...

RF Modulator: Working Principle and Development Trend

Cecelia   utmel.com   2021-10-21 17:41:46

RF modulators are used to change signals from devices such as VCRs, DVD players, media players, and game consoles into a format that can be processed by devices designed to receive modulated RF input (for example, radio and television receivers). The RF modulator converts the video (and/or audio) output of a DVD player (or camcorder or video game console) into a frequency that can be assigned to any channel compatible with the TV cable or antenna input. The RF modulator can also be used to obtain audio and video signals from PAL or NTSC/ATSC composite video, RGB, or other composite AV sources, and generate broadcast RF signals, which can be fed to the TV’s antenna/coaxial connector.


Catalog:


I Working principle of RF modulator

1.1Working principle of Diode hybrid integrated modulator
1.2 Working principle of Gilbert integrated modulator
II Development Trend of RF Modulator


I Working principle of RF modulator


1.1 Working principle of Diode hybrid integrated modulator

EKIN2-960 is a typical diode hybrid integrated modulator. Let's use EKIN2-960 as an example to introduce the working principle of the analog modulator.

05677a527a80935dcf98e23351e983c.png

Figure 1: Internal structure diagram of EKIN2-960


As shown, the mixer of the modulator is composed of diode tube stacks, and the phase-shifting network is composed of LC phase shifters. The carrier signal with a power of 10dBm is input from the local oscillator port (LO port), A carrier signal with a power of 10dBm is input from the local oscillator port (LO port), and the single-ended signal is converted into a differential signal through a transmission transformer, which is divided into two channels with the same amplitude and phase and sent to the diode tube stack for mixing. The I and Q signals (frequency near the local oscillator frequency) after mixing are output from the secondary of the transformer, and then pass the 3dB directional coupling phase shifter, making the mixed I and Q signals phase-shifted by 90 degrees and added together. Finally, the modulated signal with suppressed sidebands is output at the RF end of the modulator. If the baseband I and Q signals are single tone signals with the same phase quadrature amplitude, the spectral characteristics of the output end are as follows:

ec78be7b1d633175145ef61e55299bb.png

Figure 2: spectral characteristics of radio frequency output


Fc is the carrier center frequency and fi is the baseband tone signal frequency. The ideal spectral characteristic after modulation should have only one component fc+fi, but the carrier and useless sidebands are not completely suppressed due to the existence of non-ideal factors. The difference between them relative to the power of the enhanced signal (in power dBm) is the carrier rejection and sideband rejection respectively. In addition, due to the nonlinearity of the mixer, the output frequency also contains the second, third, fourth, and fifth harmonics of the baseband signal and frequency components modulated by the carrier signal.


Take EKIN2-960 as an example. The local oscillator is a single-to-double conversion transmission line transformer with a 1:4 impedance conversion function, that is, the impedance obtained at the single end is 1/4 of the impedance of the differential end. Under the action of the local oscillator signal, the two mixer diode stacks inside the EKIN2-960 are in a repeated switching state. At any time, each mixer stack has two diodes connected in series, so the impedance of the differential end of the transmission line transformer is equal to 1/2 of the sum of the on-resistances of the two diodes (the DC resistance of the diodes). In order to make the impedance of the local oscillator port close to 50 ohms, the on-resistance of the diode is generally close to 200 ohms.

The signals of I and Q ports are low-frequency baseband signals, and the port impedance is equal to the parallel connection of the on-resistance of two diodes (the DC resistance of the diodes), which is about 100 ohms. The DC resistance of the baseband port is measured and the result is 70~80 ohms, which is closer to the ideal value.

The internal RF output end of EKIN2-960 is a 3dB direct coupling phase shifter. The coupler is a capacitive transformer component designed according to the port impedance of 50 ohms, so the characteristic impedance of the radio frequency output port is 50 ohms.

According to the above port characteristics, the local oscillator and the RF port are designed to match 50 ohms. The signal power of the baseband port is based on the amplitude of the baseband signal, considering the internal resistance of the signal source and the impedance of the baseband port of the modulator as the load to calculate the baseband input power value.


1.2 Working principle of Gilbert integrated modulator

The reason why it is called Gilbert integrated modulator is that it is mainly composed of a local oscillator power division phase shifter, two Gilbert mixers, and an output power synthesis amplifier. The core component─source mixer was designed by Gilbert in 1967.

8c2c2bb3b5dabc64cd3b90456f53755.png

Figure 3: Internal structure diagram of Gilbert integrated modulator


As shown in the figure, LOIN and LOIP are the differential input terminals of the local oscillator, internally connected with a two-pole amplification and phase-shifting network in order to improve the phase orthogonality and amplitude balance of the RF phase shifter. The local oscillator signal is phase-shifted and amplified to provide a local oscillator drive for the Gilbert mixer of the subsequent stage.

IBBP and IBBN, QBBP and QBBN are the differential input terminals of two quadrature baseband signals, which are sent to the Gilbert mixer through the voltage-current conversion amplifier. The I and Q signals are directly superimposed with the local oscillator signal after being mixed in the mixer, and finally, the modulated signal is output after being amplified by the radio frequency amplifier.


The local oscillator end of the Gilbert integrated modulator is generally differential because the differential input is beneficial to improve the rejection of the local oscillator signal. The reason is that after the local oscillator signal passes through the differential amplifier, the differential input form can minimize the common mode local oscillator signal, thereby improving the carrier rejection of the modulator.

In order to improve the accuracy of the RF phase shift of the modulator, the RF phase shift part of the local oscillator of the general modulator contains a phase shift amplification network composed of a multi-stage RC phase shifter and a differential amplifier. The first stage of some modulators RF phase shifters is an RC phase shifter, and some first stages are differential amplifiers. In design, the differential input impedance of the local oscillator port theoretically should close to the standard impedance of 50 ohms. The bias of the amplifier at the input end of the local oscillator is provided by the energy of the modulator. When there is no integrated DC blocking capacitor inside the local oscillator end, an additional DC blocking capacitor is required at the input end of the local oscillator.

The inside of I and Q baseband signal input ends is the input stage of the differential amplifier, and the input impedance is generally several thousand ohms. Considering that the baseband signal contains a lot of low-frequency components, a non-DC connection is generally required for baseband port applications. Therefore, the DC bias at the input of the differential amplifier needs to be provided by an external circuit.

The I and Q signals are respectively mixed with the local oscillator signal in a double-balanced mixer, and output through a differential amplifier after being superimposed. When the differential amplifier is a double-ended output, the output structure of the radio frequency port is a differential form. Some modulators add a single-double conversion and matching stage after the double-ended output stage, and the RF output is in single-ended form. Regardless of single-ended or differential form, its output impedance is generally close to 50 ohms. The DC bias of the amplifier of the RF output stage is provided by the power supply of the modulator. When the RF output terminal does not contain a DC blocking capacitor, an additional DC blocking capacitor is required.


II Development Trend of RF Modulator


The overall development trend of radiofrequency modulators continues to develop towards miniaturization, low cost, and multi-function. The traditional diode-type RF modulator adopts manual craftsmanship, which has poor index consistency and high cost. The general market price is around $9. One of the development trends of diode-type radio frequency modulators is the conversion from the manual manufacturing process to the LTCC (low temperature baked ceramic) process, which greatly reduces the volume and production cost of the device, and improves the index consistency. Its representative devices have IQBG-2000 etc.

Gilbert integrated modulator is made by traditional IC process, with lower volume and cost, higher index consistency, and overall radio frequency index is better than diode integrated modulator. After properly adjusting its static operating point, the harmonic suppression of some types of devices can be more than 60dBc. Gilbert integrated modulators generally have additional functions such as turn-off control of the radio frequency output. Analog IC manufacturers such as ADI and RFMD mainly recommend such radio frequency modulators. The general market price is about $5. According to the development trend of modulators, in the mobile communication frequency band, Gilbert integrated modulators will replace diode integrated modulators.

Selection point 5: When selecting a radio frequency modulator, the Gilbert integrated modulator is preferred, the diode integrated modulator of the LTCC process is secondly selected, and the hand-made diode integrated modulator is the last choice.


RF modulators are used to change signals from devices such as VCRs, DVD players, media players, and game consoles into a format that can be processed by devices designed to receive modulated RF input (for example, radio and television receivers). The RF modulator converts the video (and/or audio) output of a DVD player (or camcorder or video game console) into a frequency that can be assigned to any channel compatible with the TV cable or antenna input. The RF modulator can also be used to obtain audio and video signals from PAL or NTSC/ATSC composite video, RGB, or other composite AV sources, and generate broadcast RF signals, which can be fed to the TV’s antenna/coaxial connector.


Catalog:


I Working principle of RF modulator

1.1Working principle of Diode hybrid integrated modulator
1.2 Working principle of Gilbert integrated modulator
II Development Trend of RF Modulator


I Working principle of RF modulator


1.1 Working principle of Diode hybrid integrated modulator

EKIN2-960 is a typical diode hybrid integrated modulator. Let's use EKIN2-960 as an example to introduce the working principle of the analog modulator.

05677a527a80935dcf98e23351e983c.png

Figure 1: Internal structure diagram of EKIN2-960


As shown, the mixer of the modulator is composed of diode tube stacks, and the phase-shifting network is composed of LC phase shifters. The carrier signal with a power of 10dBm is input from the local oscillator port (LO port), A carrier signal with a power of 10dBm is input from the local oscillator port (LO port), and the single-ended signal is converted into a differential signal through a transmission transformer, which is divided into two channels with the same amplitude and phase and sent to the diode tube stack for mixing. The I and Q signals (frequency near the local oscillator frequency) after mixing are output from the secondary of the transformer, and then pass the 3dB directional coupling phase shifter, making the mixed I and Q signals phase-shifted by 90 degrees and added together. Finally, the modulated signal with suppressed sidebands is output at the RF end of the modulator. If the baseband I and Q signals are single tone signals with the same phase quadrature amplitude, the spectral characteristics of the output end are as follows:

ec78be7b1d633175145ef61e55299bb.png

Figure 2: spectral characteristics of radio frequency output


Fc is the carrier center frequency and fi is the baseband tone signal frequency. The ideal spectral characteristic after modulation should have only one component fc+fi, but the carrier and useless sidebands are not completely suppressed due to the existence of non-ideal factors. The difference between them relative to the power of the enhanced signal (in power dBm) is the carrier rejection and sideband rejection respectively. In addition, due to the nonlinearity of the mixer, the output frequency also contains the second, third, fourth, and fifth harmonics of the baseband signal and frequency components modulated by the carrier signal.


Take EKIN2-960 as an example. The local oscillator is a single-to-double conversion transmission line transformer with a 1:4 impedance conversion function, that is, the impedance obtained at the single end is 1/4 of the impedance of the differential end. Under the action of the local oscillator signal, the two mixer diode stacks inside the EKIN2-960 are in a repeated switching state. At any time, each mixer stack has two diodes connected in series, so the impedance of the differential end of the transmission line transformer is equal to 1/2 of the sum of the on-resistances of the two diodes (the DC resistance of the diodes). In order to make the impedance of the local oscillator port close to 50 ohms, the on-resistance of the diode is generally close to 200 ohms.

The signals of I and Q ports are low-frequency baseband signals, and the port impedance is equal to the parallel connection of the on-resistance of two diodes (the DC resistance of the diodes), which is about 100 ohms. The DC resistance of the baseband port is measured and the result is 70~80 ohms, which is closer to the ideal value.

The internal RF output end of EKIN2-960 is a 3dB direct coupling phase shifter. The coupler is a capacitive transformer component designed according to the port impedance of 50 ohms, so the characteristic impedance of the radio frequency output port is 50 ohms.

According to the above port characteristics, the local oscillator and the RF port are designed to match 50 ohms. The signal power of the baseband port is based on the amplitude of the baseband signal, considering the internal resistance of the signal source and the impedance of the baseband port of the modulator as the load to calculate the baseband input power value.


1.2 Working principle of Gilbert integrated modulator

The reason why it is called Gilbert integrated modulator is that it is mainly composed of a local oscillator power division phase shifter, two Gilbert mixers, and an output power synthesis amplifier. The core component─source mixer was designed by Gilbert in 1967.

8c2c2bb3b5dabc64cd3b90456f53755.png

Figure 3: Internal structure diagram of Gilbert integrated modulator


As shown in the figure, LOIN and LOIP are the differential input terminals of the local oscillator, internally connected with a two-pole amplification and phase-shifting network in order to improve the phase orthogonality and amplitude balance of the RF phase shifter. The local oscillator signal is phase-shifted and amplified to provide a local oscillator drive for the Gilbert mixer of the subsequent stage.

IBBP and IBBN, QBBP and QBBN are the differential input terminals of two quadrature baseband signals, which are sent to the Gilbert mixer through the voltage-current conversion amplifier. The I and Q signals are directly superimposed with the local oscillator signal after being mixed in the mixer, and finally, the modulated signal is output after being amplified by the radio frequency amplifier.


The local oscillator end of the Gilbert integrated modulator is generally differential because the differential input is beneficial to improve the rejection of the local oscillator signal. The reason is that after the local oscillator signal passes through the differential amplifier, the differential input form can minimize the common mode local oscillator signal, thereby improving the carrier rejection of the modulator.

In order to improve the accuracy of the RF phase shift of the modulator, the RF phase shift part of the local oscillator of the general modulator contains a phase shift amplification network composed of a multi-stage RC phase shifter and a differential amplifier. The first stage of some modulators RF phase shifters is an RC phase shifter, and some first stages are differential amplifiers. In design, the differential input impedance of the local oscillator port theoretically should close to the standard impedance of 50 ohms. The bias of the amplifier at the input end of the local oscillator is provided by the energy of the modulator. When there is no integrated DC blocking capacitor inside the local oscillator end, an additional DC blocking capacitor is required at the input end of the local oscillator.

The inside of I and Q baseband signal input ends is the input stage of the differential amplifier, and the input impedance is generally several thousand ohms. Considering that the baseband signal contains a lot of low-frequency components, a non-DC connection is generally required for baseband port applications. Therefore, the DC bias at the input of the differential amplifier needs to be provided by an external circuit.

The I and Q signals are respectively mixed with the local oscillator signal in a double-balanced mixer, and output through a differential amplifier after being superimposed. When the differential amplifier is a double-ended output, the output structure of the radio frequency port is a differential form. Some modulators add a single-double conversion and matching stage after the double-ended output stage, and the RF output is in single-ended form. Regardless of single-ended or differential form, its output impedance is generally close to 50 ohms. The DC bias of the amplifier of the RF output stage is provided by the power supply of the modulator. When the RF output terminal does not contain a DC blocking capacitor, an additional DC blocking capacitor is required.


II Development Trend of RF Modulator


The overall development trend of radiofrequency modulators continues to develop towards miniaturization, low cost, and multi-function. The traditional diode-type RF modulator adopts manual craftsmanship, which has poor index consistency and high cost. The general market price is around $9. One of the development trends of diode-type radio frequency modulators is the conversion from the manual manufacturing process to the LTCC (low temperature baked ceramic) process, which greatly reduces the volume and production cost of the device, and improves the index consistency. Its representative devices have IQBG-2000 etc.

Gilbert integrated modulator is made by traditional IC process, with lower volume and cost, higher index consistency, and overall radio frequency index is better than diode integrated modulator. After properly adjusting its static operating point, the harmonic suppression of some types of devices can be more than 60dBc. Gilbert integrated modulators generally have additional functions such as turn-off control of the radio frequency output. Analog IC manufacturers such as ADI and RFMD mainly recommend such radio frequency modulators. The general market price is about $5. According to the development trend of modulators, in the mobile communication frequency band, Gilbert integrated modulators will replace diode integrated modulators.

Selection point 5: When selecting a radio frequency modulator, the Gilbert integrated modulator is preferred, the diode integrated modulator of the LTCC process is secondly selected, and the hand-made diode integrated modulator is the last choice.


0755-82568876

gdl02@szgdl.com.cn
0