P1dB using CW vs Complex waveform

Thread Starter

Tom Hong

Joined Feb 27, 2016
Hi experts,

First, I need a sanity check on my basic understanding of P1dB...

If a spec sheet for a PA says it has a gain of 27dB and a output referred P1dB of 34dBm, that means the maximum power I can feed at the input cannot exceed 7dBm if I want to keep it in the linear region.

Is that right?

But is that the case for ‘any’ signal feeding into the PA? From my very limited experience testing transmitter chain, when feeding a VCO+ PA chain with a single tone vs a complex waveform, say, a 64 QAM 54 Mbps signal, that PA’s max power output will be lower, because the PA may not have the response time to react to spin up it’s max power when responding to VCO driven by a complex waveform. From my recollection I think when I run tests feed complex waveforms on a VCO+PA chain the measured power output I got is usually 10-30dBm less than what I measured when I feed a single tone.

So would P1dB still be the same for both single tone and complex waveform at the input?

The reason I am asking, is because I am trying to see if I can get more dBm at the output of this TX chain, consist of a VCO, a 12dB 'linearizer', and a PA, by simply adding another PA at the end.

The TX chain as is gets about 29dBm at the output.

If I were to add another 12dB attenuation and this PA aforementioned (a gain of 27dB and a output referred P1dB of 34dBm), the power output after the second PA would add up to be at 44dBm, which is 10dBm higher than the P1dB listed.

But I am wondering if that’s still the case with a complex waveform, given my past measured power is always 10-30dB less than a single tone, I should be able to get an extra 15dBm without driving the second PA into non linear region, assuming the P1dB stays the same.

Also, is there a co relation between P1dB, P2dB, P3dB and EVM? That is, how far outside of linear region could I drive a PA before it start messing up the information that passes thru it?