Using an MMIC LNA IC as a Power Amplifier
I have a power consumption sensitive application where I need to amplify a 0dBm RF signal to about 16dBm in the 400MHz band. While I can find some specific PA ICs to get me there, I frequently came across LNA IC's which calculated to be more power efficient at the job. That leads me to a question, is it inappropriate to use an LNA amplifier in a PA application? If so, why is this?
That's surprising, since LNAs are not usually optimized for efficiency. But 16dBm isn't really the domain of "PAs" anyways, more like general purpose gain blocks. What components are you looking at?
Now depends by the linearity requirements of your design.
Most of the MMIC available on the market have P1dB higher than +16dBm, but their current consumption is in 40mA to 100mA range.
Using a single stage transistor amplifier (LNA if want to name), you can get +16dBm output power for less DC current consumption.
I am looking at the Avago MGA-62563 right now. The traditional PA I was also looking at was the RFPA0133 in its 16dBm mode.
According to my calculations, the RFPA0133 would use around 230mW of DC power, where the MGA-62563 would use around 150mW of power for about the same power level (Technically, the MGA would give me 17dBm).
The RF signal is narrow and just needs a tuning range of less than 1MHz, but also requires very low phase error.
No, it would be perfectly fine.
LNA are usually designed for low noise, but some of them are also designed for high IP3 or high dynamic range so they can be great low level PAs.
Go for it!
You can safely use a LNA if it has high enough 1 dB compression point in the specified frequency. Efficiency of LNA's are pretty good if they have high enough P1db because to have low Noise Figure they tend to use low Collector currents and voltages around 10-20 mA=Ic and Vc=3-4 V typically resulting in low power consumption.
If the harmonics are not very serious issues for you, a simple Class-E PA ( even with a ordinary pHEMT transistor) will give you better efficiency.The design process is obviously more complicated but it's worth to try.
at 400 MHz you might be able to do a Class E power amplifier stage that would be super efficient. Maybe 90% efficient:
http://www.skyworksinc.com/downloads...IMS_062013.pdf