WO2003079573A1

WO2003079573A1 - Method and apparatus providing calibration technique for rf performance tuning

Info

Publication number: WO2003079573A1
Application number: PCT/IB2003/000901
Authority: WO
Inventors: Pauli Seppinen; Aarno Parssinen; Mikael Gustafsson; Mika Makitalo
Original assignee: Nokia Corpration; Nokia Inc.
Priority date: 2002-03-15
Filing date: 2003-03-12
Publication date: 2003-09-25
Also published as: US20030176174A1; CN101189804A; EP1486012A1; US20060160510A1; KR20040097171A; EP1486012A4; AU2003209549A1; US7203472B2

Abstract

A method is disclosed for operating a RF receiver of a communications equipment, as is circuitry for implementing the method. The method includes, while operating under the control of a data processor of the communications equipment, generating a calibration signal; injecting the calibration signal into a low noise amplifier (LNA) of the RF receiver; measuring a downconverted response of the receiver at a plurality of different frequencies of the calibration signal, or measuring the downconverted response of the receiver at a plurality of different LNA tuning combinations using a fixed calibration frequency, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response so as to compensate at least for variations in component values that comprise the at least one resonator, or adjusting the linearity of the receiver. In one embodiment the calibration signal is generated using one of a noise source located upstream from the LNA and a transmitter signal, having an AM component, that leaks into the receiver through a duplex filter.

Description

METHOD AND APPARATUS PROVIDING CALIBRATION TECHNIQUE FOR

RF PERFORMANCE TUNING

TECHNICAL FIELD:

These teachings relate generally to radio frequency (RF) receivers and, more specifically, relate to methods and apparatus for optimizing the performance of RF receivers, such as those found in cellular telephones and other types of mobile communication devices and terminals.

BACKGROUND:

The following abbreviations are herewith defined.

ADC analog-to-digital converter AM amplitude modulation

ASIC application specific integrated circuit

BB baseband

CDMA code division multiple access

CMRR common-mode-rejection ratio CPU central processing unit

DS-CDMA direct sequence CDMA

DSP digital signal processing

FDD frequency division duplexing

FM frequency modulation FPGA field programmable gate array

IC integrated circuit

ICP input compression point

IF intermediate frequency

IIP2 second-order input intercept point IIP3 third-order input intercept point

IMD2 second-order intermodulation product

IMD3 third-order intermodulation product

LNA low noise amplifier

LO local oscillator PM phase modulation

PD phase detector

PDF phase-frequency detector

RX receiver RF radio frequency

RSSI received signal strength indicator

TX transmitter

VCO voltage controlled oscillator

WCDMA wide-band CDMA 3G third-generation (cellular system)

As is well known, passive components that are used in RF ICs typically have relatively large process variations. This leads to a direct trade-off between the accuracy of the resonant or resonance frequency and the bandwidth of the circuit. As a result it is common practice to use relatively low-Q resonators in the RF signal path in order to ensure a sufficiently wide bandwidth and, thus, sufficient performance without requiring calibration. Additionally, calibrations performed during fabrication are preferably avoided in order to reduce cost. The use of a narrow bandwidth (narrow band) LNA in the RF receiver enables the elimination of a bandpass filter after the LNA, and thus reduces cost. Since the passive component process variations can be large, however, some calibration is normally needed, and the cost savings may not be as great as one would at first expect.

As such, what is required is a simple implementation of a calibration technique that can be used to tune resonators in analog circuits, as well as a technique to utilize (relatively) narrow-band resonators in a radio system.

hi most applications a relatively wide-band LNA is used, which is insensitive to process variations, and if necessary an external filter is placed between the LNA and a downstream mixer in order to reduce transmitter leakage (an undesired signal coupled into the receiver from the transmitter) . hi addition, some structures that use additional resonators in the LNA- mixer interface, or in the LNA topology itself, have been previously presented. Reference in this regard can be made to J. A. Macedo, M. A. Copeland, "A 1.9-GHz Silicon Receiver with Monolithic Image Filtering", IEEE J. Solid-State Circuits, vol.33, pp.378-386, March 1998, as well as to H. Samavati, H. R. Rategh, T. H. Lee, "A 5-GHz CMOS Wireless LAN Receiver Front End", IEEE J. Solid-State Circuits, vol. 35, pp. 765-772, March 2000.

While primarily intended for image rejection purposes, the problems associated with filtering out-of-band signal components are basically the same as when filtering transmitter leakage. However, although the LNA structure with two resonators has been described in the prior art, an adequate solution to the calibration and optimal scaling with current of the two resonator LNA has not previously been proposed.

It is noted that instead of two resonators, one high Q value resonator can be used. However, this approach can be assumed to have the same or similar tuning needs as the two resonator topology.

Reference can also be had to commonly-assigned International Patent Application No.: WO02/084859 Al, Kalle Kivekas and Aarno Parssinen, "Balanced Circuit Arrangement and Method for Linearizing such an Arrangement", International Filing Date 18 April 2001 , that describes a technique to calibrate the second order nonlinearity of a receiver.

Fig. 1 shows a conventional direct conversion receiver 1. After a received signal passes through an antenna 2 the desired radio band (e.g., WCDMA/GSM/or other) is selected using a bandpass filter 3 in front of the first (variable) amplifying stage 4. Bandpass filter 3 may be part of a dual filter arrangement, such as a duplexer that includes a transmit side bandpass filter as well. The filtered received signal is then downconverted with mixers 5 to a zero IF (i.e. direct conversion) using quadrature local oscillator (LO) signals 6 (in 90 degree phase shift) that are tuned with a receiver RX synthesizer 7 at the carrier frequency of the received channel. After downconversion the signal is applied to baseband amplifiers 8, 10 and filters 9, and in a digital communications systems the information is converted into digital form with an analog-to-digital converter (A/D) 11 and is then further digitally filtered 12. Channel decoding 13 and other necessary digital functions to recover the transmitted information are be performed after the A/D 11. Gain control is an important function to extend the input signal range in all receiver architectures, and is used as well in the instant invention to adjust signal levels during calibration. A Received Signal Strength Indicator (RSSI) block 14 provides a signal to a gain control logic block 15 that functions to adjust the gain of the amplifiers 4, 8 and 10 to maintain the received signal at a desired level. It should be noted that the foregoing description of a conventional direct conversion receiver architecture is not to be read or construed in a limiting sense upon the practice and scope of the embodiments of this invention, as it will be described further below.

Fig. 2 A shows the construction of a typical differential low noise amplifier (LNA), having a single resonator (Zjj, and Z_T^ ), that is used as the first amplifier 4 in the receiver chain, while Fig. 2B shows a typical voltage controlled oscillator (VCO) 7A used in the synthesizer 7. The VCO 7A generates the high-frequency signal from which the quadrature LO signals are generated. Both of these devices use a resonator that can be implemented on the RF IC or with an external tank circuit. Although the resonators in Figs. 2A and 2B appear slightly different, they perform electrically exactly the same resonance function. Generally, for the purposes of this invention all described resonators can be considered to be an LC tank circuit containing an inductor, a capacitor and a resistor. The resistor is not necessarily shown in all drawings, and in most cases the resistor is actually the parasitic resistance, which degrades the quality of the tank circuit in all physical realizations, and hence must be taken into account.

A wide-band LNA can be implemented according to Fig. 2A using a high-quality LC tank and a parallel separate resistor R (not shown) to enhance the bandwidth by lowering the quality factor (Q value) of the tank circuit. This technique is generally used in many LNA implementations, as the current IC technologies provide inductors and capacitors that produce too high a Q value for the tank circuit, if process variations are taken into account. This problem is illustrated in Fig. 3A. Due to process variations the center frequency can vary by too large an amount between samples in order to cover the entire band of interest (i.e., bandwidth of the system) without requiring tuning (calibration). Therefore, tuning is required in the case of a narrow-band LNA. In the case of a wide-band structure (shown by the dashed lines), the process variations have a much smaller effect on the amplification in the band of interest. The difference between the two approaches in attenuating the out-of- band interferer, such as the transmitter leakage, is shown in Fig. 3B. The benefit of using a narrow-band LNA is obvious, and even relatively small improvements in the attenuation can relax the receiver specifications significantly.

It can be shown that 6 dB attenuation in the maximum transmitter (TX) power leakage at some distance (in MHz) from the desired signal can relax the mixer 5 specification sufficiently so as to remove a filter from the LNA-mixer interface. The intermodulation of the TX leakage with an unwanted spurious signal is considered in this estimation. While a resonance circuit with sufficient performance can be implemented using current IC technologies, the accuracy of the resonant frequency is not acceptable without tuning. However, and as was described above, the requirement to provide tuning increases the cost, and is thus not desirable.

SUMMARY OF THE PREFERRED EMBODIMENTS

The foregoing and other problems are overcome, and other advantages are realized, in accordance with the presently preferred embodiments of these teachings.

This invention describes an apparatus, a method and an algorithm for providing a calibration technique that allows for the use of narrow bandwidth resonators in the RF signal path and thus reduces the linearity requirements of those circuit blocks that benefit from additional attenuation of the out-of-band interferers. The presently preferred calibration technique also allows the enhancement of linearity performance with certain tuning techniques, without increasing the power consumption. Because in most cases the linearity requirement dominates the power consumption, the calibration technique preserves the same linearity performance with a reduced current consumption, and thus contributes to reducing the overall power consumption in the system. Another important benefit realized by the presently preferred calibration technique is that when certain performance parameters are relaxed after frequency band limitation, it becomes possible to eliminate external filtering, while achieving acceptable performance with on-chip devices. One example of an important application for this technique is in the removal of external band selection filtering located between the LNA and the downconversion mixers, which are typically required due to the leakage of the transmitter power into the receiver input in full duplex systems having a simultaneous transmission and reception mode. The 3G CDMA systems are particular examples of such full duplex systems.

In accordance with this invention, logic is provided that changes the LNA resonator and detects the strongest response, as well as an algorithm that selects the resonance frequency based on the strongest response, and that can shift the resonance from the nominal according to the current LO signal. The invention also provides a method to shift the resonance 901

according to a current scaling of the LNA. The invention also provides an additional oscillator mode in which the output signal is amplitude and/or frequency modulated. The invention also provides an algorithm that tunes the linearity performance of the receiver to a maximum by using the signals, produced by the additional oscillator or LO synthesizer and modulator, and measures the results at baseband. The invention beneficially provides a tuning method for a narrow-band LNA structure using either one high Q resonator, or a plurality of resonators coupled together in various ways, for obtaining narrow-band performance or a larger tuning range for, as an example, multi-band applications. An additional oscillator is further provided that uses the LNA resonator in a calibration mode, preferably in combination with a phase lock loop.

In accordance with an embodiment of this invention, an additional high frequency signal with desired properties maybe connected to the LNA resonance circuit and the input stage of the LNA is switched off. By tuning the resonance frequency (and possibly the LO frequency) the maximum output signal level of the RF front-end of the receiver is detected. This maximum level indicates that the resonance of the LNA is approximately the same as with the known LO frequency. With this setting the resonance is brought closer to the band of interest and any process variations that affect the resonance frequency are thereby compensated. By using this calibration procedure a relatively narrow bandwidth LNA can be employed in the receiver to filter out the out-of-band interferers, such as the own- transmitter leakage power when operating in the full duplex mode.

The tuning can also be accomplished without using an additional high frequency signal source, by detecting the maximum gain of the receiver from noise generated internally in the receiver. That approach is possible, in a typical case, when a significant part of the noise in the receiver is generated before the resonator, and thus the control word producing the strongest signal defined by the noise indicates the maximum gain, i.e., the desired resonant frequency.

The calibrated LNA resonance may also be tuned during reception based on the synthesizer frequency to which the receiver is to be tuned. In that the transmitted signal in the case of full duplex communication is located typically at a fixed separation in frequency as compared to the reception, the gain of the LNA with respect to the TX attenuation can be optimized for all frequency channels separately if the resonance of the LNA can be shifted according to the received radio channel. This property is beneficial in, for example, the full duplex 3G WCDMA system.

The resonance frequency may also change slightly if the LNA biasing current varies. However, because the process variations of the resonator are calibrated for by the use of this invention, the known frequency shift as a function of the biasing current can be also taken into account when the bias of the LNA is changed.

A narrow-band LNA can be established using a high-quality tank circuit in the resonator load. However, current IC technologies do not provide very high Q- values for the inductors using standard processing steps, and thus restrict the bandwidth limitation capabilities of the LNA. To circumvent this problem an LNA structure that uses two resonators is employed in order to reduce the LNA bandwidth. A narrow-bandwidth LNA can also be constructed using one high Q value resonator (or with a plurality of interconnected resonators) for obtaining narrow band performance or a large tuning range. In accordance with an aspect of this invention, a procedure is provided for tuning the one or more resonators during calibration and during operation of the LN A/receiver.

In order to maximize the linearity of the receiver, the calibration signal can be amplitude or frequency modulated. The amplitude modulated (AM) signal can be used for IIP2 tuning in the receiver while the frequency modulated (FM) signal can be used for IIP3 tuning.

This embodiment of the invention thus provides circuitry and a method to generate suitable test signals for enabling internal receiver tuning, to detect the results and to perform an internal calibration cycle in the radio receiver to achieve an optimum level of performance.

The calibration is preferably performed during idle time slots in a TDMA communications system, such as GSM, and/or during the start up of the mobile station, and/or during any other available idle time in any other radio system. Because the calibrations performed are related mainly to characterizing component process variations, they maybe performed only once. However, the calibration can be performed as often as desired, so long as the LNA of the receiver can be switched off or at least partially disabled from operation. For example, the calibration procedure can be performed in a multi-mode or multi-band receiver when the particular receiver front-end is not in use. One important benefit that is derived from the use of the teachings of this invention is that it becomes possible to use a relatively narrow-band LNA in the radio receiver, and to then perform the calibration internally to the RF IC. The narrow-band LNA attenuates more effectively the out-of-band interferers, which enables the linearity requirements of the receiver blocks following the LNA, as well as the LNA itself, to be relaxed. This has particular significance in the full duplex case where reception occurs simultaneously with transmission, and where the transmitter power leaks into the receiver input, h that linearity performance requirements typically dominate the receiver power consumption, by relaxing the linearity performance requirements, the average power consumption can be reduced significantly.

Another important benefit that is derived from the use of the teachings of this invention is that due to the large leaked transmitter power an external filter is typically needed between the output of the LNA and input to the mixer, even in the direct conversion receiver, otherwise the mixer linearity requirements become unreasonable. However, by enabling the use of a narrow band LNA, in combination with the calibration procedure therefor, the large and expensive external filter can be eliminated, and the linearity requirements of the mixer can be relaxed as well. This results in both cost and power consumption savings.

The use of this embodiment of the invention also provides the possibility to employ the same oscillator used in the resonance calibration for tuning the IIP2 of the receiver by a simple amplitude modulation method in the oscillator. In this case there is potentially no need to calibrate the IIP2 during fabrication. The IIP3 can also be maximized by correctly biasing or switching certain structures. This invention provides an opportunity to tune or calibrate IIP3 performance internally in order to optimize the circuit performance. In general, the internal optimization of IIP3 can lead to reduced power consumption. A two- tone test signal for the IIP3 optimization can be generated either internally or externally.

A first embodiment of this invention provides a RF receiver of a communications equipment, such as a mobile station or a base station, that has calibration circuitry that operates under the control of a data processor of the communications equipment for use in calibrating RF circuitry of the communications equipment in the field, as opposed to only calibrating the circuitry in a factory in a production and/or testing environment. Furthermore, all of the disclosed calibrations can also be performed during production. In this case no external test signals are required to be connected to the device under test, leading to simplifications and increased efficiencies during the fabrication test cycle.

A method is disclosed for operating a RF receiver of a communications equipment, as is circuitry for implementing the method. The method includes, while operating under the control of a data processor of the communications equipment, generating a calibration signal; injecting the calibration signal into a low noise amplifier (LNA) of the RF receiver; measuring a downconverted response of the receiver at a plurality of different internal states of the receiver using the calibration signal at more than one frequency, or measuring the downconverted response of the receiver at a plurality of different LNA tuning combinations using a fixed calibration frequency, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response so as to compensate at least for variations in component values that comprise the at least one resonator, or adjusting the linearity of the receiver.

The receiver internal states are preferably those associated with an adaptive receiver wherein receiver internal blocks including, but not limited to, the bias current(s), gain(s) and linearity can be tuned or adjusted, resulting in some state or performance of the receiver.

In a further aspect this invention operates a mobile station data processor, during a time that a receiver is not required, for at least partially disabling a receiver low noise amplifier, for generating a calibration signal within the mobile station and coupling the calibration signal into the receiver, measuring a downconverted response of the receiver to the calibration signal, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response, or adjusting the linearity of the receiver chain.

This first embodiment of the invention operates a mobile station data processor, during a time that a receiver is required, for at least partially enabling the receiver low noise amplifier, for generating the calibration signal within the mobile station and coupling the calibration signal into the receiver, measuring a downconverted response of the receiver to the calibration signal, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response, or adjusting the linearity of the receiver chain. The downconverted calibration signal is preferably located outside of a receiver passband transfer function so that the calibration signal is not totally rejected, and preferably the downconverted calibration signal is separated from the received signal spectrum by bandpass filtering in the digital domain.

Further in accordance with a presently preferred, second embodiment of this invention there is described an algorithm that selects the LNA load resonance frequency based on a strongest response/gain, and that can shift the resonance, according to a current LO signal, from the nominal value.

Also described is an algorithm for tuning the second order linearity performance of the receiver to a maximum value by using signals, produced by the transmitter, and by measuring results at baseband.

An aspect of this invention is a novel technique to calibrate the LNA center frequency, as well as the second order non-linearity of the mixer.

In the second embodiment of this invention the calibration of the LNA center frequency is performed by measuring the noise power at the digital baseband. By tuning the resonance frequency (and possibly the LO frequency as well) the maximum output signal level of the RF receiver can be determined. The maximum output signal level indicates that the resonance and the peak gain of the LNA are at approximately the same frequency as the known LO frequency. By using this determined setting, the resonance is made to be sufficiently close to the band of interest, and the process variations of the resonance frequency are thereby compensated for. Using this calibration procedure a narrow-band LNA can be used in the receiver to filter out the out-of-band interferers, including transmitter (TX) leakage power as well as external signal sources.

In order to maximize the second order linearity in the direct conversion receiver, the calibration signal is preferably amplitude modulated (AM). The TX leakage signal, in systems that have a duplexer and an AM component in the used modulation technique, can be used as the calibration signal for IIP2 tuning directly. If an antenna switch is used, the attenuated TX signal can be connected to the input node, or to some other RF node of the receiver. In accordance with the second embodiment of this invention the inclusion of additional circuitry, such as the inclusion of calibration signal generators, is not required, thereby decreasing both cost and complexity, while conserving IC area and reducing power consumption.

The presently preferred embodiments of the invention calibrate the center frequency of the LNA 4 by using only the thermal noise of the background radiation, duplexer and balun, and allows the use of a narrow-band LNA in a radio receiver that can be calibrated in situ. This calibration involves isolating the LNA input from external interference that could disturb the calibration either by opening a switch to isolate the LNA input or by shielding to isolate the antenna or mobile station from receiving any interfering signals.

The same circuitry can also be used also to tune the linearity of the receiver by powering up the TX signal and measuring the IMD2 product, or some other impairment in the frequency band of interest. An embodiment enables the tuning of the IIP2 without requiring an external test signal. Furthermore, power measurements (RSS) can be performed using existing digital measurement blocks, and no additional IC area is required. The calibration procedures (LNA and mixer) need not occur at the same time as the other measurements during fabrication.

Aspects of this invention are directed, at least in part, to generating or using pre-existing test signals for internal tuning, to detecting the results, and to performing an internal calibration cycle in the radio receiver to achieve an optimum level of performance.

Because the calibrations are related mainly to process variations the receiver may need to be calibrated but once, and the calibration parameters stored for later use. Ideally the calibration is performed once during production, and the settings are stored into non- volatile memory. Also, it is possible to carry out the calibration during idle time slots in TDMA communications (such as EDGE/GSM), or during the start up of the mobile station, or during any other available idle slot or time in any other radio system. For example, the calibration procedure can be performed in the multi-mode or multi-band receiver when the particular receiver front-end is not in use.

BRIEF DESCRIPTION OF THE DRAWINGS The foregoing and other aspects of these teachings are made more evident in the following Detailed Description of the Preferred Embodiments, when read in conjunction with the attached Drawing Figures, wherein:

Fig. 1 is a simplified block diagram of a conventional direct conversion receiver;

Figs.2A and 2B, collectively referred to as Fig. 2, are schematic diagrams of a conventional differential common emitter low noise amplifier (LNA) having two resonators, and a conventional voltage controlled oscillator (VCO), respectively;

Figs. 3 A and 3B, collectively referred to as Fig. 3, are graphs wherein Fig. 3 A shows the effect of process variations on a narrow bandwidth LNA, contrasted with a wide bandwidth LNA; and Fig. 3B shows the effect of the narrow band LNA on the attenuation of an out-of- band interferer, such as leaked transmitter power;

Fig. 4 is a block diagram of a direct conversion receiver having calibration circuitry in accordance with this invention, where the switch indicates possible connection points or nodes of the calibration signal;

Fig. 5 is a block diagram of the direct conversion receiver having LNA center frequency calibration circuitry using an amplitude modulated (AM) local oscillator (LO) signal in accordance with an aspect of this invention;

Fig. 6 is a schematic diagram showing one embodiment for coupling the calibration signal to a LNA resonator, where the LNA is in the calibration mode where the input stage is switched off;

Fig.7 is a logic flow diagram that illustrates an embodiment of a resonator tuning algorithm;

Fig. 8 shows a dual resonator LNA embodiment;

Fig. 9 is a graph that depicts resonance frequency as a function of supply current for the LNA, with and without digital tuning; Fig. 10 is a logic flow diagram that illustrates an embodiment of second-order input intercept point (HP2) calibration of a direct conversion receiver;

Fig. 11 is a schematic diagram showing an embodiment of an auxiliary VCO coupled to the LNA resonator in accordance with an embodiment of this invention;

Figs. 12A and 12B, collectively referred to as Fig. 12, are schematic diagrams of first and second alternative embodiments, respectively, of an auxiliary VCO coupled to the LNA with two resonators;

Fig. 13 is a block diagram of an LNA tuning function that uses a frequency divider, phase- frequency detector, counter and associated logic, where the LNA is assumed to be in oscillation with an additional oscillator core;

Fig. 14 is a block diagram of a wireless communications system that includes a mobile station with a direct conversion receiver that is operated in accordance with this invention;

Fig. 15 is a logic flow diagram that illustrates an embodiment of a resonator tuning algorithm in accordance with a further aspect of this invention, where the Received Signal Strength (RSS) measurement may be the sum of a plurality of measurements in the same state;

Fig. 16 is a logic flow diagram of a presently preferred method to calibrate the second-order input intercept point (IIP2) of a direct conversion receiver, where the RSS measurement may be the sum of a plurality of measurements in the same state;

Fig. 17 is a graph that plots integrated power versus the LNA tuning word, and shows the LNA center frequency tuning measured from the RSS indicator, with only noise present, and further illustrates that the curve follows the gain of the LNA, where the desired calibration result is one that corresponds to the maximum power (MaxP);

Fig. 18 is a graph that plots RSS versus the mixer tuning word, and shows the how the mixer IIP2 is tuned utilizing the TX leakage, where the TX modulation has an AM component, where the upper curve shows the I channel mixer calibration when the Q channel mixer is un-calibrated, while the lower curve shows the Q mixer calibration after the I mixer is calibrated, and where the desired calibration result is one that corresponds to the minimum power (MinP);

Fig. 19 is a block diagram of a direct conversion receiver that includes tuning LNA and mixer tuning logic in accordance with this invention; and

Fig.20A shows an embodiment of the mixer having tunable load resistances (R_T1 and R_T2), and Fig. 20B shows an embodiment of one of the tunable load resistances.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

In order to place this invention into a suitable technological context, reference is first made to Fig. 14 where there is illustrated a simplified block diagram of an embodiment of a wireless communications system that includes at least one mobile station (MS) 100 that is suitable for practicing this invention. Fig.14 also shows an exemplary network operator having, for example, a GPRS Support Node (GSN) 30 for connecting to a telecommunications network, such as a Public Packet Data Network or PDN, at least one base station controller (BSC) 40, and a plurality of base transceiver stations (BTS) 50 that transmit in a forward or downlink direction both physical and logical channels to the mobile station 100 in accordance with a predetermined air interface standard. Each BTS 50 is assumed to support a cell. A reverse or uplink communication path also exists from the mobile station 100 to the network operator, which conveys mobile originated access requests and traffic.

The air interface standard can conform to any suitable standard or protocol, and may enable both voice and data traffic, such as data traffic enabling Internet 70 access and web page downloads. One suitable type of air interface is based on TDMA and may support a GSM or an advanced GSM protocol, although these teachings are not intended to be limited to TDMA or to GSM or GSM-related wireless systems, hi fact, another wireless system and air interface, such as a WCDMA system, may serve at least a part of the geographical area served by the wireless communication system shown in Fig. 4, and the MS 100 may be a multi-band terminal that is capable of operating with either the GSM or the WCDMA network, or it may be a single-band terminal that operates only with the WCDMA system. The network operator may also include a suitable type of Message Center (MC) 60 that receives and forwards messages for the mobile stations 100. Other types of messaging service may include Supplementary Data Services and one under currently development and known as Multimedia Messaging Service (MMS), wherein image messages, video messages, audio messages, text messages, executables and the like, and combinations thereof, can be transferred between the network and the mobile station 100.

The mobile station 100 typically includes a microcontrol unit (MCU) 120 having an output coupled to an input of a display 140 and an input coupled to an output of a keyboard or keypad 160. The mobile station 100 may be a handheld radiotelephone, such as a cellular telephone or a personal communicator. The mobile station 100 could also be contained within a card or module that is connected during use to another device. For example, the mobile station 100 could be contained within a PCMCIA or similar type of card or module that is installed during use within a portable data processor, such as a laptop or notebook computer, or even a computer that is wearable by the user.

The MCU 120 is assumed to include or be coupled to some type of amemory 130, including a non-volatile memory for storing an operating program, as well as a random access memory (RAM) for temporarily storing required data, scratchpad memory, received packet data, packet data to be transmitted, and the like. A separate, removable SIM (not shown) can be provided as well, the SIM storing, for example, a preferred Public Land Mobile Network (PLMN) list and other subscriber-related information. The non- volatile memory is assumed, for the purposes of this invention, to store a program enabling the MCU 120 to execute the software routines, layers and protocols required to operate in the wireless communications system, as well as to provide a suitable user interface (UI), via display 140 and keypad 160, with a user. The stored program also is operable for executing the methods and algorithms in accordance with these teachings, such as and including the methods illustrated in Figs. 7 and 10. Although not shown, a microphone and speaker are typically provided for enabling the user to conduct voice calls in a conventional manner.

The mobile station 100 also contains a wireless section that includes a digital signal processor (DSP) 180, or equivalent high speed processor or logic or control unit, as well as a wireless transceiver that includes a transmitter (Tx) 200 and a receiver (Rx) 220, both of which are coupled to an antenna 240 for communication with the network operator via the BTS 50. At least one local oscillator (LO) 260 forms a part of a frequency synthesizer and is provided for tuning the transceiver. Data, such as digitized voice and packet data, is transmitted and received through the antenna 240.

Of most interest to the teachings of this invention is the receiver 220. The receiver 220 is assumed to be a direct conversion receiver, although in other embodiments other types of receivers, such as superheterodyne receivers, could be employed. The receiver 220 includes LNA and mixer calibration functions in accordance with one or more of the various embodiments of this invention, as will now be described in further detail.

Fig. 19 shows an embodiment of the receiver 220. Note that a number of blocks are the same as the conventional direct conversion receiver of Fig. 1. However, also shown in Fig. 19 is an optional antenna switch 2A, and the bandpass filter 3 A in front of the LNA 4 that is part of the dual filter arrangement given by the duplexer 3 that includes a transmit side bandpass filter 3B. Fig. 19 also shows a tuning logic block 16 that receives the output of the RSSI block 14 (the value of the measured RSS) and that provides tuning outputs to the mixers 5 and the LNA 4 in accordance with this invention. The transmitter circuitry 210 is shown generally as a single block . The combination of the transmitter 210 and receiver 220 forms a transceiver.

The tuning logic block 16, in the preferred embodiment of this invention, can be a part of, or a software function executed by, the MCU 120 of Fig. 14.

This invention allows the use of a narrow-band LNA 4 in the radio receiver 220 by providing a convenient mechanism and method to calibrate the LNA. The calibration method can be done using an internal calibration procedure. While some calibration typically is performed after RF IC fabrication, the use of this invention can eliminate the requirement to provide external test signals and/or equipment during mobile station production. The calibration can also be performed later if desired, during idle operation modes and possibly during reception as well. Additionally, the calibration does not require another switch to connect the calibration signal to the LNA 4.

For the purposes of this invention a wide bandwidth LNA 4 is one that is suitable for operation over a wide frequency bandwidth, whereas a narrow bandwidth resonator will typically have a higher Q value than the wide bandwidth resonator.

The same circuitry can be used also to tune the linearity of the receiver 220 by modulating the LO/auxiliary VCO signal, or by directly using the transmitter 210 as a source. This also permits the tuning of LTP2 without requiring external test signals, and also provides a technique to optimize HP3.

While the calibration procedure consumes some amount of power during calibration, since the calibration needs to be performed either only once or at most infrequently, the increment in the average power consumption is negligible.

When conducting the calibration procedures any signals received through the antenna 2 to the input of the LNA 4 are preferably isolated or shielded sufficiently so that they do not have any impact on the calibration procedure. Especially during the IIP2 calibration, shown in Fig. 16, when the resulting 1MD2 is at a very low level any signal close or above the thermal noise level will disturb the calibration procedure.

Referring to Fig. 4, in accordance with a first embodiment of this invention an internally generated test or calibration signal (CS) is connected to the LNA 4 resonator in order to calibrate the receiver 220. The calibration signal can be taken from the synthesizer 7 as shown in Fig. 4, or from an auxiliary oscillator as discussed below. The calibration signal is first attenuated in attenuator 20, as the typical signal levels output by the synthesizer 7 would easily saturate the signal path of the receiver 220. The calibration signal can be located at the same frequency as the LO 6 of the downconversion mixers 5, or at some other frequency. In the case of the direct conversion receiver, as shown in Fig. 4, the former alternative is preferred as the same signal used for the LO 6 can be used as well for the calibration signal CS. However, the calibration signal is preferably modulated by modulation block 22 such that the calibration signal is not located exactly at DC (or as a passband signal in general) after downconversion by mixers 5. This is so because the DC component is typically filtered out, and at least the accuracy of the calibration would suffer from other interferences at DC. The calibration signal CS can be amplitude modulated (AM), phase modulated (PM), frequency modulated (FM), or data modulated. The last alternative (data modulation) implies that the calibration signal can be modulated with some known data sequence that can be detected in the receiver 220. The frequency of the modulated calibration signal is preferably located within the passband of the device that is measuring the response. In the preferred mode, the same signal path as the actual reception signal path is used for the calibration, although in other embodiments a separate, parallel signal path could be used, hi the case of at least the preferred mode the calibration signal is modulated with a signal that results in the generation of frequency components that fall within the passband of the channel selection filter.

A receiver 220 embodiment, according to this invention, that uses an AM modulator 22 is shown in Fig. 5. The AM modulation is formed by switching the LO signal on and off with a digital clock signal of frequency f₀. The AM modulated signal is downconverted by the RF mixers 5 so that only frequency f₀, having an amplitude proportional to the amplitude of the calibration signal after the LNA 4, is detected by the RSSI block 14 or by some equivalent voltage/current/power detector.

Fig. 6 shows one embodiment of the LNA 4 and the connection of the calibration signal (CS) to the LNA resonator. In this embodiment, and contrasting same with the prior art example shown in Fig. 2A, it can be seen that the input transistors Q_1P and Q_1M are paralleled with calibration signal input (CAL_p and CAL_m) transistors Q_3p and Q_3m, respectively. By connecting the calibration signal to the transistors Q_3p and Q_3m, the use of a calibration switch upstream form the LNA 4 is avoided. This can be beneficial in some cases, as the presence of the calibration switch may increase the complexity and decrease the performance of the receiver 220. Note the two LNA resonators, where the first resonator is comprised of C_Lp, L_φ, C_Lm, L_m, C_t (and associated parasitic resistances) and the second resonator is comprised of L_r2 and C_a (and associated parasitic resistances). C_t and C^ represent switchable tuning capacitors for tuning the resonance frequency of the first and second LNA resonators. The bases of Q_2p and Q_2m are normally connected to a suitable DC bias source.

Fig. 8 shows an embodiment of the LNA 4 having two resonators, where Z^ and Z_m form the first resonator, and where variable (tuning) capacitance C_β and L_r2 form the second resonator. Variable capacitance C_t can also be used for taring purposes.

The two resonators shown in the embodiments of Figs.6 and 8 can be tuned separately. The overall transfer function is the combination of the transfer functions of the two resonators. hi the case of an unmodulated signal, or if the receiver 220 uses some intermediate frequency (IF), the calibration signal is not located at the same frequency as the LO signal. Hence, an additional oscillator is preferred for use. An embodiment that combines the LNA 4 and the VCO for this purpose is described below. However, in the case of IIP2 (or IIP3) trimming the calibration signal (CS) is preferably located outside of the passband of the channel selection filter. That can be done, for example, by using a higher frequency in the modulated signal or preferably, in the case of I1P2, by using the TX local oscillator signal from the synthesizer 7 in Fig. 4. This signal is also strong enough, and typically at a convenient distance from the reception frequency band of interest, that it can be used as a carrier signal for appropriate modulation. An AM component can be also added to the TX signal. The calibration signal (CS) can be fed to the circuit from the input, output or some interstage node of the LNA 4, as shown by the placement of the switch SW in Fig. 4. In addition, the HP2 calibration can be realized by switching on and off the current source of an additional oscillator (see Figs. 11 and 12 that are described below), thereby producing a strong AM signal. In this case the downconverted oscillation frequency is outside of the channel selection filter, and the amplitude of the oscillation signal may be controlled by the current source of the oscillator.

An embodiment of the calibration algorithm is shown in Fig. 7. hi the algorithm the resonator tuning range is swept over the total tuning range and the strongest response that can be measured using the RSSI block 14 of the receiver 220 defines the optimum setting. The sweep over the tuning range can be performed using all states or, more intelligently, by searching for an optimum state with a more efficient algorithm to increase the search speed. This configuration can then be stored in the memory 130 for use during reception. Preferably, the resonance tuning is performed at the middle-most radio channel. An important further embodiment provides the possibility to shift the tuning word of the resonator depending on the received radio channel to which the receiver 220 is being tuned. This means that once the process variations are calibrated for, it is possible to change the resonance frequency of the LNA 4 according to the received channel. In WCDMA, this has particular importance if the highest frequency channel centered at 2167.5 MHz is to be received. If the resonance frequency is located at the middle of the band i.e. 2140 MHz, some additional loss for the received channel will exist due to narrow bandwidth LNA 4. However, an additional benefit is realized due to the fact that the attenuation of the TX leakage at 1977.5 MHz will significantly increase due to the shifted resonance. This option is significant in all RF communications systems having a wide bandwidth and full duplex operation.

Discussing Fig. 7 in greater detail, at Step A the calibration procedure is started, and at Step B the LNA input stage(s) are switched off. In the example of Fig. 6 these would be the transistors Q_lp and Q_lm. At Step C the calibration signal (CS) is switched on, and a loop defined by Steps D, E and F is then entered where the frequency of the calibration signal, or the resonator center frequency, or the oscillation frequency of the additional oscillator (see Figs. 11 and 12) is swept over the LNA 4 tuning range. The downconverted response is measured at several points, and the process is repeated until all measurements are made (and stored in the memory 130.) At Step G the strongest response is selected, the calibration signal is switched off at Step H, and the resonance frequency of the LNA 4 is set accordingly at Step I. At Step J the calibration procedure terminates with the LNA 4 properly calibrated so as to compensate for variations that occur in the passive and other components in the RF IC.

Note that the method can include changing the resonance frequency after calibrating, and during normal operation, based on a current local oscillator frequency. That is, the center frequency of the LNA resonator is changed based on the operational channel. In this manner the rejection of the transmitted signal in frequency division duplex (FDD) systems, such as the WCDMA system, is maximized.

Further in accordance with this invention, for the case where the LNA 4 has two resonators, a procedure can be performed to improve the out-of-band attenuation performance and internal linearity. Such a LNA 4 structure is shown in Fig. 6. The basic LNA structure is similar in some respects to the one shown in Fig. 2A. However, this improved LNA 4 has a tuning mechanism that includes tunable capacitors C_t and C_a, and provides for tuning both the resonance frequency and the supply current. Because both resonators, made up of C_Lp, L_jp, Ci_jn, L_JΠJ, C_t (and associated parasitic resistances) and L_r2, C^ (and associated parasitic resistances), interact with one another in this topology it is possible to use one capacitor for resonance frequency tuning during the calibration, and the other capacitor to fine tune the resonance when the supply current is varied. The latter option can be used in adaptive receivers, and is of considerable importance in narrow bandwidth LNAs as changes in power supply current are known to also shift the resonance frequency slightly. In the preferred mode, the capacitor C_t is used for coarse tuning during calibration and the capacitor C_a is used for fine tuning when the supply current is changed. However, other functionality of the capacitors C_t and C_a is possible and operable in practice.

Fig. 9 shows the effect of fine-tuning, as a function of power supply current, for the resonance frequency. The resonance frequency can be seen to shift by almost 100 MHz over the supply current range without the use of fine-tuning. With fine tuning, however, the resonance frequency can be maintained to within +15 MHz, an amount of frequency shift that is acceptable in most applications. The fine tuning the resonance to compensate for variations in power supply current can use predetermined information, or it can performed by executing the calibration procedure at different power supply current levels.

Fig. 10 shows an embodiment of the IIP2 calibration algorithm for the direct conversion receiver 220, and may use the calibration arrangement shown in Fig. 4, wherein the modulated signals that are used to tune the IIP2 are outside of the channel selection filter. The main differences as compared to the algorithm for searching for the resonance frequency (Fig. 7) are the use of an amplitude modulated out-of-band signal and the search for the weakest, as opposed to the strongest, response. The latter is due to the fact that the HVLD2 product is an unwanted effect.

More particularly, in the receiver 220 the baseband signal processing blocks 8, 9 andl 0 have an ideally infinite and a typically very high common-mode-rejection ratio (CMRR), and thus only the signal that is differential can pass through the chain. In principle, the IMD2 product is a common mode signal that can be rejected totally with an infinite CMRR. In practice, the process variations of the mixer 5 and baseband blocks following it will produce a small differential signal that originates from the common mode IMD2 product. This signal is undesirable and can be minimized by using the tuning procedure in accordance with this invention. By tuning the mixer 5 in such a way that the EV1D2 products are again common mode, the CMRR of the analog baseband chain can suppress these signals automatically. One advantage of this tuning method is it's simplicity, as no additional circuit blocks at the analog baseband side are needed.

If several settings during the tuning cause the smallest response, then the measurement accuracy may not be sufficient. In that case it is possible to increase the measurement accuracy by increasing the signal level of the AM tone, or increasing the gain of the signal path. In the case of signal saturation of the receiver output, the opposite procedure can be performed during calibration. One possible LTP2 tuning technique is described in EP 0951138A1, "Method for reducing envelope distortion in radio receiver" by K. Kivekas & A. Parssinen, application date 18.04.2001, incorporated by reference herein in its entirety. This embodiment of the present invention extends the use of those techniques and presents a new calibration algorithm.

I

Discussing Fig. 10 now in greater detail, at Step A the IIP2 calibration procedure is started, and at Step B the LNA input stage(s) are switched off. At Step C the AM calibration signal (CS) is switched on, as shown in Fig. 4, and the loop defined by Steps D, E and F is then entered where the 1TP2 of the receiver 220 is tuned, the downconverted response is measured at different settings, and the process is repeated until all measurements are made. At Step G the weakest response is selected (the AM calibration signal is switched off) and an IIP2 tuning code is sent to the receiver 220 at Step H. At Step I the calibration procedure terminates with the ILP2 properly calibrated.

As was discussed earlier, instead of the synthesizer 7 an additional or auxiliary oscillator, which uses the same resonator as the LNA 4, can be used to generate the calibration signal (CS). Fig. 11 shows a preferred mode of implementation if a single resonator is used, while Fig. 12 shows two different embodiments of the LNA 4 with two resonators. The oscillator 4A uses the same resonator as the LNA 4, and is formed from cross-coupled transistors M_lp and M_lm and current source I_VCo The input stage(s) of the LNA 4 are switched off during the operation of the oscillator. Additionally, and as is shown in Fig. 13, the oscillator 4A can be connected to a phase lock loop (PLL) 30 formed using a divide by N frequency divider 32 that is connected to the output of the LNA 4, a phase detector (PD) 34, logic 36 to tune the capacitor switches of the resonator and a counter 38. This embodiment of the invention can be used for LNA 4 center frequency tuning, as the oscillation frequency of the auxiliary oscillator 4A is the same as the center frequency of the LNA 4.

The logic 36 may operate as follows: the logic 36 counts the up and down pulses of the PDF 34 separately and decides if the signal is lower or higher in frequency than the target frequency by comparing the output of the counters 38. If up counter has a larger count than the down counter, the decision is ' go up' and vice versa. The PLL 30 would thus oscillate between the two tuning words that are both acceptable. Other possibility is to begin from the lowest (or highest) frequency and count the tuning words up (or down) until the state of the phase detector 34 changes. In this case the logic 36 may count up and down pulses. If the system starts from the lowest frequency the result of the calibration is the state where the number of the counted down pulses exceeds the number of counted up pulses. The correct frequency is then one between the last tuning word that does not change the state of the phase detector 34 and the tuning word that does change the state of the phase detector 34. If the resonator includes an analog capacitance the tuning voltage is preferably set into the middle of the analog tuning range. Additionally, this method can be used to tune the local oscillator of the synthesizer 7.

A further, non-limiting embodiment for tuning the VCO is described in a commonly assigned patent application "Self-contained tuning of the VCO center frequency", by Pauli Seppinen and Kalle Asikainen, U.S. Patent Application Serial No.: 10/024,084, filed 17 December 2001, and incorporated by reference herein. In this approach there is a method for tuning an adjustable oscillator having at least one resonance circuit. The frequency of the oscillator is adjusted by changing the resonance frequency of the resonance circuit by means of a control signal for which a minimum value and a maximum value are selected. In the execution of the method at least one target value is selected for the control signal, the frequency of the adjustable oscillator is adjusted to substantially correspond to the target value, and the value of the control signal and the target value are compared. When the value of the control signal is substantially different from the target value, a tuning signal is produced to change the resonance frequency of the at least one resonance circuit.

The nodes to which the auxiliary oscillator is connected in Figs. 11 and 12 (collectors of Q_2p, Q_2m or the emitters of Q_2p, Q_2m) also illustrate examples of the nodes to which the other calibration signals can be connected in the LNA 4.

The foregoing embodiments of this invention can be implemented with just a few additional active components including RF transistors, possibly a simple AM modulator (a simple switching device in its basic form, as shown in Fig.5) and simple logic. While the invention could be implemented using external devices, the most preferred embodiments place the additional RF components on the same die as the RF front end. In any case, the IC area increase is very small. Very simple digital control logic is sufficient in this invention and the logic can be implemented with customized logic on the same die with any of the analog or mixed-mode chip, on the digital ASIC or programmable digital signal processor (DSP 180) or the central processing unit (CPU, such as the MCU 120). The memory requirements are also modest.

A description is now provided of further embodiments of this invention that do not require the inclusion of additional circuitry, or the inclusion of calibration signal generators, thereby decreasing both cost and complexity, while conserving IC area and reducing power consumption.

A presently preferred embodiment of the invention calibrates the center frequency of the LNA 4 by using only the thermal noise of the background radiation, duplexer andbalun, and allows the use of a narrow-band LNA in a radio receiver that can be calibrated in situ. The same circuitry can also be used also to tune the linearity of the receiver 220 by powering up the TX signal and measuring the LMD2 product, or some other impairment in the frequency band of interest. The tuning of the HP2, without requiring an external test signal, is thus made possible. Furthermore, power measurements (RSS) can be performed using existing digital measurement blocks, and no additional IC area is required. The calibration procedures (LNA and mixer) need not occur at the same time as the other measurements during fabrication.

In the presently preferred embodiments of this invention the noise generated by the duplexer 3, and other components connected in front of the LNA 4, are received by the LNA 4 (no changes are required to the receiver topology) in order to calibrate the receiver circuitry. While these embodiments may be used in all radio architectures, the discussion herein focuses on the direct conversion receiver architecture.

The performance of these calibration procedures can be facilitated in the field if the antenna switch 2 A of Fig. 19 is present so that the antenna 240 is isolated from the transceiver, and so that a resistive load R_Lis connected in place of the antenna 240. The transceiver in this case is isolated from external signals. The antenna switch 2A is not, however, required in order to employ the calibration procedures in accordance with the second embodiment of the invention. The basic concept is that the noise power received by the RSSI measurement block 14 is at it's maximum when the gain of the receiver 220 is maximized. An example of a measurement is plotted in Fig. 17. The center frequency can be tuned by tuning the LNA 4 resonator resonance frequency using C_t and C_a.

The second-order linearity calibration method can be performed without using any external calibration signals, as only the TX leakage signal is needed. In GSM systems the GMSK modulation cannot be used directly for the calibration, but the burst mode transmission can be readily used. In the case of EDGE, the 8-PSK modulation has strong AM component and thus this is suitable for use as a calibration signal for the ILP2 tuning. In WCDMA, the TX signal can be used as is. The second order linearity timing measures the 1MD2 products with the RSS indicator output from the RSSI block 14. By integrating the power within the channel band it is possible to distinguish small differences in the IMD2 levels and to determine the optimum linearity.

The second-order linearity calibration begins by taning on the TX circuitry 210. If the switch 2A between the antenna 240 and the filter 3 is used, it is switched to open state, where the antenna 240 is isolated from the receiver 220 and the resistive noisy element is connected to the input of the transceiver. A transmitted signal pattern can then be selected so that the AM signal component is maximized. It is also possible to transmit pseudorandom data for this purpose.

A presently preferred method to calibrate the mixer 5 linearity is to tune the load resistors of the mixer 5. In this case, changing the balance of the load resistors functions to linearize the mixer 5.

Reference can be made to Figs. 20A and 20B, where Fig.20A shows an embodiment of the mixer 5 having tunable load resistances (R_τl and R_T2), and Fig. 20B shows an embodiment of one of the tunable load resistances. In this case the tunable load resistance R_τ is embodied as a switched resistance network controlled by a 6-bit control word (MixRes(0)-MixRes(5)).

During operation, first one of the mixers 5 (I or Q) is calibrated. As is shown in Fig. 18, upper trace, the measured power curve reaches a minimum (note that there is still some remaining IMD2 power). The calibrated mixer 5 is thus set to the minimum TMD2 power setting. Next, the other (as yet uncalibrated) mixer is calibrated, and a similar EMD2 power minimization is performed. Again, the minimum power value corresponds to the optimum IIP2 result. At this point the ITP2 of the receiver is calibrated. The minimum of the measured curve can be determined using any suitable search algorithm utilizing one or several points during the calibration procedure.

Fig. 15 is a logic flow diagram that illustrates an embodiment of the resonator tuning algorithm in accordance with this second embodiment of this invention.

At Block A the calibration procedure is initiated, and at Block B a determination is made if the antenna switch 2A is present. If it is not, control passes to Block C to isolate the resonator from the air, such as by shielding, so as not to receive signals, while if the switch 2A is present control passes to Block D to isolate the antenna 240 from the LNA 4 by opening the switch 2A. At Block E the method sets a new LNA resonator state, such as by changing the digital control word for one or both of C_t and C_t2 for changing the resonant frequency of the LNA 4 resonators. At Block F the RSS is measured by the RSSI block 14 and recorded. Note that the RSS measurement may be the sum of several different measurements made at the same LNA 4 resonator setting. A check is made at Block G to determine if all of the I-channel settings have been tested (the Q-channel could be used instead of the I-channel). If not, control passes back to Block E to set the next LNA 4 tuning state, while if all settings have been tested control passes to Block H to select the LNA resonator setting that exhibits the largest measured RSS signal. At Block I this LNA 4 resonator setting is stored in non- volatile memory of the mobile station, and at Block J the calibration is completed.

It should be noted that the foregoing measurements and calibration of the LNA resonators is accomplished without injecting an external or an internal calibration test signal, but is accomplished instead using thermal noise from the circuitry prior to the LNA 4, such as the load resistance R_L of the antenna switch 2 A of Fig. 19, if present, or simply the thermal noise of the background radiation, duplexer 3 and balun, or generated internally in the LNA 4.

Fig. 16 is a logic flow diagram that illustrates an embodiment of the ITP2 calibration algorithm in accordance with this second embodiment of this invention. At Block A the calibration procedure is initiated, and at Block B the transmitter circuitry 210 is energized and the antenna 2 or the 220 receiver input is preferably shielded or isolated from disturbing signals. At Block C the method sets a new tuning state for the I- channel mixer 5, such as by changing the digital control word for the mixer load resistors. At Block D the RSS is measured by the RSSI block 14 and recorded. Note that the RSS measurement maybe the sum of several different measurements made at the same setting. A check is made at Block E to determine if all of the I-channel settings have been tested (the Q-channel could be tested first, instead of the I-channel). If not, control passes back to Block C to set the next mixer 5 tuning state, while if all settings have been tested control passes to Block F to select the mixer setting that exhibits the smallest measured RS S signal. The I-channel mixer is left at this setting while the Q channel mixer is tested (or vice versa, depending on which channel is tested first). At Block G the method sets a new timing state for the Q-channel mixer 5. At Block H the RSS is measured by the RSSI block 14 and recorded. Note again that the RSS measurement may be the sum of several different measurements made at the same setting. A check is made at Block I to determine if all of the Q-channel settings have been tested. If not, control passes back to Block G to set the next mixer 5 tuning state, while if all settings have been tested control passes to Block J to select the Q-channel mixer setting that exhibits the smallest measured RSS signal. At Block K the selected mixer settings are stored in non- volatile memory of the mobile station, at Block L the transmitter is deenergized (if appropriate), and at Block M the mixer IIP2 calibration is completed.

It should be noted here as well that the foregoing measurements and calibration of the mixers 5 is accomplished without injecting an external or an internal calibration test signal, but is accomplished instead using an AM modulated transmitter signal component that leaks through the duplex filter 3 into the front end of the receiver.

As was noted above, in order to obtain sufficient accuracy for the RSS measurement it is preferred to accumulate several measurements in the same tuning state. For ease in making the IIP2 tuning measurement it is also preferred to disable the input signal to RSS measurement block 14 from the I or Q branch that is not being calibrated.

The teachings of this invention can be applied to all digital radio systems. On the other hand, the benefits may vary between systems. For example, the use of the narrowband resonator is important in CDMA systems in order to establish the maximum attenuation for the TX leakage. On the other hand, in a GSM system it becomes possible to calibrate IIP2 internally. The feature may relax the calibration requirements during production, and/or it may add the possibility to use adaptivity in the mixers during operation. This is true because the ILP2 can be calibrated at a larger set of test points without adding production cost. This invention is suitable for use in both the mobile station 100 and in the base stations 50.

While described in the context of presently preferred embodiments of this invention, those having skill in the art should appreciate that variations in form and detail maybe made, and that all such changes will still fall within the scope of the teachings of this invention. For example, this invention should not be viewed as being limited to the specific LNA circuit embodiments shown in the Figures and described above, nor for use with only specific frequencies, ranges of frequencies, types of receivers, or with only certain wireless communications standards and protocols, such as only with GSM or WCDMA systems.

Furthermore, in some circumstances the calibration can be made during a reception time, instead of during a receiver idle time. In this case, and during a time that the receiver 220 is required, the data processor 120,180 of the mobile station 100 is operated to enable the LNA 4, as well as for generating the calibration signal within the mobile station 100 for the first embodiment described above, coupling the calibration signal into the receiver 220, measuring the downconverted response of the receiver 220 to the calibration signal, and then at least one of tuning the resonance frequency of at least one LNA resonator based on the measured downconverted response, or adjusting the linearity of the receiver chain. In this embodiment the downconverted calibration signal is preferably located outside of a receiver passband transfer function so that the calibration signal is not totally rejected. That is, the downconverted calibration signal is not totally rejected, but is instead rejected or attenuated 20dB/decade, if the filter or filters (e.g., filters 9 and 12) are first order filters. In this case the calibration signal is attenuated by a known amount as it passes through the receiver chain, and can thus be measured and characterized as described above. In this embodiment the downconverted calibration signal may be separated from the received signal spectrum by bandpass filtering in the digital domain.

It should be noted as well that the second embodiment of this invention pertains in part to a simple a method to perform the mixer 5 calibration that does not require separate calibration signal generators). It should be further noted that in some circumstances, such as in a WCDMA receiver, the linearity calibration may not be required if there is an external filter between the LNA 4 and the mixers 5. However, it becomes possible to remove the external filter if the mixer 5 linearity performance is adequate and, in this case, it becomes desirable to calibrate the mixers 5 in order to achieve an adequate level of IIP2. With this arrangement it becomes possible to remove a filter, thereby saving cost and circuit area. In this embodiment of the invention it has been shown that the calibration can be performed on-chip using an internally generated signal (such as the TX leakage signal). Therefore, the calibration procedure can be performed autonomously, which is preferable in a production environment, as no additional parts or external test equipment are required. Note that the disclosed calibration procedure differs significantly from the AM suppression test used in GSM receivers. In order to remove the external filter, the disclosed calibration technique is very desirable, especially since it can be shown that the transmitter 17 is a major source of second order nonlinearity in a direct conversion WCDMA receiver.

The LNA 4 tuning, on the other hand, can be performed in a straightforward fashion by receiving only white (e.g., thermal) noise that is generated internally in the receiver 220, and preferably by isolating or shielding the LNA 4 input from external signals. This approach enables the use of tunable RF structures with internal resonators, and is very beneficial when narrow band RF IC structures are used. Because of inevitable process-related spread in the resonant frequencies, this calibration procedure can be used to compensate the individual receivers for variations in resonant frequencies due to the process spread. These types of narrow band structures are beneficial, for example, in the case of TX leakage in the WCDMA receiver, and also to enable the removal of external filters.

Based on the foregoing description of the presently preferred embodiments, it can be appreciated that each has in common the internal (to the mobile station or the base station) generation of a calibration test signal or signals for use in calibrating the center frequency of the LNA, as well as for calibrating, if desired, the linearity of the receiver, such as by calibrating the mixer IIP2 characteristics. In one embodiment the internally generated calibration signal is sourced from a calibration signal generator component under control of a data processor, such as the MCU 120, within the mobile station 100, while in another embodiment the internally generated calibration signal can be white (thermal) noise inj ected into the input of the receiver chain, or a transmitter signal that leaks into the input of the receiver chain, again under control of the data processor 120 within the mobile station 100.

Claims

CLAIMSWhat is claimed is:

1. A method for operating a radio frequency RF receiver of a communications equipment, comprising:

under the control of a data processor of the communications equipment,

generating a calibration signal;

injecting the calibration signal into a low noise amplifier LNA of the RF receiver;

measuring a downconverted response of the receiver at a plurality of different internal states of the receiver using at least one frequency of the calibration signal; and

at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response so as to compensate at least for variations in component values that comprise the at least one resonator, or adjusting the linearity of the receiver.

2. A method as in claim 1 , where the calibration signal is generated using at least one of a frequency synthesizer of the communications equipment and an oscillator that comprises said at least one LNA resonator.

3. A method as in claim 1 , where the calibration signal is generated using at least one of noise and a transmitter signal that leaks into the receiver.

4. A method as in claim 1, where the resonance frequency is tuned based on a strongest measured downconverted response, and where the linearity is adjusted based on a weakest downconverted response.

5. A method as in claim 1, where the step of tuning a resonance frequency of the at least one LNA resonator comprises also fine tuning the resonance to compensate for variations in power supply current using one of predetermined information, or executing the calibration procedure at different power supply current levels.

6. A method as in claim 1, where generating the calibration signal comprises generating a modulated calibration signal, and where adjusting the linearity of the receiver comprises making an adjustment for either the second order input intercept point IIP2 or the third order input intercept point IIP3.

7. A method as in claim 1, where an output of the LNA is coupled to an input of a downconversion mixer, and where the step of measuring observes an output of a received signal strength indicator RSSI that is located downstream from the downconversion mixer.

8. A method as in claim 1 , where the receiver is a direct conversion receiver, where an output of the LNA is coupled to an input of a downconversion mixer, and where the calibration signal is modulated so as to avoid the generation of a DC or a passband signal at the output of the downconversion mixer during normal downconversion operation.

9. A method as in claim 1, where generating the calibration signal comprises attenuating the calibration signal.

10. A method as in claim 1, where injecting the calibration signal includes one of disabling a normal received signal input to the LNA, or electrically decoupling an antenna from the input of the LNA.

11. A method as in claim 1, where the communications equipment comprises a mobile station that operates in accordance with a TDMA protocol.

12. A method as in claim 1, where the communications equipment comprises a mobile station that operates in accordance with a CDMA protocol.

13. A method as in claim 1, where the RF receiver comprises a direct conversion receiver, and where the communications equipment comprises a mobile station that operates , in accordance with a CDMA protocol.

14. A method as in claim 1, where the communications equipment comprises abase station that operates in accordance with one of a TDMA protocol or a CDMA protocol.

15. A method as in claim 1, and further comprising changing the resonance frequency after calibrating, and during normal operation, based on a current local oscillator frequency.

16. A radio frequency RF receiver of a communications equipment, comprising circuitry that operates under control of a data processor of communications equipment for calibrating RF circuitry of said communications equipment, said calibration circuitry comprising a source of a calibration signal and circuitry for coupling a RF receiver calibration signal to a low noise amplifier LNA of said RF receiver; further comprising circuitry for measuring a downconverted response of said RF receiver at a plurality of different internal states of said receiver using at least one frequency of the calibration signal and for tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response so as to at least one of compensate at least for variations in component values that comprise said at least one resonator, or for adjusting the linearity of said RF receiver.

17. A RF receiver as in claim 16, where said calibration signal source comprises at least one of a frequency synthesizer of said communications equipment and an oscillator that in turn comprises said at least one LNA resonator.

18. A RF receiver as in claim 16, where said calibration signal source comprises at least one of a noise source and a transmitter signal that leaks into said receiver.

19. A RF receiver as in claim 16, where said resonance frequency is tuned based on a strongest measured downconverted response, and where said linearity is adjusted based on a weakest downconverted response.

20. A RF receiver as in claim 16, where said circuitry for tuning said resonance frequency also fine tunes said resonance to compensate for variations in power supply current using one of predetermined information or executing the calibration procedure at different power supply current levels.

21. A RF receiver as in claim 16, and further comprising a modulator for modulating said calibration signal, and where said circuitry for adjusting said linearity of said RF receiver comprises making an adjustment for either a second order input intercept point IIP2 or a third order input intercept point IIP3.

22. A RF receiver as in claim 16, where an output of said LNA is coupled to an input of a downconversion mixer, and where said measuring circuitry observes an output of a received signal strength indicator RSSI that is located downstream from said downconversion mixer.

23. A RF receiver as in claim 16, where said RF receiver is a direct conversion receiver, where an output of said LNA is coupled to an input of a downconversion mixer, and where said calibration signal is modulated to avoid generating a DC signal or a passband at an output of said downconversion mixer during normal downconversion operation.

24. A RF receiver as in claim 16, where said source comprises an attenuator for attenuating an output of a frequency synthesizer to provide said calibration signal.

25. A RF receiver as in claim 16, where when said calibration signal is coupled to said LNA a normal received signal input to said LNA is disabled.

26. A RF receiver as in claim 16, where said communications equipment comprises a mobile station that operates in accordance with a TDMA protocol.

27. A RF receiver as in claim 16, where said communications equipment comprises a mobile station that operates in accordance with a CDMA protocol.

28. A RF receiver as in claim 16, where said RF receiver comprises a direct conversion receiver, and where said communications equipment comprises a mobile station that operates in accordance with a CDMA protocol.

29. A RF receiver as in claim 16, where said communications equipment comprises a base station that operates in accordance with one of a TDMA protocol or a CDMA protocol.

30. A RF receiver as in claim 16, where said communications equipment changes the resonance frequency after calibrating, and during normal operation, based on a current local oscillator frequency.

31. A low noise amplifier for use in a radio frequency receiver chain of communications equipment, comprising at least one gain element and at least one resonant LC tank circuit coupled to said gain element, further comprising at least one further gain element coupled to said LC tank circuit for operating as an oscillator for generating a calibration signal for calibrating at least said low noise amplifier.

32. A low noise amplifier as in claim 31 , and further comprising a phase lock loop circuit coupled to said oscillator, where said oscillator output is set to a center frequency of said low noise amplifier.

33. A method for operating a radio frequency RF receiver of a communications equipment, comprising:

under the control of a data processor of the communications equipment,

generating an internal calibration signal;

injecting the internal calibration signal into a low noise amplifier LNA of the RF receiver;

tuning the LNA to a plurality of different tuning points and measuring a corresponding downconverted response of the receiver; and

at least one of taring a resonance frequency of at least one LNA resonator based on the measured downconverted response so as to compensate at least for variations in component values that comprise the at least one resonator, or adjusting the linearity of the receiver.

34. A method as in claim 33, where the calibration signal is generated using at least one of a frequency synthesizer of the communications equipment and an oscillator that comprises said at least one LNA resonator.

35. A method as in claim 33, where the calibration signal is generated using one of a noise source located upstream from said LNA and a transmitter signal, having an AM component, that leaks into the receiver through a duplex filter.

36. A method as in claim 33, where the resonance frequency is tuned based on a strongest measured downconverted response, and where the linearity is adjusted based on a weakest downconverted response.

37. A method as in claim 33, and further comprising changing the resonance frequency after calibrating, and during normal operation, based on a current local oscillator frequency.

38. A method for operating a mobile station comprising, during a time that a receiver is not required, operating a data processor of said mobile station for at least partially disabling a receiver low noise amplifier LNA, generating a calibration signal within said mobile station, coupling said calibration signal into said receiver, measuring a downconverted response of the receiver to said calibration signal, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response, or adjusting the linearity of a receiver chain.

39. A method for operating a mobile station comprising, during a time that a receiver is required, operating a data processor of said mobile station for enabling a receiver low noise amplifier LNA, generating a calibration signal within said mobile station, coupling said calibration signal into said receiver, measuring a downconverted response of the receiver to said calibration signal, and at least one of tuning a resonance frequency of at least one LNA resonator based on the measured downconverted response, or adjusting the linearity of a receiver chain.

40. A method as in claim 39, where the downconverted calibration signal is located outside of a receiver passband transfer function so that the calibration signal is not totally rejected.

41. A method as in claim 39, where the downconverted calibration signal is separated from the received signal spectrum by bandpass filtering in the digital domain.

42. A method for operating a radio frequency RF receiver of a communications equipment, comprising:

generating an internal calibration signal from a noise source;

tuning a resonance frequency of at least one LNA resonator based on a maximum measured value of the downconverted response so as to compensate at least for variations in component values that comprise the at least one resonator.

43. A method as in claim 42, where generating the internal calibration signal comprises decoupling an antenna from an input of the receiver and coupling a resistance to the input of the receiver, and where the noise source comprises the resistance.

44. A method as in claim 42, where generating the internal calibration signal comprises shielding or isolating at least one of the antenna and the at least one LNA resonator from interfering signals.

45. A method as in claim 42, where said noise source comprises any noise generated in the communications equipment prior to a component being calibrated.

46. A method as in claim 42, further comprising storing at least one digital value that represents an optimum value for tuning the center frequency of said LNA resonator.

47. A method for operating communications equipment to adjust the linearity of an RF receiver, comprising:

energizing a transmitter of the communications equipment;

leaking a part of a transmitter signal having an AM component into an input of the RF receiver and then into inputs of I-channel and Q-channel downconversion mixers;

adjusting a first one of the I-channel or Q-channel mixers at a plurality of different points and measuring a corresponding downconverted response of the receiver;

setting the first one of the mixers based on a minimum measured value of the downconverted response;

adjusting the other one of the I-channel or Q-channel mixers at a plurality of different points and measuring a corresponding downconverted response of the receiver; and

setting the other one of the mixers based on a minimum measured value of the downconverted response.

48. A method as in claim 47, further comprising storing digital values that represent optimum values for adjusting load resistances of each of the I-channel and Q-channel for calibrating at least the receiver second-order input intercept point.