Looking for efficient 900-930 MHz power amplifier chip?

[biff44]

Efficient 915 MHz amp?

What would be your suggestions for the most efficient power amp chip at 900 - 930 MHz and 300 mW output power?

 

[vfone]

Re: Efficient 915 MHz amp?

It depends what modulation are you going to use. Some modulations require linear amplification and minimum phase distortion.
I am attaching info about PA efficiency (%) vs PA classes.

 

[biff44]

Re: Efficient 915 MHz amp?

Good point! Saturated FSK modulation.

 

[vfone]

Re: Efficient 915 MHz amp?

OK.
Because FSK is constant-envelope you can use a nonlinear PA.
For 300mW output power theoretically would be ok class C or E.

 

[biff44]

Re: Efficient 915 MHz amp?

I am curious. You have a class E design that works at 915 MHz???

 

[vfone]

Re: Efficient 915 MHz amp?

At the beginning the class-E was available only up to VHF frequencies…but not anymore.

https://scholar.lib.vt.edu/theses/available/etd-07152001-172453/unrestricted/chap5.PDF
https://scholar.lib.vt.edu/theses/available/etd-07152001-172453/unrestricted/Chap6.PDF

 

[biff44]

Re: Efficient 915 MHz amp?

That is all very interesting. But I was hoping someone had an off the shelf part number chip that provided 55% efficiency.

 

[r-snake]

Re: Efficient 915 MHz amp?

Good point! Saturated FSK modulation.

Hi ,biff44.
Can you give me some suggestion on modulation vs PA classes? such as ,GMSK, QPSK, 8PSK, AM, PM, FM......

 

[biff44]

Re: Efficient 915 MHz amp?

Just in a general sense. Some modulation schemes allow a fully compressed PA to be used.

Some modulation schemes, like QAM have half of the modulation "information" in the amplitude part of the signal. So, obviously, if you compress the amplitude of the largest signals by 1 or 2 dB due to saturation, you can not send the data at all. Those types of modulation require you to "back off" the output power. I think I recall that 16QAM requires around 10 dB of backoff. i.e. you can only run the output power to the P1db - 10 dB level.

Some modulation schemes that you think are constant envelope, really arent. in QPSK, for instance, if you go from the 0 to 180 degree state, the signal actually goes thru the origin, ie winks out in amplitude inbetween states. You amplifier has to be able to handle that without lingering am to pm conversion, or you will have poorer bit error rate because the phase state does not arrive at 180 degrees, but something a little different until the PA settles out.

Some times you can send the data thru just fine, but for emissions mask reasons you have to back off a little to please the government. Overdriven amplifiers can generate 3rd order, 5th order, etc. adjacent spurious emissions.

Hope that helps some.