Power Amplifier Linearity and Predistortion

[@nezbrun872]

I worked on this extensively about 20 to 25 years ago, when technology was emerging to make this practical with off the shelf DSPs, without having to rely on dedicated ASICs or a large component count solution. There were other complexities in my application, as it was for space, and I was also using additional methods applied to the PA itself, based on Doherty techniques, to improve efficiency to reduce power budget and improve thermals. I never thought I'd ever see a KZbin video on it though! Bravo!

[@AllElectronicsChannel]

Thanks!!

[@MohammadInsafrapz]

very informative and helpful, with some background knowledge in RF I was easily able to grasp the general idea of distortion and DPD, Thank you so much and keep up the amazing work.

[@AllElectronicsChannel]

Glad you liked !

[@steniomedeiros6709]

Fantastic Video! Although the application in telecommunications has been mentioned, this demonstration can help us a lot in the analysis of the data captured via transducers for the most diverse purposes. Very enlightening, thank you!

[@AllElectronicsChannel]

Great to hear!

[@user-yk5fm1br6w]

great job Greogry! You have solid RF fundamentals knowledge- Thank you

[@AllElectronicsChannel]

My pleasure!

[@condor7001]

There are also a few tricks to get distortion products in a cascade of amplifiers to cancel, or at least to add at slower rates than normal. Comes up a lot in CATV work, although it's a bit of a niche application

[@AllElectronicsChannel]

Very interesting! How could I search for this techniques?

[@ricksampson6780]

Excellent tutorial.

[@gammaleader96]

Thanks for the video! You made the concepts very clear, I especially liked how you explained the OIP3, it is always a little strange to think about it in a way. Mobile basestations use predistortion as well. I have even seen systems, that use a directional coupler on the antenna output, where they have a complete IQ downconverter into baseband to monitor and dynamically tune the input signal, very fascinating.

[@AllElectronicsChannel]

🍺🍺🍺

[@markboldyrev8321]

Around minute 21 - I think that receiver-side compensation for the amplifier non-linearity is possible (for example, using known calibration symbols), however the issue will be different - the resulting transmitted signal will be much more susceptible to noise due to the decreasing distance between the constellation symbols.

[@aduedc]

Great presentation. It would be good to see some circuit implementation of pre-distortion.

[@AllElectronicsChannel]

I think that old analog pre distorters used diode structures to create the distortion profile

[@aduedc]

@@AllElectronicsChannel I saw an active cancellation in one chip, by taking part of output filtering out the band using band stop filter and adding the rest with correct phase and amplitude to the input to cancel them out at the output. The control loop tries to minimize out of band signals by adjusting the feedback amplitude and phase. They do that to say class D amp, so they can go above 1dB compression even higher than 3rd order intercepts. Since in class D the output transistor is either on or off, only the duty cycle changes. Remember the switch transistor must be much faster than carrier signal. So for 10MHz transmitter you need 1GHZ capable switch. Of course the output has a band pass tank (LC parallel circuit) in front of it.

[@atSievna]

Excellent video. Very well explained and concise. I subscribed!

[@layt01]

Very good, thanks!

[@abhisheksourya9033]

superbly explained. Thank you

[@AllElectronicsChannel]

Glad it was helpful!

[@plastilinovbly]

Thank you for the video! Great expanation

[@AllElectronicsChannel]

Glad it helped!

[@2011HPS]

Excellent! Many thanks.

[@AllElectronicsChannel]

You are welcome!

[@ahmadhasan8607]

Sir. I found your session very helpful, and informative. My question is that, can we apply some negative feedback from output to input of the amplifier to increase its linearity.

[@AllElectronicsChannel]

Yes and it is a common practice!

[@ahmadhasan8607]

@@AllElectronicsChannel Thank you Sir for your reply. Can you show us in a video how it affects the linearity of a PA on simulator.

[@gamachutafari3588]

Thanks Brother; please How do you simulate The Amplifier Design ? to know the Linearity difference and Energy efficiency?

[@Donquixote-qv7sp]

Great video. Can you share the code that was used to generate all those graphs?

[@AllElectronicsChannel]

Think it is posted on Patreon!

[@theoryandapplication7197]

thank you

[@AllElectronicsChannel]

Welcome!

[@FarzinAhmadian-ej9wl]

How can we have these matlab codes?

[@like_cabezon3810]

In the industry all the amps of the same manufacturer are predistorted with the same model or each system is tested and then it’s nonlinearities are measured to calibrate and correct it’s behabiour?

[@AllElectronicsChannel]

Both.. advanced systems use a real time receiver that samples the output and adjust the pre distortion system in operation

[@ostrov11]

Спасибо.

[@AllElectronicsChannel]

🙏🏼🙏🏼🙏🏼

[@dn0hr9879]

🎉🎉

[@AllElectronicsChannel]

🍺🍺🍺

[@markboldyrev8321]

Around minute 21 - I think that receiver-side compensation for the amplifier non-linearity is possible (for example, using known calibration symbols), however the issue will be different - the resulting transmitted signal will be much more susceptible to noise due to the decreasing distance between the constellation symbols.

[@AllElectronicsChannel]

What I think very interesting is that the "spectrum" is a linear concept. Is the frequency domain point of view of the time domain world. I'm not sure, but I don't know if exists some fundamental limitation for the existence of other points of views, like nonlinear ones