An X-band GaNHEMT power amplifier design using an artificial neural network modeling technique


Creative Commons License

LEE S., CETINER B., Torpi H., CAI S., LI J., ALT K., ...Daha Fazla

IEEE TRANSACTIONS ON ELECTRON DEVICES, cilt.48, sa.3, ss.495-501, 2001 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 48 Sayı: 3
  • Basım Tarihi: 2001
  • Dergi Adı: IEEE TRANSACTIONS ON ELECTRON DEVICES
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.495-501
  • Anahtar Kelimeler: device modeling, gallium nitride (GaN), HEMT, neural network technique, power amplifier, DEVICES
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

In this paper, the first gallium nitride (GaN) based high electron mobility transistor (HEMT) power amplifier design using an artificial neural network (ANN) modeling technique is presented. The ANN technique was used to model the small signal behavior of a device with a gate periphery of 1 mm and a gate length of 1 mum over the broad frequency range from 1 GHz to 26 GHz with multiple bias points, based on fitting calculated S-parameters to measured S-parameters. A single stage amplifier constructed using these parameters showed a gain of about 7 dB and an output power of 1.2 W at 8 GHz when biased at V-ds = 20 V and I-ds = 220 mA in class AB mode. The good agreement between measured and simulated results was shown in both S-parameter modeling and in amplifier design.

—In this paper, the first gallium nitride (GaN) based high electron mobility transistor (HEMT) power amplifier design using an artificial neural network (ANN) modeling technique is presented. The ANN technique was used to model the small signal behavior of a device with a gate periphery of 1 mm and a gate length of 1 m over the broad frequency range from 1 GHz to 26 GHz with multiple bias points, based on fitting calculated -parameters to measured -parameters. A single stage amplifier constructed using these parameters showed a gain of about 7 dB and an output power of 1.2 W at 8 GHz when biased at = 20V and = 220mA in class AB mode. The good agreement between measured and simulated results was shown in both -parameter modeling and in amplifier design. Index Terms—Device modeling, gallium nitride (GaN), HEMT, neural network technique, power amplifier.