The Class D RF amplifier - Basics (1/3)

[@KT819GM]

Underrated channel for sure, quality (niche thought) content as this creator provides should get more likes comparing to it's subscriber number. People, don't forget to press that thumbs up, it should help to make channel more visible for a masses.

[@xenoxaos1]

And in the next episode, we'll see how resilient the RFI filtering on his camera, microphone, and computer are.

[@MegaCadr]

I’m excited for this!! I’ve tried implementing Class D RF amplifiers in the past and never had any luck. Looking forward to the rest of the parts in this series.

[@FesZElectronics]

I'm sorry to hear the circuits did not work... May I ask what sort of problems you ran into?

[@MegaCadr]

@@FesZElectronics kept smoking one of the transistors.

[@pieters286]

Yes, also looking fwd at implimentation of snubber circuits /dead band to manage transistor heat generation - and effect on noise.

[@joejane9977]

thank you for your time and the excellent teachings

[@thomassorensen7907]

Hi At 4:16 how come the current is 200mA? For the full-bridge the current is 100mA both with a supply of 10V.

[@thomassorensen7907]

Ahh inductive kickback, just like in a boost-converter.

[@GRAHAMAUS]

In an audio class-D amplifier, the output is the information in the PWM of the switching carrier, and the filter is there to remove the switching frequency (or looked at another way it stores the energy in the switching carrier and that is integrated to become the energy of the delivered PWM converted signal in the load). Here, you are coupling the switching frequency to the load and removing its harmonics, and there is no information content in the duty cycle (PWM) of the carrier. To my mind, what you have then is a push-pull class-C amplifier, not a class-D. Its efficiency will be similar. I have designed RF transmitters* using bipolar transistors in similar class-C configuration and I don't see this as a different class, even if you use MOSFETs or whatever instead. If the sinewave you wanted in the load was carried in a PWM of the switching frequency, I would agree that would be class-D. *I should say that my time as a RF engineer in industry was back in the late 70s and early-80s, class-D wasn't a thing then, but class-C was (and still is) common. nevertheless, I found this discussion interesting, and a good introductory treatment of the matching and filtering networks which can be a difficult subject for those not familiar with RF.

[@ohmschool]

Very nice, what is the root cause of harmonics at 2:30? The circuit is very simple, so I am not understanding what causes harmonics to be formed in the first place. Also, why not generate a clean sine wave first , this way there will be no harmonic to filter.

[@FesZElectronics]

The whole idea of the class D is to generate a square wave by using the transistors in saturated/blocked mode. Its more efficient than going trough a linear mode (like class A/B). When you generate the square wave, you inevitably generate also the upper harmonics.

[@szekerespista3758]

Excellent! 👌. Waiting for the next video!

[@lcarliner]

It appears that what you are describing seems to have more in common with the classic C-class functionality, which is highly non-linear and cannot amplify a fully modulated input!

[@GRAHAMAUS]

I agree Louis, this is class-C. The output frequency is not the PWM of the carrier, but is the carrier itself. Class-C can amplify FM perfectly alright, it just can't do AM if the AM is on the input, but you can modulate the supply with AM and that works OK. Other modulations that use a combination of phase and amplitude would also need to be applied to the drain (collector), not the input (base or gate).

[@lcarliner]

@@GRAHAMAUS This may true with any modulation format that is devoid of any amplitude component, However, with any pure frequently of phase modulation, proper filtration of the output to keep extraneous harmonics down to acceptable levels to avoid overmodulation like splatter would be a design challenge.

[@thomassorensen7907]

For the full-bridge, is the choke really necessary to get the fundamental frequency?

[@FesZElectronics]

The choke is there to limit and stabilize the current. If there is no choke, the parallel load would just be switched between Vcc and -Vcc - it would see a voltage square wave, not a current one.

[@thomassorensen7907]

@@FesZElectronics Yes I see, thank you for the answer. I guess it also make no sense to have a voltage source over a parallel resonant circuit. So in principal it has to be converted to a current source. Maybe that is not the correct way to analyse it?

[@FesZElectronics]

The supply of a circuit in general will be a voltage source (a battery or the output of an electronic power supply usually is voltage stabilized so it behaves like a voltage source); having the inductor in series will help it behave like a current source

[@AnalogDude_]

How about using DG442 or ADG418 analog switches instead of transistors? would that be better? Where you inject the audio signal? and how does the audio signal gets squared?

[@FesZElectronics]

In this video I discuss the RF class D amplifier; I made a few videos a while back about the Audio Class D amplifier.

[@Scrogan]

My question would be how you’d go about modulating the output. OOK would be pretty simple (though maybe with significant ringdown time) but amplitude modulation would need shorter pulse lengths or pulse skipping. Phase modulation would likely be difficult, and frequency modulation may not be possible unless you’re willing to sacrifice your sinusoidal waveform to run at a lower Q. Feels like a push-pull version of the class-C, maybe that’s why it’s the class-D (i.e. what the B is to the A) and the audio version came later.

[@FesZElectronics]

I think for AM you can go with supply modulation (the output of the class D is proportional to the supply voltage); I did see a really complex version of AM modulation though, but I cannot explain it; For FM it should not be a problem since usually the signal BW is rather small - for the commercial FM the BW is 120KHz is I'm not mistaken, so the effect on Q factor should not be that bad. In general you can apply similar principles regarding modulation for class D as for Class C

[@SignalProduct-be2js]

will this RF Class D amplifier works with PSK\QPSK phase-shift modulation?

[@alo1236546]

Dear sir, could you check if module tpa3118 class d as am transmitter

[@piclife1178]

Can you please mention some practical uses for this type of amplifier. I am interest to hear what an amplifier that is essentially single frequency, non-linear for am modes and requires extremely fast transistors (relative to operating frequency) might be used for? Have really enjoyed watching your videos, keep up the good work.

[@AdityaMehendale]

It's in the title :) "RF" --> For communication-system (AM/FM/etc) Other industrial uses include RF-welding for plastic parts. A large RF voltage is applied between two electrodes, and plastic parts sandwiched between these melt due to dielectric-losses in the plastic, effectively "sealing" (fusing) the parts together.

[@AdityaMehendale]

Radio --> Take a typical FM channel - say 101MHz. Its bandwidth is say 200kHz (so only +/-0.1% of the center-frequency) - this is as good as "one frequency" seen from the viewpoint of a filter with Q=10.

[@MegaCadr]

I was planning on using them for digital modes like FT-8. They are efficient, so good for portable operations. I have used Class E up until this point, but would like to experiment.

[@piclife1178]

@@AdityaMehendale I can see it would work for narrow band FM but surely it cannot pass AM signals without total distortion of the modulation. I guess only HAMs use AM these days anyway lol

[@playitlouder451]

@@piclife1178 You'd high level modulation for AM. I.e. Apply the modulation to the supply voltage, say with a PWM circuit, like a class D audio amp.

[@davidkclayton]

I'm curious to see how a voltage switching device such as a tube would work in this scenario

[@FesZElectronics]

As long as you are using the transformer coupled version, it should not be a problem getting it to work; at least at relatively low frequencies.

[@rehandrone7146]

Pls keep up the good work

[@cyril1522]

thanks

[@vakeesan3138]

Please do some advertisement to your channel, this is pure gem 💎 .

[@Radplus]

I have some problems!!! How rf power amplifier works I have a walkie talkie that have a rf power amplifier module inside and powered by 7.5v and have 5watts of power into the antenna How this is happening? We know our antenna have a impedance of 50 ohm so if: I = v÷r P= v × I I= 7.5v ÷ 50ohm=150mA P=7.5v×150mA=1.12 watts So why ican see 5watts in my power meter

[@embeddedlab6669]

what books on ltspice do you know. maybe you have an e-book version?

[@pieters286]

I've only see him use the manual in help. This channel is the "book" on matching ltspice simulations with practical circuits.

[@FesZElectronics]

I do have a book on LTspice from Wurth, but it mainly covers what is in the help file, and that much more. But if you find articles on analog.com its usually more usefull; just google whatever you need to do and usually you will be able to find and example somewhere.

[@troy6882]

So would I 😁.was wondering about the npn gate blow out when specific on type all they had then was npn wasn't expecting it to short blowing the out gate one after the next move one on the list and it's fine. 😃😄😁😆😅5 Smile's. You can even have them off start up it has a saturation jump issue yet you can while having the positive line in cut to saturate the circuit to stop that jump 4second lock out the negative and electron roll still can pass this and on some 555 jump the reset button turning it back on???😁 multi line system use stuff.

[@jimsonweed1846]

Резистор МЛТ-2, ваттафак?

[@FesZElectronics]

I have a lot of old components lying around :D

[@playitlouder451]

The voltage switching class D amp looks a lot like an LLC DCDC converter.

[@diegogarciamedina6363]

can you share ltspice file? thanks