US 20060220749 A1
An automatic quadrature phase compensation system comprises an on-chip analog phase sense circuit capable of detecting small differences in quadrature phase error and providing a corresponding DC voltage, a voltage-controlled or programmable phase delay circuit to implement quadrature phase error correction, and a feedback system or compensation engine used to process the sensed error voltage and apply a corresponding correction signal to the adjustable phase delay.
1. A quadrature phase compensation system, comprising:
a voltage-controlled, phase delay circuit configured to implement quadrature phase error correction on quadrature signals generated by a quadrature signal generator;
an analog phase sense circuit coupled with the voltage controlled, phase delay circuit, the analog phase sense circuit configured to receive the quadrature signals from the voltage-controlled, phase delay circuit, sense a phase error related to the quadrature signals, and produce a DC voltage that is representative of the sensed phase error; and
a feedback system coupled between the analog phase sense circuit and the voltage-controlled, phase delay circuit, the feedback system configured to process the DC voltage and apply a corresponding correction signal to the voltage-controlled, phase delay circuit.
2. The quadrature phase compensation system of
3. The quadrature phase compensation system of
4. The quadrature phase compensation system of
5. The quadrature phase compensation system of
6. The quadrature phase compensation system of
7. The quadrature phase compensation system of
8. The quadrature phase compensation system of
9. The quadrature phase compensation system of
10. The quadrature phase compensation system of claim I, wherein the voltage-controlled, phase delay circuit is configured to adjust the phase of the quadrature phase signal of the quadrature signals in order to implement the quadrature phase error correction.
11. The quadrature phase compensation system of
12. The quadrature phase compensation system of
13. The quadrature phase compensation system of
14. A receiver, comprising:
a quadrature signal generator configured to generate in-phase and quadrature signals;
an in-phase VCDL coupled with the quadrature signal generator and configured to implement quadrature phase error correction on the in-phase signal generated by the quadrature signal generator;
a quadrature phase VCDL coupled with the quadrature signal generator and configured to implement quadrature phase error correction on the quadrature phase signal generated by the quadrature signal generator;
an analog sense circuit coupled with the in-phase and quadrature phase VCDLs, the analog sense circuit configured to receive the in-phase and quadrature phase signals from the VCDLs, sense a phase error related to the signals, and produce a DC voltage that is representative of the sensed phase error; and
a feedback system coupled between the analog sense circuit and at least on of the VCDLs, the feedback system configured to process the DC voltage and apply a corresponding correction signal to at least one of the VCDLs.
15. The receiver of
16. The receiver of
17. The receiver of
18. The receiver of
19. The receiver of
20. The receiver of
1. Field of the Invention
The invention described herein relate generally to transceivers requiring precision quadrature local oscillator (LO) signals.
Many communications standards, for example, 802.11 a/b/g for WLAN and DVBS2 for Digital Satellite TV, demand fully integrated transmitter and receiver architectures offering high performance at low cost. Direct conversion receivers and image-reject receivers are examples of such architectures. In these architectures, excellent phase and amplitude balance are critical to prevent quadrature inaccuracy from limiting the overall performance. In particular, inaccuracy in the phase and amplitude balance can result in poor image rejection.
In a receiver, image rejection refers to the receiver's ability to reject signals at its image frequency. In a transmitter, poor image rejection can produce an image signal that falls within the receive band of adjacent channels, which will interfere with devices operating on the adjacent channels. The image and desired inputs both mix with the Local Oscillator (LO) signal and are downconverted to the same frequency. This poses a problem in conventional double-balanced mixers because the two downconverted products interfere with each other, since they exit at the Intermediate Frequency (IF) output port together. Image rejection is thus defined as the ratio of the downconverted image signal power exiting at the IF output port, to that of the desired signal, exiting the same IF output port. For example, if the downconverted image and desired signal levels at are −30 dBm and −10 dBm respectively, then the image rejection is 20 dB. As mentioned, good image rejection requires close amplitude and phase matching.
Conventional double-balanced mixers use filters to block the image from entering the mixer. This prevents the mixer from generating a down-converted image signal. It will be understood, however, that as the IF is reduced, the desired and image signals move closer together in frequency, converging on the LO frequency, which limits the effectiveness of filtering and/or increases the complexity and cost of filtering solutions.
In comparison to conventional double-balanced mixers, image-rejection mixers, for example, achieve image-rejection through phase cancellation, not filtering, so the frequency spacing between the image and desired inputs can be negligible. Conventional image rejection receivers can achieve an image rejection ratio in the range of approximately 20-30 dB, which corresponds to a phase imbalance that can be as high as 10°. The standards referred to above, however, can require a phase accuracy of better than 1°. For example, such standards can require 60 dB of image rejection, which can require a phase imbalance that is as precise as 0.1°.
Phase accuracy in conventional receivers is often limited by the on-chip matching of devices within the quadrature generation circuit. Careful layout can help minimize phase inaccuracy; however, typical performance with good matching is still typically limited to ±3°. This corresponds to an Image Rejection Ratio (IRR) of less than 30 dB, which often results in a corrupted signal constellation and high bit error rate.
Given the mismatch limitations of conventional designs, successful implementation of the architectures referred to above requires additional phase calibration either inside the tuner included in the RF front end or inside the baseband demodulator circuit. Many conventional designs use the later approach to address the phase imbalance issue. The problem with suoh solutions, however, is that the RF front end and the baseband demodulator circuit are often produced by two different entities. As a result, the designer of the RF front end is faced with a dilemma. The designer can assume that the baseband circuit will include the requisite compensation and not include any compensation in the RF front end. But if the baseband circuit does not include sufficient compensation, then the receiver will not perform adequately. Thus, it can be preferable for the RF front end to include the requisite compensation, because it makes the RF front end independent of the baseband circuit. Alternatively, the system design may be forced to purchase both the baseband circuit and RF front end form the same supplier as part of a chip set solution, in order to ensure that the chip set has adequate compensation, which limits the designers options. Further in an implementation of an RF frequency translation modular based on an RF-Analog Baseband-RF conversion architecture for emerging digital satellite TV broadcasting systems, it is imperative to have an analog method to generate an accurate quadrature LO signals for both RF to analog baseband conversion (receiver portion of the FTM modular) and analog baseband to RF conversion (transmitter portion of the FTM modular).
A few analog-based, front-end solutions have been proposed; however, none have implemented continuous phase error correction with <1° performance. One such solution is based on least-mean squared (LMS) algorithm. The main drawback of such solutions are increased power consumption, increased design complexity, and a one-time only calibration run at startup. Therefore, such designs require additional startup time and phase error performance may drift with temperature variation after power-up.
Another solution provides programmable amplitude and phase. The main drawback of such a solution is that they do not include an on-chip sense circuit or calibration engine. Therefore, such solutions still require support from the baseband circuit.
Still another solution generates N harmonics of the LO frequency using a high frequency delay locked loop. The main drawbacks of such an architecture are
An automatic quadrature phase compensation system comprises an on-chip analog sense circuit capable of detecting small differences in quadrature phase error and providing a corresponding DC voltage, a voltage-controlled or programmable phase delay circuit to implement quadrature phase error correction, and a feedback system or compensation engine used to process the sensed error voltage and apply a corresponding correction signal to the adjustable phase delay.
In one aspect, the automatic quadrature phase compensation system comprises a Phase Locked Loop (PLL) in which the quadrature phase difference between I and Q local oscillator signals is locked to 90°. Other aspects of the embodiments of the invention are further described in Exhibit A, appended to this application and incorporated herein by reference.
In another aspect, since the compensation PLL requires only phase adjustment and is not required to generate a frequency output, it can be simplified further into a Delay-Locked Loop (DLL). A DLL is a specific type of PLL used to synchronize clock edges, where the VCO is replaced by a voltage controlled delay line (VCDL).
These and other features, aspects, and embodiments of the invention are described below in the section entitled “Detailed Description.”
Features, aspects, and embodiments of the inventions are described in conjunction with the attached drawings, in which:
The systems and method described below relate to various embodiments and implementations for an automatic quadrature phase compensation system. The systems and methods described herein can apply equally to receivers and transmitters. While certain embodiments are described below, this should not be seen as limiting the systems and methods described herein to any particular embodiment, or implementation as it will be apparent that many variations are possible.
Feedback system 108 can comprise an analog feedback system or a digital feedback system. Analog feedback systems are described in the embodiments below. A digital feedback system 108 can comprise a filter for filtering a signal generated by phase sense circuit 106, a sampling circuit, and an Analog-to-Digital (A/D) converter. The filtered signal can then be sampled by the sampling circuit and converted to a digital representation using an the A/D converter. The generated digital representation can then be processed by a digital controller to set switches on a programmable delay circuit to adjust the delay accordingly.
One continuous implementation of phase compensation system 100 is equivalent to a Phase Locked Loop (PLL) in which the expected phase difference between I and Q local oscillator signals is locked to 90°. Thus, any variation from the expected 90° phase difference between the I and Q local oscillator signals will be detected as a phase error by sense circuit 106. Since system 100 is only designed for phase adjustment and is not required to generate a varying frequency output, it can be simplified further into a Delay-Locked Loop (DLL). A DLL is a specific type of PLL used to synchronize clock edges, where a Voltage Controlled Oscillator (VCO), needed to produce such a varying frequency output, can be replaced by a Voltage Controlled Delay Line (VCDL). Certain DLL implementations are described in detail below.
Thus, quadrature local oscillator signal generation performed in accordance with the systems and methods described herein can be broken down into two parts: a coarse quadrature signal generation; and a refinement, or compensation, of the coarse quadrature signals to produce finely compensated quadrature signals. Coarse quadrature signal generation is accomplished via quadrature signal generator 102, which can, depending on the embodiment, be implemented in the form of a conventional divide-by-two circuit and corresponding PLL. Such divide-by-two circuits inherently generate I and Q outputs with phase accuracy approximately limited to the range of ±3°, based upon the matching of certain time delays associated with latches that comprise part of the divide-by-two circuit. The fine tune, calibration circuit can be implemented by loop 110 in
It should be noted that some or all of the components in
Implementing the fine tune calibration circuit in the RF front end as an analog solution allows the RF front end to work with a variety of baseband demodulators as a stand alone product without a full chip set solution. Further, it makes re-transmission of the signal possible before demodulation. This makes possible potential applications in the implementation of a frequency translation module (FTM) for digital satellite television broadcast. Further, an analog solution configured in accordance with the systems and methods described herein can be configured to continuously correct for phase imbalance. Therefore, compensation provided is valid over temperature, supply voltage, mismatch, process variation, and LO frequency.
As mentioned, loop 110 can take the form of a DLL.
XOR 214 can be configured to provide a pulse width modulated (PWM) error voltage signal corresponding to the quadrature phase error between the I and Q signals. Loop filter 212 can then translate the PWM error signal into a DC voltage corresponding to the phase error. Loop filter 212 can, for example, be implemented as a low pass filter; however, it will be clear that other implementations for loop filter 212 are also possible. The DC voltage generated by loop filter 412 can then be provided as the control voltage for VCDLs 208 and 210. As will be explained below, the control voltage controls VCDLs 208 and 210 so as to reduce the quadrature phase error between the I and Q signals.
In the example of
Many circuit topologies can be used for implementing VCDLs 208 and 210.
The output of delay controlled, slow signal path 612 is then combined with the output of fast signal path 610 in combiner 606. Therefore, the overall interpolated phase delay of the combined signal (VOUT) can be controlled via the control of the gain coefficient on the buffers 604 and 608. In one embodiment, the delay can be adjusted from 35ps to 85ps for an input control voltage swing of 400mV as illustrated by the graph of
In the example of
Thus referring back to
It will be clear that in other embodiments, all of the tuning can be allowed to occur on the I signal, while the Q signal remains unchanged, i.e., VCDL 806 is biased to a specified set point. Further, while not shown, DLL 110 can comprise a loop filter, such as loop filter 212.
An analog solution configured in accordance with the systems and methods described herein requires,very low current, as the only extra blocks consuming current are the phase detector and an extra stage of the I/Q buffer(s) for implementing the VCDLs. Moreover, the solutions described can implement phase imbalance compensation directly on the quadrature circuit of the LO, thereby preventing any degradation in the quadrature amplitude matching of the receiver.
An automatic quadrature phase compensation system 100, such as described herein, can be included in a variety of devices. Such devices can include a settop box, television, satellite receiver, or a tuner, to name just a few.
While certain embodiments and/or implementations of the inventions have been described above, it will be understood that the embodiments described are by way of example only. Accordingly, the inventions should not be limited based on the described embodiments. For example, it will be clear that the systems and methods can be applied to any communication architecture that requires a precision quadrature LO, whether in the receiver or the transmitter. Rather, the scope of the inventions described herein should only be limited in light of the claims that follow when taken in conjunction with the above description and accompanying drawings.