Alumina Substrate

I think you're referring to a dielectric resonator. Those work from about 500Mhz to maybe 30 GHz. Typical dielectric constant is the usual 10, although some exotic ceramics go up to 45. They're really "cool" for ultra low loss cryogenic cavity resonators. I don't think I can give a sane explanation of how they work. Maybe later.

The others devices you mentioned are possibly SAW (surface acoustic wave) filters and oscillators, which owe their small size to the relatively slow propagation time of "sound" waves across the substrate.

Anway, the OP doesn't need a 0.25" thick chunk of ceramic. A thin layer, sufficient to act as a substrate for the circuitry is all that's required. Back that up with an aluminum heat sink (RT/Duriod) and it's MUCH easier to machine.

Don't worry. When we get going on optically coupled circuit elements, the need for switches and modulators will need PLZT devices. Patience.

--
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558

Engineer at hybrid mfg handed me a 0.7mm srap piece. "Here, thick enough, you won't be able to break that with your fingers". Took it between thumb and index fingers ... CRUNCH! "Ok, you guys get the 1mm then ..."

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Every other manufacturer is cheaper than Rogers. Great product, but awful pricing.

In about 1970, I was working for Alpha Electronics in Stanton CA. We did hybrids for CTCSS tone boards and the first LED digital watches. At first, we trimmed values using an S.S. White sand blaster. Later, we built our own gas laser. Nobody had a clue what they were doing so we proceded to duplicate every mistake possible. Full blast laser (10 watts) on a ceramic substrate resulted in a fragmentation bomb. No air flow near the optics resulted in pitting the optics. The front surface silvered mirrors got rotted by the CO2 turning to carbolic acid. Burning a hole in the x-y table was standard practice. A good time was had by all involved.

Surface mount killed hybrids. The big advantage I found in hybrids was building circuits with resistors that would all track together. I think (not sure) that less than 1ppm tracking over the entire mil spec temp range was possible.

--
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558

In Europe we never built our own. In a particularly difficult case we even had an engineer from the laser mfg in Oregon fly in.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Yep. I used to buy them in Germany because of better pricing. Looks like those:

Sometimes you can buy them in the HW stores over there. But after the slide of the USD I might not do that anymore next time.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

I'm a fan of aluminum nitride; thermal conductivity is close to BeO.

John

No, I meant the little ceramic puck resonators...

Neither the coaxial nor the puck things are acoustic/mechanical resonators.

John

Certainly. Regular drills, and core drills for big holes. Still slow and needs fluid. Not really expensive, just slow.

John

We had some good help. There was a small laser company nearby that did most of the engineering. They also supplied us with parts and pieces. As long as we didn't burn too much of their time, we could get fairly good advice. Our chief engineer was self taught and could literally do anything. In 1970, high power lasers (suitable for scribing hybrids) were few and very expensive. So, we built our own laser and hi-v power supply. We also mixed our own inks, did our own silk screening, and baked the results. Trimming was automatic, but with an analog, not a digital computer.

The laser wasn't too difficult to make work. The problem was keeping it working. The gasses would leak and/or diffuse through the glass. Keeping the mixture correct was an ordeal. Under the high power, various parts would deteriorate or move, which would wreck the focus. We used microscope optics, which were easy to destroy and replace. Consistancy was also difficult to achieve. It was not unusual for us to run a lot of hybrids well into the night, simply because stopping meant that we would have to re-align everything from scratch the next morning.

There was also a small problem. The laser was mounted on a concrete block wall with various plumbing parts forming the optical bench. On the other side of the wall were some railroad tracks. The trains were never on schedule, so I built a train detector, that would warn the laser trimmer operators that a train was coming and to shut down. If they forgot, the hybrid would have a perfect seismograph waveform burned into the resistor material. Someone gave one of these seismo scribed resistors to a vendor, who promptly leaked to the competition that the unique laser scribe pattern was the real secret to maintaining constant frequency accuracy (twin-T filter) over a very wide temp range. Over the years, we got pumped for information by various spies as to how this was achieved. I suspect there was some poor hybrid specialist, that tore his hear out trying to reverse engineer the secret. Heh-heh.

I didn't have much to do with the hybrid line or production. Mostly, I ran the RF and radio parts of the business and helped clean up the mess after the latest laser initiated disaster.

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558            jeffl@comix.santa-cruz.ca.us
# http://802.11junk.com               jeffl@cruzio.com
# http://www.LearnByDestroying.com               AE6KS

Egads. Dielectric constant of 89. I didn't know they went that high.

Highest dielectric constant mentioned is 38.

Y'er right. Sorry about the bad guess.

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558            jeffl@comix.santa-cruz.ca.us
# http://802.11junk.com               jeffl@cruzio.com
# http://www.LearnByDestroying.com               AE6KS

Sand blasters are cheap... particularly when (in my case) the employees were Papago Indians.

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
         America: Land of the Free, Because of the Brave

Hmm, so that's like saying high K insulators act like good conductors to HF (e.g. ceramic bypasses short RF to GND), except applying it to travelling waves instead? And of course, that "conduction" comes with a substantial phase shift, and because of the high permittivity, a high index of refraction (= low speed of light), making them nice and small too.

Do they have to be canned or do they work in free space? Seems to me Q would roughly correspond to K if unshielded, since that's about how much would reflect around inside. Assuming an optical-ish model, which of course is probably completely wrong...

Tim

-- Deep Fryer: A very philosophical monk. Website @

I don't know a lot about the puck things; I've only used the coaxial ceramic resonators, in the 600 MHz range.

We make one delay generator that uses a coaxial resonator as a startable oscillator; when we get a customer trigger, we instantly start a 600 MHz oscillator, and count that to make delays. It works more like a transmission-line oscillator than like an LC.

I have seen the aspirin-type pucks just sitting on a pc board, with some nearby traces somehow coupling into the puck, with a transistor to make an oscillator. That's essentially free-space. I think the resonance is quasi-optical, in that the electric field waves are trapped inside the puck by the equivalent of total internal reflection, possibly in a whispering-gallery mode. Some interested person could look it up and post some links.

I think ceramic-filled cavities exist, too, but the pucks aren't usually confined. There are certainly liquid-filled cavities in use.

Hey, you could tune a cavity by pouring sand or something into it.

John

I have seen resonators that had so much in nicotine deposits and other gunk on it that I was very surprise they still worked. Probably it's better these days since people aren't allowed to smoke in labs anymore.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Of course, we've been using ~100-layer alumina/refractory metal bricks (more recently Cu/glass ceramic with Cu/polyimide on top) to build computers for almost 30 years now...ever since the 3033 ;)

Cheers,

Phil Hobbs

But they sure couldn't stomach a 170V/3A pulse going clear across the alumina and listen to uV-level signals a microsecond later :-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Sure they could...a microsecond after you did that to it, the brick would be _really_ quiet. Doing it twice would be the trick.

Cheers,

Phil Hobbs

The other type of machinable ceramic is Macor:

AFIK this is the only one which is easily machinable in the full hard condition rather than bisque or green, but someone else may have something similar by now as it has been around since about 1980. The drawbacks are that the full hard condition is not very hard or strong, and the electrical properties are not the same as alumina.

It probably would also instantly fix any potential EMI issues :-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.