Intended to simplify circuit design and save space, Ericsson Components is the first company to develop a range of ýhybridý RF power transistors. The devices feature built-in 50 Ohm matching networks for input and output impedances, along with bias circuitry and temperature compensation.

With an output power of 12W, Ericssonýs PTF31042 is for use in the 1.9 to 2.0GHz PCS frequency band. With a footprint of less than 25x21mm, the HIT (High Impedance Technology) device provides a modular solution and can conveniently save board space as well.

Input and output impedances are set to a standard 50 Ohms, while the bias circuit can be set via an input pin to adjust IDQ. As well as making circuit design easier, the reduction in external components results in an increase in reliability and ruggedness of the overall circuit into which the PTF31042 is incorporated.

Temperature compensation included in the PTF31042 provides feedback via the bias circuit to the gain loop. This keeps the drain current working point constant, resulting in an increase in output power and lower intermodulation distortion. The class AB 2-tone IMD3 figure for the part is a low -40dB at 5W.

Possessing excellent linearity, the part exhibits a power gain of 12.5dB (typ). Operating from a standard 28V supply, efficiency at 12W output power is a minimum 40%. At 1.96GHz, return loss is as low as -25dB. In line with Ericssonýs commitment to quality, gold metallization is employed to improve linearity and stability, while 100% lot traceability is standard across the range. The PTF31042 is part of a large family of hybrid RF devices from 10 to 90W power output in the range 800MHz to 2.2GHz, covering GSM, CDMA, PCS and PCN bands.

I don't believe Ericsson's claim to produce the first "hybrid" RF transistor is valid. The first devices I remember came out of France in the mid-1970's mainly developed for the burgeoning CATV industry but they were unreliable and difficult to work with. Certainly the PTF31042 is the first 2-GHz hybrid.

I like the -40 dB two-tone intermodulation distortion, which is probably not as low as the PCS base station manufacturers are asking for, but they should be delighted by the constant nature of the number caused by the innovative gain loop to keep the drain current working point constant. I love the 25 dB return loss -- not taken nearly seriously enough by a lot of manufacturers -- but I am curious why it is quoted as a negative number; is that a Scandinavian thing?

The efficiency numbers for this and similar parts are good for the applications and with the linearity achieved are quite acceptable for most station designs. These parts will represent good business for Ericsson and keeps more of their own OEM profits in house.