[Amps] Pre-Distortion Linearizer

[jeff millar]

Hi Paul..

Check out US Patent 4588958 by Allen Katz K2UYH.  It's a common way to linearize 
solid state PAs for the cellular market back in the day.   The basic idea is to 
emulate the compression curve of the amplifier (gain rolloff as power input 
increases) using the nonlinearity of a diode.  The trick Al discovered is how to 
use a 90 deg hybrid to reflect the nonlinearity back into the input path in a 
useful way.  I don't know if that's what AC2IQ uses.  He definitely has some 
cool ideas.

Another good source on the description of the causes of linearity and the 
requirement on linearization comes from Joel Vuolevi's PhD thesis

"Analysis, measurement and cancellation of the bandwidth and amplitude 
dependence of intermodulation distortion in RF power amplifiers"

and later book

"Distortion in RF Power Amplifiers"

I learned a tremendous amount from studying the Katz and Vuolevi work when I 
worked on high linearity cellular systems.

Amateurs should really to attack the splatter problem.  Most of the interference 
on the bands comes from splatter from many -30 dBm IMD transmitter and 
amplifiers.  It's easy to get 10 dB improvement in IMD  with a pre-distort 
linearizer.  Getting up to 30 dB is feasible but takes a digital system and 
complex design.

Many high power, high efficiency, low IMD cellular amplifiers use a Doherty 
circuit with one amplifier handling the low level parts of the signal and 
another handling peaks.

jeff, wa1hco



On 05/03/2013 08:18 AM, Paul Christensen wrote:
> Not exactly Motorola transistor related but the subject reminded me of this: 
> AC2IQ has developed an interesting pre-distortion "linearizer."
>
> http://www.qrz.com/db/AC2IQ
>
> Linearizers have been used in the cell phone industry to more efficiently pack 
> more digitally-modulated channels into licensed spectrum.  Broadcasters have 
> been using this technology for the last decade.  Swiss ADAT was the first 
> amateur manufacturer to use the technology in their ADT-200A transceiver.  In 
> fact, it works so well that within limits, the ADAT will sample the output of 
> a non-linear amp, and correct for the amp's non-linearity back in the 
> transceiver.
>
> Apparently, AC2IQ is using an analog, rather than a digital method to achieve 
> extreme linearity from his SD2931x16  MOSFET amp.  Not sure what class the amp 
> is biased.  Scroll down to the middle of the page and you can see the result 
> of his work.  This isn't just lab performance.  Folks have observed his 75m 
> signal with no IMD "wings" on their SDR receivers.
>
> I'm aware of digital adaptive pre-distortion techniques, but I've not seen 
> this done using analog components.  Anyone familiar with this type of circuit?
>
> Paul, W9AC
>
> ----- Original Message ----- From: "Paul" <w8aef at q.com>
> To: "Amps" <amps at contesting.com>
> Sent: Tuesday, April 30, 2013 7:12 PM
> Subject: [Amps] Motorola transistors
>
>
>>
>>
>> From: Paul
>> Sent: Tuesday, April 30, 2013 4:11 PM
>> To: ScQRPions
>> Subject: Motorola transistors
>>
>> Advanced Semiconductor continues to make most Motorola, Philips, and SGS 
>> Thompson RF transistors.
>>
>> http://www.advancedsemiconductor.com/index.html
>>
>> de Paul, W8AEF
>>
>> ZF2JI/ZF2TA 8Q7AA FO8DX/FO0PLA XZ0A VU7RG/VU3PYM TX5A A52PP
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps 
>
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps