User:Grantwt/sandbox/Phasing Quadrature Amplification

Background
The phasing modulator has been around since the 1940’s, in it’s early form it was used to generate SSB as a more efficient transmission format over AM that was widely used at that time. This is when we started working with In-phase (I) and Quadrature (Q) inputs, to represent each part of the waveform as Phase and Amplitude.

Single-sideband modulation []

Up until now we have been generating various wave forms and measuring the effects of the pulse widths to work out the minimum required bandwidth. Where in this process works the opposite way and used pulses to generate various wave forms, this technique is able to works both ways. This form of Quadrature amplification was invented in 2017, after experimenting with an optical road safety system called the Electronic eye project, it was soon discovered that the same process could be modified to work at radio frequencies. This form of switching amplification is made up of two parts, one been a phasing modulator using In-phase and Quadrature inputs, the other is switching output stage that acts as the amplifier. For this process to work it must have minimum of four pulses, two for the In-phase components positive and negative going and same for both Quadrature components.

In-phase & quadrature []

Classes of amplification
From the beginning of electronic amplification devices there was a requirement to understand how the amplification process is been done. The way this was worked out in the analog classes was by using angles to specify the on time in degrees. So therefore you had Class A, that on for 360° of the cycle, with Class B that was on for 180° x 2 of a cycle. Class C, is on from a few degrees and used a LC tuned circuit combination to generate a full cycle. With switching amplification classification is based on the type of switching and the way the output filtering is being done.

Power amplifier classes []

Types of pulse modulation
Where Pulse Width Modulation (PWM) has the same information on both sides of the pulse, but is mirrored or 180° out of phase, whereby the phase information is canceled out leaving just amplitude information. By converting PWM to Pulse Phase Modulation (PPM) by removing one side, by doing this you also keep all the encoded information as well as the all important phase information. This is in a way like what you would get with Amplitude modulation with the side bands on each side of the carrier, where all that is needed is just one of the side bands.

There are three types of pulse modulation that can be used to build other wave forms with. There are PWM, PPM and Pulse Location Modulation (PLM). Whereby PPM can used to generate the other two forms, PWM has amplitude information and PLM has phase information. So to be able to move from one form to the other depending on what is required Amplitude, Frequency Modulation or both. By using PPM it can do both phase and amplitude, as demonstrated with this new phasing modulator amplifier design. This technique is able to work with both Radio Frequencies and with light at optical wavelengths.

Radio frequency []

PWM amplification
Class D & I are switching amplifiers, where class D uses PWM this process chops the sine wave into wide or narrow pulses. At the widest point of the pulse is at the peak of the sine wave and the opposite at the minimum point. With Class I there is two in-phase PWM carriers that are connected to a common clock, using a differential process where one input is offset to the other by 180°. This means the audio input needs to be phase shifted by 0° and 180° to drive each PWM input. Both classes of amplification therefore are linear, what goes in comes out with very good efficiency. These are known as switching classes and all require filtering after their output stages to remove unwanted harmonics, in class D & I a low-pass filter is used. The efficiency of these classes comes from the output device been tuned hard on and off, minimizing power been dissipated within this switching device.

Pulse-width modulation []

Quadrature amplification
Starts out with two signals that have the same frequency and are offset by 90°, which is expanded out to four phase angles that have an offset of 90° (0°, 90°, 180° & 270°). Unlike Class D, Quadrature amplification works at minimum of four times the highest frequency where Class D is a minimum of two times. Another difference between the other switching classes is that Quadrature amplification uses PPM not PWM. Where PWM has no phase information, therefore it is used to very the amplitude, however if one side is remove you end up with both the phase and amplitude components. In quadrature amplification the amplitude part is not used, whereby it is possible to process the phase information within logic gates and by adding I & Q pulses together it is possible rebuild any type of analog waveform. This is where Nyquist is very miss leading, stating that you only need two pules to regenerate a sine wave, not true for phase integrity this is where you need minimum of four. This is the key difference between quadrature amplification and what happens in Class D and in many other switching classes.

CLASS-P™ and CLASS-Q™
These classes of amplification are unique due to the way that they are based on phasing principles, so you will have a Sine and Cosine parts to the step waveform. Therefore with these amplifiers are made up of four pulses, two positive going and two for negative going, as parts of the generated analog waveform. This approach is used in these new forms of amplification, moving on from the limitations of Class D.

There are two forms of Phasing Quadrature amplification which are called Class P and Class Q, in Class P (pulse) you have four PPM pulses that are offset by 90° from each other, in Class Q (quadrature) each side of the pulse has the in-phase and quadrature information.

In class P each pulse must be lass than 25% on time, where you have a gating window that the pulse must fall within. The PPM therefore is between 0% and a maximum of 25%, it must be trigged to start at 0°, 90°, 180° & 270°, with PLM the pulse just needs to be within the gating window. The output waveform therefore has 0° & 90° positive going and 180° & 270° are negative pulses. Possible uses for Class P would be in applications where you need a extra level processing between input / output stages, where Class Q has the higher efficiency of the two.

With class Q the maximum on time is 50% and 50% off, where you will end up moving into Class E (Square wave). Therefore when amplifying a modulated signal you always will be less than the maximum of 50% on the positive and negative going cycles to provide room for modulation. Where there is a sharp cutoff between the linear and nonlinear zones, this starts to have an impact above 25% pulse average until you reach 30% where it mostly becomes nonlinear. Another way to look at Class Q is that it provides the linearity of Class A with the efficiency of Class E, making ideal for many forms of analog and digital modulation systems.

With quadrature amplification it is also possible to be used for audio applications, but there is no real advantage over existing classes like D & I, so therefore the focus has been on RF applications where I & Q inputs are used.

As with Class D and all the other switches classes output filtering becomes very important to rebuild the analog waveform, both Class P and Q used with low-pass, band-pass or a combination of both. Where you have both positive and negative going pulses that you are working with, as in figure 4.

Amplifier grouping types:

With amplification classes it is possible to group them by their operating type, helping to show where fall in relation to each other.

Working prototypes; Class Q, AM broadcast transmitter
By using a dual phase modulators at twice the frequency and in logic doing a divide by two, bringing the operating frequency back down to 660 kHz from 1.32 MHz. With this version it is possible to modulated both digital and analog wave forms with very good linearity, for testing analog AM stereo (C-QUAM) and for digital Digital Radio Mondiale (DRM) was used at 64QAM.

AM radio []

C-QUAM []

Digital Radio Mondiale []

Whereby in this configuration it is able operate up to the maximum frequency of 10 MHz, this is well within the range of the AM broadcast band from 540 to 1700 kHz. All the testing was done over three frequencies, 660, 1110 and 1500kHz, where there were gaps found between local radio stations. For the output power it was possible to drive to a maximum of 200 Watt at 100% modulation, using a LDMOS switching device. With a number of working prototypes all using some form of Quadrature amplification, has shown that these amplifier classes have many possible applications.

Where the amplitude was driven by the In-phase pulses and the Quadrature pulses control the phase components of amplified wave form. For AM stereo the sum of Lift + Right makes up the amplitude, and the difference Lift - Right is within the phase of the modulated carrier. When DRM is used, this done by having a 12 kHz intermediate frequency generated in software, with the In-phase and Quadrature outputs coming from a sound device.

Intermediate frequency []