Ericsson Microelectronics introduced two RF power FETs designed specifically for world-wide use in UHF and the new generation of digital television transmitters.

optimized for the 470 to 860MHz band, the family comprises of the 85W PTF10049 and the 120W PTF10037, both for use in push-pull configuration. The devices are intended to be run from a standard 32V supply, and exhibiting ruggedness, both feature a drain-source breakdown voltage of 65V.

Thanks to a high efficiency figure (typically over 50%) and the use of LDMOS (Laterally Diffused Metal Oxide Semiconductor) technology, the wideband devices exhibit an impressive gain of 14dB for the PTF10037 and 13.5 dB for the PTF10049 - making them ideal for UHF applications.

To simplify matters for circuit designers, the transistors are internally matched, allowing complete host circuits to be built with fewer components. Thermal runaway is eliminated because the drain current in each LDMOS device actually decreases with an increase in temperature.

This is another major advance by Ericsson in making high-power LDMOS products. It must be a little scary for Motorola that their process is resulting in major, faster developments for one of their competitors. A couple of years ago you thought of Motorola and Philips as the two suppliers for power semiconductor devices for RF; Ericsson is trying to push both of them out of the way. The efficiencies quoted are high if they are valid for 8-VSB transmissions, where vacuum devices are down into the single and lower teens. For most applications these products would not be used as a single power module but would be combined with more to get the required input power. Because of the performances here a lot fewer modules will be needed.

The gains are extremely useful for the application they are intended for allowing for a driver in the 10-W range for the PTF10037, for which there are a number of choices available. The internal matching is a dream for the design engineer. I had not realized that drain currents went down in LDMOS as the temperature increases; something I missed in LDMOS briefings and a major selling point in "must-not-fail" environments.