set up my equipment. It's getting the key component for the transmitter, the gunn diode itself.

[Ben321]

If you want someone that will sell single units retail, here is one:

https://www.rf-microwave.com/en/gunn-tunnel-diodes/445/

Unfortunately those are listed as new-old-stock. Does nobody make gunn diodes anymore? Has some new technology been introduced that fills the same role, that I'm not aware of?

[RoV]

Does nobody make gunn diodes anymore? Has some new technology been introduced that fills the same role, that I'm not aware of?

Gunn diodes have been obsolete for a long time. They are quite inefficient, rather noisy and not very stable without a PLL. Besides, they require a cavity to oscillate.
DR (Dielectric Resonator) oscillators initially replaced fixed frequency Gunn cavities, being rather stable even without a PLL. These also have disappeared, but you can find many on ebay working around 10 GHz with a decent output power. An example: http://www.telecomponents.com/catalog/oscillatore-dro-10-ghz-mizar-oscillator-microwave-p-172.html
Synthesized oscillators than followed, and became smaller and smaller over the years. Now the VCO at their basis is monolithic, like https://www.mouser.it/ProductDetail/Analog-Devices/HMC733LC4BTR?qs=pceeu5JH%2FH%2Ffv1ib8cM6Og%3D%3D

Don't think that designing a cavity for a gunn is an easy task: The cavity must be polished (and best if silver coated) and accurately sized. The diode position and its electrical connections must be very cleverly chosen for the device to work and correctly dissipate the heat it produces. The cavity coupling with outside is also complicated to design. Finally, also the power supply is critical: the gunn requires typically 6-9 V according to the exact diode. Too low a voltage can damage the diode for excessive current, too high will damage it for excessive voltage...

[Ben321]

Gunn diodes have been obsolete for a long time.

That's what I was afraid of. I wonder where Pasco Scientific gets their supply of gun diodes for their microwave optics kit though? https://www.pasco.com/products/lab-apparatus/light-and-optics/advanced-optics/wa-9316

Since they are selling to the education community, I doubt they are buying these old Russian gunn diodes from eBay. I'm guessing they have access to a microwave devices manufacturing company that makes these gunn diodes still.

And yes, the transmitter in this kit does use a gunn diode. It says so clearly in the manual https://cdn.pasco.com/product_document/Microwave-Optics-System-Manual-WA-9314C.pdf
The manual also says that a schottky diode is the component used in the receiver for detecting the signal.

Unfortunately, there's a couple things wrong with this kit. The most obvious is the price. At $2999 it's well outside my hobbyist budget. Secondly, I've researched it (even contacting Pasco directly), and found that the transmitter is continuous wave only. No possibility to modulate the signal. And the receiving end has a large value smoothing capacitor, designed to give a stable reading on the included analog meter, so you can measure the microwave beam intensity at different locations for the purpose of microwave optics experiments, but this also means that the capacitance is so large it would block any audio frequency signals being received. So it's not suited for a communication link. Furthermore, even if I was able to modify these (add my own wiring and connectors for modulation on the transmitting module, and remove the smoothing capacitor on the receiving module, or even avoid the receiving module and use a cheap Ku band LNB instead) I would still have the problem of the HUGE COST. At just one dollar short of $3000, it's not worth the money, in any way shape or form. I mean, when I saw that price, my jaw dropped. I'm hoping to find a more reasonably priced solution.



I can buy the "basic system", instead of the "advanced" system, for a couple hundred dollars less, but it's still $2799. And even if I buy the transmitter and receiver separately, they are still $825 each for a total of $1650. And even if I bought only the transmitter, and went with a much cheaper Ku band LNB for the receiver, that still is $825 dollars. That may be a reasonable price for professional grade ready-to-use microwave communications transmitter, but this transmitter is only continuous wave capable without modifications. And really all it is is a gunn diode inside a resonant cavity with a horn antenna for output. I'm guessing that the individual parts that make up the transmitter (the metal for the cavity and horn antenna, and the gunn diode itself) don't cost more than $200 for Pasco to purchase and assemble into the product in question.

[LaserSteve]

Its the height of the can that matters,  the diameter has to be small for a high quality factor.

 Round resonators for Gunn diodes died with the Dodo for a variety of very good reasons, including the fact that an iris / hole in a thin brass or copper sheet across the waveguide makes it very easy to control the amount of local oscillator power applied to the mixer diode. Tuning a round resonator on a small scale is, well, difficult.  It takes around 15 mm of WR90 to make a Gunn LO, and another 20 mm or so to couple that LO to an SMA connector.

A 10 Ghz round resonator would be roughly sewing thimble sized.


If you want an ultra cheap 10 ghz source to learn with , HB100
Motion detectors on Ebay are 5-10$, the antenna can be bypassed with miniature hardline coax. They are DRO based.
The stability is poor, but it would give you a learning tool

Sometimes it is easier to start with Harmonic mixing and multiplication as gain is cheap.

This is a Brick, it has a phase locked sapphire rod oscillator followed by a step recovery diode multiplier.
I use bricks because the frequency stability is amazing and the phase noise is low:

https://www.ebay.com/itm/225229582241?hash=item3470baafa1:g:G74AAOSwnCtjXpcT

Another way is 576 doubled to 1152 then multiplied to 10 Ghz.  This is a PCRO

https://www.ebay.com/itm/224629640503?hash=item344cf84d37:g:jMsAAOSwai5hVecv

576 applied to a varacter diode mounted in WR90 followed by a filter made in the same piece of WR90 would give you a few hundred microwatts of 10 Ghz that could be amplified to drive a mixer.

I bought two of those for beacons.

Gunn Diodes are fun for Doppler radar and short
Distance wideband communications. They suck as local oscillators in modern systems. Modern systems need phase locked, narrowband, low phase noise, high frequency and high power stability oscillators. Gunn diodes don't qualify for any of that.

Google W1Ghz Personal Beacon.
It is a harmonic upconverter   using pipe cap filters. With a doubler between it and the PCRO you would have a nice, clean source with moderate phase noise and excellent frequency stability.

Your also looking for 10 Ghz transverter designs.

When you said you want to apply circular polarization to a mixer, that tells me you have much, much, to learn.

In that case find copies of the various older RSGB Microwave Handbooks.

Just email Macom and ask what they have in stock

Good Luck.

Steve

[Ben321]

When you said you want to apply circular polarization to a mixer, that tells me you have much, much, to learn.

I didn't say that I want to apply circular polarization. I actually said I want to avoid it, because the LNB I'd be using to receive the signal would be a cheap linear-polarized Ku band LNB.

Also this https://www.ebay.com/itm/225229582241?hash=item3470baafa1:g:G74AAOSwnCtjXpcT microwave oscilator is too low of a frequency. The cheap Ku band free-to-air TV satellite LNBs that you can find on Amazon (and which I would use on the receiving end of my system) have a relatively narrow bandwidth of about 1GHz. For example, this Ku band receiving LNB https://www.amazon.com/dp/B09JZDFMB5 has a pass band from 10.7GHz to 11.7GHz. The transmitter you showed me was well under 10GHz. Unfortunately that's too low of a frequency for the Ku band receiving LNB.

Also note that there's a discrepency between Ku band used for free-to-air TV satellites, and the Ku band used by other equipment. The Ku low band in satellite TV terminology is 10.7GHz to 11.7GHz, and the Ku high band is 11.7GHz to 12.75GHz (high band is selected by sending a 0.5V p-p square wave tone to the LNB input along with the DC power, as otherwise it uses only the low band). However Ku band in other uses seems to be a much higher frequency, from 12.4 to 18GHz according to https://www.microsemi.com/product-directory/rf-discretes/3305-diodes-gunn#selection-table. And the TV satellite signals are in what is actually called the X band 8GHz to 12.4GHz, according to that webpage. So there seems to be a bit of confusion there about the names of the microwave bands.

[LaserSteve]

DID you even look at the Ebay link to the brick?  115$ Dollars and shipping and you have far more power then you need to drive a high level mixer. Same Brick would get you on the order of 30-50 miles of transmit range on a good day with a 19 inch satellite dish.  With five to six orders of magnitude better frequency  stability then a Gunn Diode.

[miken]

I remember seeing a box full of Gunn diode assemblies at a ham swap fest, maybe five years, ago so they're still out there.

[Ben321]

DID you even look at the Ebay link to the brick?  115$ Dollars and shipping and you have far more power then you need to drive a high level mixer. Same Brick would get you on the order of 30-50 miles of transmit range on a good day with a 19 inch satellite dish.  With five to six orders of magnitude better frequency  stability then a Gunn Diode.

You mean this one? https://www.ebay.com/itm/225229582241?hash=item3470baafa1:g:G74AAOSwnCtjXpcT
This one has an output frequency that is too low for receiving with a Ku band satellite TV LNB. Also why would I need to power a mixer with it? Wouldn't it be simpler to modulate the voltage powering the device, to make amplitude modulated microwaves. As microwave components are expensive, I'm trying to keep most of the system at lower RF frequencies, with up-conversion to microwaves being the last step (and directly transmitting from the microwave generator module, without further microwave optics or electronics).

I'm picturing a circuit like this for the transmitter.

[Gyro]

I remember seeing a box full of Gunn diode assemblies at a ham swap fest, maybe five years, ago so they're still out there.

Irrc, Gunn diodes are pretty ESD sensitive (much more so when out of the assembies). From memory, the assemblies were supplied with shorting links.

[Hamelec]

the shorts were for the mixer diode.

[Ben321]

Now the VCO at their basis is monolithic, like https://www.mouser.it/ProductDetail/Analog-Devices/HMC733LC4BTR?qs=pceeu5JH%2FH%2Ffv1ib8cM6Og%3D%3D

I checked that out. It appears unfortunately that only one company is making those microwave VCOs, or at least only one company is selling them through Mouser. This means little competition, and thus high prices. For just one of those chips, it costs over $300. And that is literally just for the chip, not a complete benchtop device (like a microwave function generator), or even a module/subsystem of such a device. Do you know of any other company that makes similar chips, preferably for a lower price?

[Marco]

Have you seen JBeale's threads and the links therein? There's a lot of cheap modules coming out of China at the moment working at the mmWave ISM bands.

https://www.eevblog.com/forum/chat/opinions-on-reasonably-cheap-mmwave-radar-module/msg3565114/
https://www.eevblog.com/forum/projects/another-small-cheap-24ghz-doppler-radar-but-more-feature-full/

[Ben321]

Have you seen JBeale's threads and the links therein? There's a lot of cheap modules coming out of China at the moment working at the mmWave ISM bands.

https://www.eevblog.com/forum/chat/opinions-on-reasonably-cheap-mmwave-radar-module/msg3565114/
https://www.eevblog.com/forum/projects/another-small-cheap-24ghz-doppler-radar-but-more-feature-full/

Unfortunately those transmitters you linked to threads about all are in the 24GHz band. I'm specifically looking for a transmitter between 10.7GHz and 11.7GHz, due to my intended setup on the receiving end.

[LaserSteve]

OK, having checked Ebay for Microwave Oscillator, there are plenty ol units for 10700 or thereabouts.  Bricks tune up or down in around 80-120 Mhz steps, depending on the crystal installed or in some cases an external reference input.  My bricks for example tune in 106.5 Mhz steps.  Pyro-Joe can certainly sell you a brick with a commercial band crystal in it and retuned for 10.7 or up. All he has to do is retune the ones he tuned for the ham band back to commercial band.

SEE ATTACHED PAPERS FOR MECHANICAL DESIGN:

Try this search you will be pleasantly surprised:

https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2380057.m570.l1313&_nkw=gunn+diode+x-band&_sacat=0

5 sellers...

Oh Canada, Thy New and Glorious Land, These need a sheet metal iris made from 0.040" brass added, and away you go...

https://www.ebay.com/itm/275495176736?mkevt=1&mkcid=1&mkrid=711-53200-19255-0&campid=5338722076&customid=&toolid=10050

These are shockingly accurate, Add WR90 detector, this is the microwave version of a Grid Dip Meter:

https://www.ebay.com/itm/114835730792?epid=1917956329&hash=item1abcbe3168:g:-U0AAOSwjJxguAAN&amdata=enc%3AAQAHAAAAsG0TKp02fqXzRw4yopxKkivdVLCrPSMtYdzkDp3Vrwnw4iuV7TYv1uxFtzMAi7mrwby%2BnHrwqUmyumL5%2BeoaUrW7twdudbWTIyS8rnMm9A7f%2FPzaeAym14N%2FEtpRreTVnBC%2FkvNeV4%2FP1PI2%2BCAT5fJ8hxAfxH1V4U7f7akagtxiguHFM9nPzWLhNoU5e7jxolUI%2BN7CnjQi43mzy03nqQwXqrtrirj%2FRGzXdTSMdKtk%7Ctkp%3ABk9SR4zc1bKJYQ

Unless your "treez|", you should be delighted.

Something Russian:



Overkill:



FM Video TX:



Totally Daz-eling...



Revenge of the Wavemeter:



See attached PAPERS

Good tutorial:





Steve

[buta]

I'm specifically looking for a transmitter between 10.7GHz and 11.7GHz, due to my intended setup on the receiving end.

Not sure your area uses Ku band for satellite TV or not.
IF your area uses Ku band for satellite TV, do you allow to transmit in 10.7 GHz to 11.7 GHz?

[Ben321]

I'm specifically looking for a transmitter between 10.7GHz and 11.7GHz, due to my intended setup on the receiving end.

Not sure your area uses Ku band for satellite TV or not.
IF your area uses Ku band for satellite TV, do you allow to transmit in 10.7 GHz to 11.7 GHz?

I live in the US. Here, we don't have free-to-air satellite TV. In the US, satellite is all paid TV. And to avoid being able to be illegally received/pirated using standard free-to-air satellite LNBs, paid satellite TV signals are transmitted with circular polarization. Normal free-to-air satellite TV (used in some other countries) uses linear polarized LNBs. In both cases, it's in the KU band, but with different polarization, it's unlikely to interfere. Those cheap KU band LNBs you can buy on Amazon or Ebay are linear LNBs for receiving free-to-air satellite signals found elsewhere in the world, but in the US, they serve very little practical use.

However they can be used for a very cheap microwave receiver for experiments, provided that you also have a signal source that operates in the same frequency. And that's the hard part, is getting a transmitter for this frequency band. KU band LNBs typically operate in one of 2 polarizations, horizontal and vertical, which is selected by the voltage used to power the device. If you power it at 12V you get one polarization. If you power it at 18V you get polarization 90deg to what you get at 12V. This is to prevent needing to manually rotate the LNB by 90 deg to switch polarization. The frequency band is selected by the presence (or lack of presence) of a 22kHz tone. Without the tone, the KU low band is used, between about 10.7GHz and 11.7GHz. With the tone, the frequency band you get is the KU high band, between about 11.7GHz and 12.7GHz. The specs for the tone is a 1Vp-p 22kHz square wave.

For simplicity of setup, I'm not going to figure out how to inject this audio tone onto the coax. It's going to be difficult enough to build a Bias-T adapter to inject DC power on the coax, and then extract the RF off of the coax. So without the tone, I'm limited to receiving the KU low band between 10.7GHz and 11.7KHz.

Now with that in mind, I'm going to just need to figure out how to generate a microwave signal within the range of 10.7GHz and 11.7GHz, so that it will be receivable by a cheap KU band linear LNB that I can buy on Amazon.

[james_s]

Unfortunately those are listed as new-old-stock. Does nobody make gunn diodes anymore? Has some new technology been introduced that fills the same role, that I'm not aware of?

So? What's wrong with new old stock? They don't expire.

[ejeffrey]

I live in the US. Here, we don't have free-to-air satellite TV. In the US, satellite is all paid TV. And to avoid being able to be illegally received/pirated using standard free-to-air satellite LNBs, paid satellite TV signals are transmitted with circular polarization. Normal free-to-air satellite TV (used in some other countries) uses linear polarized LNBs. In both cases, it's in the KU band, but with different polarization, it's unlikely to interfere.

Circular polarization overlaps 50% with linear and will definitely cause interference with only 3dB rejection.   Polarization is also not really a means of piracy prevention, it's mostly chosen for transmission and scattering characteristics.  Afaik all current paid satellite TV is digital and encrypted, but older analog systems used scrambling.  The only way circular polarization could be considered to prevent reception by unapproved devices is that most satellite transmissions use both polarizations for different channels.  A linearly polarized receiver will see both equally and possibly be unable to resolve either.  But that only works because the signals are close to the same strength due to originating at the same transmitter and following the same path.