Fast Rise Time InP Oscillators Employing Cavity Stabilisation and Pulsed Priming Techniques

Abstract

INTRODUCTIONThis paper describes the design of a fast rise time pulsed InP oscillator for operation at J-band. R.F. circuit techniques are described which simultaneously satisfy the conflicting requirements of a fast R.F. rise time and good frequency stability. R.F. CIRJUIT DESIGNThe R.F. circuit (see Figure 1) consists of a waveguide oscillator of the reaction cavity type. A single InP device is mnounted at the waveguide centroid on cylindrical posts. Experinentally it is observed that the loaded Q factor (QL) of this type of oscillator must be less than approximately 150 to achieve R.F. rise times faster than 5 nS, but the chirp is then greater than the required 3 MHz/300 nS. Although other methods of chirp reduction have been reported (1) in order to achieve the required stability over all environmental conditions cavity stabilisation techniques must be used. X Referring to Figure 1, a resonant cavity separated from the device by g/2 is iris-coupled in the waveguide broad-wall in the absorption mode. This method of coupling does not affect the basic oscillator rise time (2) since at the instant of switch-on the internal field in the resonant cavity is not established and no energy is stored. However, after a period of approximately three times the Q limited rise time of the resonant cavity, the amplitude of the internal field has reached its steady state value and the oscillator stability is then controlled by the resonant cavity. The design of the resonant cavity is treated in detail below but in outline the unloaded Q (QO) and the coupling iris size must be chosen such that the cavity QL is high enough to ensure adequate stability yet low enough such that its time constant is short compared with the pulse length. The oscillator is frequency stabilised over temperature as described by Myers and §tevens [3). The dF/dT of the oscillator, normally of the order -1 MHz/ C, is reduced by means of the compensatos to match or be less than that of the resonant cavity, typically -300 KHz/ C. Using this technique, chirp reduction has been maintained over the operational temperature range of -40 to +700C. The oscillator also incorporates a GaAs Schottky barrier detector diode, which provides the overall system with a monitor pulse within a few nanoseconds of the start of the R.F. pulse. The oscillator is fabricated in WG19 and fitted with a waveguide transformer feeding a WG18 isolator.The high efficiency InP pulsed device exhibits both a large delay between the application of the bias voltage and the start of the R.F. oscillations and an uncertainty in the magnitude of this delay. These are commonly referred to as switch-on delay and jitter respectively; measure-558 ments over a temperature range are shown for a typical device in Figure 2.The performance is better at higher temperatures but operation at peak efficiency is almost always accompanied by poor switch-on characteristics irrespective of ambient temperature. One solution to this problem is to employ phase priming techniques in which a small amount of power at a...

A High Efficiency InP Transferred Electron Device for High Peak Power, High Mean Power Sources in J-Band

“…Applications for this type of InP TED have in the past been exclusively for use in low duty cycle systems (<2%) (4,5), where the average power dissipation is small. In these cases, the CW thermal impedance of the device is of little consequence but if high duty cycle operation-is required care must be taken to optimise the die attachment to the device package.…”