Nuclear Quadrupole Resonance (NQR) - instrument assembly

Building a drug and bomb detector? Cool. What size (volume) is your sample? That determines the RF power required. Digging ideas out of patents is useful: See references and citations for additional related patents.

In a past life, I helped design and build an early NMR FET (VMOS) power amplifier that eventually ended up in a Toshiba/Diasonics NMR system. The key specification was very low IMD (intermod distortion), which was not easily achievable with communications grade power amplifiers. At the time, it was difficult to see the 3rd order IMD during a 2 tone test on a spectrum analyzer. Any distortion in the excitation waveform and you'll have spurious frequencies to deal with. Fortunately, the state of the art for clean signal sources, power amps and test equipment have improved, so you should have fewer problems today:

The above patent offers some useful clues as to what you might need:

The power requirements of the invention are generally proportional to the detection coil volume. An explosives scanner for mail packages with a 25 liter detector coil volume might have an RF power amplifier rated at about 25 Watts, peak value, for example. The amplifier produces a uniform RF field of about 1 gauss over the entire 25-liter volume. In other applications, such as in narcotics detection, the RF field may be greater than this value. For airline baggage, an explosives detection head of about 300 liters (10 ft3) volume within the coil requires a 1 to 2 KW RF power amplifier. These parameters are provided for reference purposes and are not meant to define or limit the actual characteristics of a practical NQR system.

The RF excitation pulses are fed from amplifier 24 into detection head 33, the operation of which will be discussed below. After the sample in the detection head has been excited by the RF pulse, a short RF coil "ring-down" or dead time occurs, during which the receiver is "deaf," before sensing occurs. This ring-down time

signals and the response is amplified by low-noise, high-gain preamplifiers 25 having a gain of 20 to 30 dB, and a noise figure of 1 to 2 dB. Examples of such preamplifiers are Anzac Model AM-110 and Mini-Circuits Model ZFL-500 LNS.

There are also clues as to the operational specifications:

For example, RDX-based plastic explosives have a resonant frequency of approximately 3.410 MHz while PETN-based plastic explosives have a resonant frequency of approximately 890 KHz. The excitation source is fed into amplifier 24 of sufficient power rating to generate about 1 gauss of RF magnetic field within the coil. The excitation frequency need not be exactly the same as the target substance NQR frequency but it should be within about 500-1000 Hz. The RF excitation for NQR detection could be a single pulse of

Such a single pulse could cause an NQR return, but the nuclei may not have reached a steady state of precess so the NQR return might not be sufficiently strong to be detectable or useful.

Also, please make an effort to shield the sample enclosure. Not only does that reduce stray pickup from other local transmitters, but also prevents your machine from becoming a jammer.

Good luck. Sounds like an interesting project.

Yet another interesting patent:

"Quadrupole resonance using narrowband probes and continuous wave excitation" This one looked interesting until I noticed that it uses high temperature superconducting wire and liquid nitrogen to increase the Q of the detection coil. An ordinary copper coil at room temperature might have a loaded Q of about 100. This thing claims a Q of over

These guys:

... seem to avoid that problem and most of the other headaches by placing t he sample between the conductors in a balanced transmission line. They loo k for stimulated emission rather than spontaneous emission when the sample molecules undergo relaxation. This enables coherent detection and (optiona l) averaging, rather than having to rely on frequency-selective, SNR-limite d detection of uncorrelated spontaneous responses.

So, apparently you just hook up a network analyzer to a balun and use it to drive a piece of ladder line that's terminated at the far end. Then you p ut a directional coupler in one leg of the line, run its output through an LNA, compare S21 before and after insertion of the sample between the trans mission-line conductors, and collect big bucks from your buddies at the TSA .

Hard to believe it's that easy/obvious, or everyone would already do it thi s way, wouldn't they?

-- john, KE5FX

the sample between the conductors in a balanced transmission line. They l ook for stimulated emission rather than spontaneous emission when the sampl e molecules undergo relaxation. This enables coherent detection and (optio nal) averaging, rather than having to rely on frequency-selective, SNR-limi ted detection of uncorrelated spontaneous responses.

to drive a piece of ladder line that's terminated at the far end. Then you put a directional coupler in one leg of the line, run its output through a n LNA, compare S21 before and after insertion of the sample between the tra nsmission-line conductors, and collect big bucks from your buddies at the T SA.

his > way, wouldn't they?

Every good idea is obvious after somebody has invented it.

One of the stronger principles of patent law is that everything is obvious to the Supreme Court ...

Thank you guys for your replies. I really appreciate that