Hello,

I need some help on peak power and efficiency measurement.

We are measuring the output power vs. drain efficiency characteristic of a power amplifier. Because the large amount of heat generated at the vicinity of the peak power region, we are stimulating the PA with pulsed modulated carrier signal, at some RF frequency. Let's the duty cycle of the signal is D.

For peak power measurement we only have 8481A power sensor plus HP438A power meter, so what we are actually measuring is the *average output power*. With the duty cycle information, we calculate the peak power by simply adding the ducy cycle value (in dB scale) to the average power value. This is how we get the *peak power* result.

We have some difficulities when measuring the current consumption. We use a digital multimeter (34401A) to measure the *average current* (Idc) during the entrie cycle. Of cause before that, we already recorded the quiescent current (Idq). Again, with the duty cycle (D), we get the peak current consumption by solving the equation: Ipeak times D + Idq times (1-D) = Idc.

However, because there is some inherent problem with the power transistor, the quiescent current of PA (actually the transistor itself) is drifting during the measurement. For example, it will be lowed after each power sweep. We believe heat is one factor, but some defects of the transistor itself also taking the responsibility. Now the situation is the Ipeak value we calculated will not reflect the real peak current consumption.

Could anybody give us any suggestions how we could do our measurements more accurately? What kind of equipments do we need to upgrade? For example some power source with pulse current capacity? How we deal with the decoupling on the PA itself? etc...

Thank you very much for any reply in advance.

Best Regards,

NL

I need some help on peak power and efficiency measurement.

We are measuring the output power vs. drain efficiency characteristic of a power amplifier. Because the large amount of heat generated at the vicinity of the peak power region, we are stimulating the PA with pulsed modulated carrier signal, at some RF frequency. Let's the duty cycle of the signal is D.

For peak power measurement we only have 8481A power sensor plus HP438A power meter, so what we are actually measuring is the *average output power*. With the duty cycle information, we calculate the peak power by simply adding the ducy cycle value (in dB scale) to the average power value. This is how we get the *peak power* result.

We have some difficulities when measuring the current consumption. We use a digital multimeter (34401A) to measure the *average current* (Idc) during the entrie cycle. Of cause before that, we already recorded the quiescent current (Idq). Again, with the duty cycle (D), we get the peak current consumption by solving the equation: Ipeak times D + Idq times (1-D) = Idc.

However, because there is some inherent problem with the power transistor, the quiescent current of PA (actually the transistor itself) is drifting during the measurement. For example, it will be lowed after each power sweep. We believe heat is one factor, but some defects of the transistor itself also taking the responsibility. Now the situation is the Ipeak value we calculated will not reflect the real peak current consumption.

Could anybody give us any suggestions how we could do our measurements more accurately? What kind of equipments do we need to upgrade? For example some power source with pulse current capacity? How we deal with the decoupling on the PA itself? etc...

Thank you very much for any reply in advance.

Best Regards,

NL