What technique you've used to improve amplifier's linearity?
1. feedback like source degen, feedback from output to input
2. use harmonic trap
3. make sure the device can handle the power not causing voltage or current clipping
4. signal feed forward for harmonic cancellation
Then I was asked: what else?
I think there are some more, like predistortion (I haven't used).
can you add to the list?
Balanced amplifier
-Choosing the right Operating point which cancels second and/or third order IM products
-Balanced configuration for lower second order IM products
-Parallel frequency selective feedback
Can you explain how "Choosing the right Operating point which cancels second and/or third order IM products" works?
When you look at-for instance GaAs PHEMT- Vgs vs Id curve of a FET device and then Gm curve by differentiating of it, you'll see Gm will increase with some sort of linearity and after that reach a constant and flat value and it finally will decrease.
It tells us by differentiating second order and third order there will be some null point for second and third order derivativatives.These are IM nulling points and at that OP, either second order or third order ( both is almost impossible) IM products will have practically zero values.
When you look at the Gm component of a nonlinear device, you'll see Gm=a0*Gm0+a1*Gm1+a2*Gm3+....
Here a0,a1,a2.. are real/complex nonlinearity coefficients, Gm0,Gm1,Gm2 etc. are n-th order derivatives.
So, when you select the right OP, one of them will be zero and its associated IM product will also be zero.
Thanks BigBoss,
Is that a practical design technique? Thanks.
Yes, it's a pratical and theoritical backgrounded technique.
Look at this paper..
predistorter works pretty well, can easily pick up 5 dB or so in 3rd order intermods. To get more, though, it gets complicated. Easiest to use if you have a constant output power. Harder if your transmit power is variable.
If you have low data rate modulation, do not forget to have very large DC bypass caps!