One of the problems will be to protect the Spectrum Analyzer / Scope inputs during the pulse, but have it
fully opened for "listening" for the response.
My approach to such isolation is keeping the Tx coil perpendicular to the Rx coil.
First, both coils should be tuned for MINIMUM reflection coefficient (or return loss or S11 or VSWR) at the target frequency. Using Litz wire for the windings is highly recommended.
Also, the Rx coil
cannot ring (self-oscillate) at the target frequency, or you will not be able to distinguish the LC ringing from the nuclear resonance - this is very important !
When positioning the TX and Rx coil in respect to each other, their Mutual Coupling should be minimized !
This can be done by twisting the coils any way you can for minimum mutual coupling and making a fixture that keeps the coils positioned appropriately with respect to each other. This also means that the coils cannot touch each other in order to avoid capacitive coupling.
Now, winding perpendicular windings on a toroidal core requires some imagination, but it is easy to come up with a way to do it on a cylindrical or square rod (a sample or sample holder).
I admit I was never able to achieve a sharp resonance with a toroidal core, but I was able to do so with a fine iron powder (easy to get) packed in a glass test tube which was purged with Argon (from a TIG welder) and bulk heated to 500ºC and spot melted shut.
After the test tube was sealed with hot Argon inside, it was heated to 400ºC for one day, wrapped in fiberglass isolation and allowed to cool slowly. This annealing process makes the resonance very sharp at 50kHz FWHM.
If you let the iron powder oxidize, then the iron oxides will resonate much higher than 45.5MHz, so be prepared to search up to 48Mhz ...and if the iron gets sulfated - even up to 76MHz.
Below is a diagram of the NMR arrangement, with the metal powder held in a glass tube (light blue).
The big magnets with the steel yoke are not needed when resonating iron because iron has a natural -33T internal field, thus a big homogeneous field is not absolutely needed... but it does not hurt (a 750mT field helps to align the ferromagnetic domains in one direction).
Also, iron features a 10
4 enhancement factor of the nuclear signal, which more than makes up for the 2% isotopic abundance of the susceptible
57Fe.
Last but not least, Solid State RF switches help to protect the sensitive receivers during the Tx pulse and an RF switch can also disconnect the Tx winding during reception, so the impedance of the disabled Tx generator does not absorb the RF energy back and leaves more for the Rx coil to pick up.
Author
Topic: Itsu's workbench / placeholder. (Read 137338 times)
