>It seems a number of NMR labs like to stick a digital oscilloscope
>online (so to speak) by using a line section/plug-in element (from Bird
>Electronic Corp. USA, for example) sitting in the transmitter line. My
>understanding is that this component picks up the reflected power from
>the probe and thereby visualize the input pulses and waveforms on the
>scope, giving a real time check on what the pulse programmer is
>producing. I understand the line section/plug-in element hardly affects
>the RF power at all.
>
A digital oscilloscope is probably not the right device to use to visualize
a pulsed RF envelope. There are as sorts of ways to produce alias effects.
I prefer an analog scope for this purpose even though the persistence of a
standard phosphor screen is a bit too short at low RF pulse rates.
The pick-up element in a Bird Wattmeter samples the traveling RF wave in
either direction depending on its orientation. RF currents are induced on
the surface of the line portion of the Bird unit. These currents are
detected by a diode which drives a DC meter in the standard Bird unit. The
diode is not DC biased, so the RF power must be sufficient to cause it to
conduct. The loss from this effect is measurable (about 2%) There is
another version that has no diode. That sensor is intended to drive a scope
directly. The version without a diode must be terminated in 50 ohms. This
load will dissipate a similar percentage of power from the RF line. There
is also an insertion loss due to the connectors. Overall, you might see a
5% loss for either device.
>I have seen this in operation in a couple of laboratories - it is really
>a nice thing to have to confirm that the pulse program is really doing
>the right thing - but it's not a cheap thing to do if you don't have a
>high frequency digital scope (eg. 500MHz or 1GHz). Can it be done less
>expensively if all you want to visualize is the presence or absence of a
>pulse rather than the high frequency oscillations within the pulse? What
>I have in mind (and this may be the mad part) is that you could
>integrate or rectify the input to a (lower frequency) scope and thereby
>simply observe the pulse envelope? If so, what sort of device could be
>to achieve this the integration step? I am obviously looking for an
>off-the-shelf component or (very) simple homemade device design.
>
There are many vendors for RF detectors. We get ours from JFW in
Indianapolis. These devices are intended to detect an RF envelope. They
are not simple diodes in a coaxial connector. The circuit includes a 50 ohm
termination, a bias resistor, and an integrating capacitor on its output
side. The capacitor will limit the rise time of the device, even when the
assembly is terminated at low impedance. It may be more effective to build
a discreet circuit in a small cast Pomona or Eddystone box. Then, you can
choose the capacitor value and terminating impedance to give a true
representation of the RF pulse. This sort of device is best operated
through an in-line directional coupler that isolates the RF line by about 30
db. If the diode is not DC biased, and there is usually a way to do this,
then the RF level must be sufficient to break over the device. The ARRL
Handbook has a circuit example that uses a transistor in a similar mode.
The transistor's input junction is what actually detects the RF envelope.
The transistor is biased to be able to detect very low levels of input RF
and amplify the resultant DC signal.
--Bob Santini
Robert E. Santini
Department of Chemistry
Purdue University
W. Lafayette, IN 47907-1393
santini@chem.purdue.edu
Voice: (317) 494-5227
Facsimile: (317) 494-0239