Sure, most capacitor companies have dielectrics with Er much higher than 400.  How do you think they cram 0.1 uF into an 0805 footprint.  Getting them to make a small batch resonator shape for you will be the trick!

Call these guys and see what they say:
http://www.dilabs.com/

The temperature characteristics of the dielectrics used for 100nF 0805 caps are pretty dire...

biff44: yes, you are right, there are materials used for capacitors, but unfortunately they have a far too high temperature dependence in eps_r (not so much in expansion). So the quest is for a material with high eps_r and a temperature compensation/stability which is unmagnetic.

Any more suggestions or ideas?

What do you care?  Sweep the frequency and find the new resonance frequency.

biff44: Indeen, I do care since I use high power pulses up to 1kW to excite the resonance. That means that the resonator gets warm and if you observe the reflection during (!) the pulse (pulselength ~ 10ms and duty cycle 1s), the resonators frequency will drift away.

I got a prototype of a non-magnetic material matching my needs with respect to eps_r and Q. Again, the temperature dependence of eps_r and the accompanied frequency shift of ~1600ppm/°K (measured in the ambient temp.range) is the main problem. And the ceramic company I got it from is hesitating to develop a temperature compensated ceramic for me, since I only need very small amounts.

So here is my question again: Does anybody know about a material having the characteristics I mentioned above? The temperature compensation might be around a specified themperature close to ambient (23°C).

Well....you get a number of different materials with differeing expansion and Er temp coefficients and play around on a good emag 3d program until you come up with the dimensions of a sandwich that is temperature stable.

And of course you do realize that the resonator absorbs more heat, and therefore has more of a temp co. problem, if it has low Q.

At lower Er's we often do this by placing a barium tetratitanate puck on top of a Saphire ring.

By sheer coincidence, I found an apnote catalogue for Trans-Tech, who apparently do, or did, dielectric resonators.

They are/were part of Alpha Industries.

www.alphaind.com

email: transtech at alphaind dot com

(address is bowdlerised, I'm sure you can figure it out...).

Can you say more about the resonator and it's use.
I'm familiar with MRI's, GE's primarily, and their typical frequency is 63 Mhz, so I'm surprised at your higher frequency which doesn't penetrate the body.

Sounds like you need to find another dielectric to make your own temp compensated material.

Have you thought of artificial dielectrics which use metal wires, I'm not sure of the range of values available though. Maybe the addition of a partial artificial dielectric to your dielectric could help compensate the temp drift.

kch

Higgler: The resonator is used not for imaging in bodies, instead for high field solids NMR. Penetration depth, at least due to parasitic e-field components, is the problem: Compared with a classical solenoid coil a DR in TE01d-mode has a reduced electrical field inside the sample volume and therefore even very salty (lossy) samples can be used (given of course, that an apropriate material is available).

Probewizard,
I was reading about high fields solid NMRs in Canada and Florida, the word Bruker kept popping up. Are you part of one of these?

Seems like 900 Mhz is on the cutting edge frequency wise.

This was a list of instrument capabilities printed below from the Florida sight, even though I see the words, their slightly Greekish still.
I assume your resonator is placed at the center of the instrument, maybe next to the object being measured, to create and then listen to the rotational shift of the electrons in different media. Or does the test media get embedded in your resonator? Seems like a high Q resonator would have its' Q lowered if you placed an item under test too close to it though.
I would think you'd use two resonator coils placed on each side of the test sample, fed out of phase from a 900 Mhz balun, with each output having a ?500 watt amp (purely a guess of course). That way you could get 4x the field at the test sample center.

I'm just a curious monkey,
thanks for your patience,

KCH


• Microimaging for ultrastructural characterization and for monitoring diffusional anisotropy.
• High resolution spectroscopy of quadrupolar nuclei for materials characterization and imaging of quadrupolar nuclei for ultrastructural characterization.
• 1H and 19F magic angle spinning solid state NMR of biological and inorganic materials.
• Structural and dynamic characterization of membrane proteins and other macromolecular structures through solid state and solution NMR.
• Characterization of nascent structure in weakly structured macromolecular systems.

Hi  Higgler,

sorry for answering that late.... (couple of months now
To answer the first question, no I am no part of bruker, although we are working with their spectrometers.

NMR spectroscopy is a method to investigate a given sample with respect to the chemical properties (chemical bound, structure, molecular motion, etc, etc) and there a plenty of applications for chemicists and physicists...
In the experiment, the sample is placed in a static magnetic field and is excited by a varying magnetic field (B1) which is generated by an RF resonator.... and here we are....

Since the resonator is somehow disturbed by the sample a lot of care has to be taken when designing it and indeed specialized resonators exists for almost every thinkable application. The idea of using a dielectric resonator is not new but to my knowledge nobody has used it up to now for the NMR frequency range....

Hope that short explanation somewhat made my questions a bit more clear. Although I currently have no time to continue the project, I am still looking for an appropriate material.... so if anybody has a hint...

best wishes
Alex