Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Need help to design low power supply final RF Power Amplifier

Status
Not open for further replies.

ganavel9783

Member level 1
Joined
Sep 6, 2005
Messages
38
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Activity points
2,257
Hi folks,

I'm trying to design a 3 stage PA line up consisting of pre-driver, driver & a final PA. The predriver & driver stage has been designed & my focus now is to design the final PA that needs the following goals to be achieved:

1) Power= 1.8 Watt or 2 Watt
2) Frequency range= 806-870MHz
3) P1dB= 36dBm
Besides that, this PA line up is powered at 3.6V.
This is how it should look like:

01 Mar. 23 00.04.jpg

The question that i would like to ask are:

1) What type of power combining techniques that i can consider to achieve these goal besides the impedance matching technique?
2) What type of power splitter techniques that i can use besides the impedance matching technique?
 

Both questions is the same one.
You can use 90 hybrid coupler.
Why do you need combining tech? One PA is enough.
 
Both questions is the same one.
You can use 90 hybrid coupler.
Why do you need combining tech? One PA is enough.

Agree. By adding a splitter/combiner pair around your finals, you simply get 3 dB more output power before you start to compress. You should also put your matching circuitry immediately before and after the amplifier (i.e. inside the splitter and combiner). That way, you transform your PA input/output impedances to 50 ohms, then go into simple 50 ohm input/output power splitter and combiner circuits.

You could use a 90-degree hybrid (@tony), or a wilkinson power divider/combiner. Both will work, one just gives you a phase shift and runs the amps in quadrature. Check out microwaves101.com for some good reading on splitter/combiner/coupler fundamentals.

You missed an important spec... gain. How much does you final stage need to do, and what are your devices capable of? If they can do it alone, then just use one part, for simplicity. If you need a +3dB to reach your goals, then add the complexity of a splitter/combiner architecture. Design one stage that will work from 50 ohms input to 50 ohms output (includes all matching), then design the splitter and combiner, and finally put them together.... IF you have to go that far!
 
Agree. By adding a splitter/combiner pair around your finals, you simply get 3 dB more output power before you start to compress. You should also put your matching circuitry immediately before and after the amplifier (i.e. inside the splitter and combiner). That way, you transform your PA input/output impedances to 50 ohms, then go into simple 50 ohm input/output power splitter and combiner circuits.

You could use a 90-degree hybrid (@tony), or a wilkinson power divider/combiner. Both will work, one just gives you a phase shift and runs the amps in quadrature. Check out microwaves101.com for some good reading on splitter/combiner/coupler fundamentals.

You missed an important spec... gain. How much does you final stage need to do, and what are your devices capable of? If they can do it alone, then just use one part, for simplicity. If you need a +3dB to reach your goals, then add the complexity of a splitter/combiner architecture. Design one stage that will work from 50 ohms input to 50 ohms output (includes all matching), then design the splitter and combiner, and finally put them together.... IF you have to go that far!

Hi Tony & Ejunear,

Here are some further details on my project:

i) The 3 stage PA line up need to have high linear ouput at 36dBm. So each stage requires a minimum linear gain of 10dB
ii) The existing predriver & driver stage is able to meet this spec but there is no final PA that can go up to P1dB of 36dB with 3.6V supply voltage at 870MHz
iii) The pre-driver devices are matched 50 Ohm input/output fixed-gain block. The device used is ADA4743 (Avago)
iv) The driver stage are matched with 50ohm reactive matching circuits at the input and output. Its serve the purpose of delivering adequate power to final PA by amplifying pre-driver output. The device is RD01MUS1 (Mitsubishi)
v) The final PA device is RQA0011DNS (Renesas) is used with power combining & power spliting method to achieve the 10dB gain at the final PA output.

If the wilkinson power divider/combiner method is used, i guess the board size will be a concern. With the 90-degree hybrid power divider method, will the board size be a concern too? Is there any other power combining methods that i can review besides wilkinson power divider method & 90-degree hybrid power divider method?
 

Hi Tony & Ejunear,

Here are some further details on my project:

i) The 3 stage PA line up need to have high linear ouput at 36dBm. So each stage requires a minimum linear gain of 10dB
ii) The existing predriver & driver stage is able to meet this spec but there is no final PA that can go up to P1dB of 36dB with 3.6V supply voltage at 870MHz
iii) The pre-driver devices are matched 50 Ohm input/output fixed-gain block. The device used is ADA4743 (Avago)
iv) The driver stage are matched with 50ohm reactive matching circuits at the input and output. Its serve the purpose of delivering adequate power to final PA by amplifying pre-driver output. The device is RD01MUS1 (Mitsubishi)
v) The final PA device is RQA0011DNS (Renesas) is used with power combining & power spliting method to achieve the 10dB gain at the final PA output.

If the wilkinson power divider/combiner method is used, i guess the board size will be a concern. With the 90-degree hybrid power divider method, will the board size be a concern too? Is there any other power combining methods that i can review besides wilkinson power divider method & 90-degree hybrid power divider method?

Both combiner methods use λ/4 (quarterwave) transmission lines, so they will probably be on the order of ~2" long (assuming Er=4 for the PCB). If board space is a concern, look at EMC/Florida RF Labs. I've used some of their Zapper products to do quadrature split/combines (and 20 dB in-line couplers) in the past. Looks like they call the new stuff HybridX. I'm sure there are other companies out there, but that's the one that pops into my head (seen other coworkers use them, too).
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top