I have been interested in Rain Scatter (RS) on the microwave bands for many years. Through tests with DL6NCI, DD7MH, HG1YA, and many others, I have confirmed that 3cms is not the only band suitable for RS. RS is also possible on 6cms (very good), 13cms (just about), and 24GHz (very rarely). 9cms is not available in Austria so I have no experience of that band.
To make RS contacts, you must first find a suitable cloud that can act as a scatter point. The finding of these clouds has until recently been mainly through listening for beacons. This only enables the direction of the scatter point to be found, but not its distance, therefore it is not possible to determine the position of the scatter point.
This brings me to my idea, to receive my own reflected signals, and to measure the elapsed time. I have had this idea for a number of years but had not acted upon it, the idea was recently re-activated through an article by LX1DU in DUBUS. I required the help of many of my fellow radio amateurs with the design and implementation of the system. Foremost is my friend Erwin (OE5UXL) who developed circuits for individual components, Michael (DB6NT) for his excellent modules and amplifiers. I had many QSOs with Ulrich (DG2MF) in which we discussed the fundamentals of the system. Ludwig (DC8NV) provided the most significant component, a circulator. Dieter (DL3NQ) gave me many important tips based on his large experience. Ferdi (DC8EC), Erhard (DC4RH), and many others helped using their specialised knowledge to help me realise this project. Based on my experience of RS and the availability of cheap components I decided to operate my Rain Scatter Detector (RSD) on 5.7GHz.
I needed a number of units for the system. A pulse generator to produce the transmitted pulse and receiver phase, a transmitter oscillator (120MHz), a x48 multiplier to produce 5760MHz at 100mw, a 5W amplifier, a converter from 5760MHz to 144MHz, a 144MHz receiver with an S-meter output, and finally a circulator. An oscilloscope is used to display the results.
The pulse generator produces a 30µs pulse with a frequency of 375Hz. This controls the transmitter which produces a 30µs carrier, 375 times a second. In order to not block the receiver with leaked RF from the transmitter, a number of stages are switched, a buffer after the oscillator and three multiplier stages in the transmit chain. The transmitted signal is amplified by a two-stage amplifier up to the 5W level, and through a circulator to the antenna (3m dish). A very good match to the antenna is a basic requirement for this to function, otherwise any reflected power would flow via the circulator to the receiver input, and the input transistor destroyed. The isolation between the transmitter and receiver ports is 33db, approximately 3mW from my transmitter appears at the receiver input. During tests I confirmed that the input transistor (NE325) could survive this power level at its input.
The antenna radiates the transmitted pulse and when no reflector is available, nothing is received back. A small amount of the transmitted signal is received via the circulator and this causes the oscilloscope to display a large deflection in the Y-axis for the duration of the pulse.
Suppose that within 300kms of my antenna is a good reflection point, a proportion of the transmitted pulse is reflected, and if strong enough, will appear on the oscilloscope as a deflection on the Y-axis at a later point in time indicated by the X-axis. This signal is the start of the time measurements on the X-axis.
The deflection in the Y-axis is proportional to the strength of the reflected signal, and the distance to the reflection point is calculated by measuring the time difference on the X-axis between the pulse being transmitted and the reflected signal being received. An elapsed time of 1ms is equal to a distance of 300kms, which is produced by 150kms there and 150kms return, indicating 150km to the reflection point. I have the X-axis time base set to 0.2 ms/div so a distance of 300kms is easily seen.
To date (beginning March) there have been no thunderstorms, and that is why I still do not have much experience with how effective my RSD is at detecting RS. Nevertheless, to the south, in the Alps I can detect the Dachstein at a distance of approximately 115kms with a 20 - 30db reflection loss. Between 80 to 100kms distance I can detect many weak reflections.
The RSD first went into service in mid February, could I by easy raised present widen, at an azimuth of 271 degrees and at a distance of approximately 370kms I receive a weak reflection, this is a hill in the Swabian mountains. To the north I can detect a large number of hills at around 200kms distance.
The next few months will show how close the theory and practice are.
The only problem with the RSD described here, is that it only works when using a large antenna (3m dish). If I had only a 1m dish, I would need a power of 500W to get the same strength from the reflections. With a 70cms dish I would need 1kW.