improvement of IMD3

Hi all,

Recently, I have designed several linear power amplifiers, but IMD3 becomes a large problem of my circuits. Can you please tell me, how I can improve the IMD3 besides increasing the RF power? Which parts of an amplifier circuit is most important for a good IMD3, the matching networks, bypass capacitors or anyting else?

You just pointed out, matching networks (for proper load-line), bypass capacitors, and correct handling of harmonics.
Bias filtering at both base/gate and collector/drain should not be very narrow, and must have a bandwidth about two times greater than the modulation bandwidth in order to maintain constant group delay and flat gain.

It depends on your process, Bipolar or CMOS?
there are some different points of view between different processes..

But in general, bias is very important, not only the decoupling cap or harmonic impedence, but some DC offset in your bias because of even-order nonlinearity.

adaptive bias is often studied as efficiency improvement, but in another hand, it could also be considered as linearity improvement method.

I don't know how to implement it. but I know some guys use an IMD
compensation diode to cancel out some phasing that helps with IMD.
You will need a uC and DAC to adjust the bias on the diode to get
optimal phasing.

Here's a starting point.
h++p://ieeexplore.ieee.org/Xplore/login.jsp?url=http%3A%2F%2Fieeexplore.ieee.org%2Fi el5%2F7260%2F28762%2F01291448.pdf%3Farnumber%3D129 1448&authDecision=-203

Cheers

Dear Element_115 :


can you give the title of the paper?thanks!