GB2249442A

GB2249442A - Correction of quadrature phase error

Info

Publication number: GB2249442A
Application number: GB9023759A
Authority: GB
Inventors: George Hedley Storm Rokos
Original assignee: STC PLC; Northern Telecom Europe Ltd
Current assignee: STC PLC; Nortel Networks Optical Components Ltd
Priority date: 1990-11-01
Filing date: 1990-11-01
Publication date: 1992-05-06
Also published as: GB9023759D0

Abstract

A method of correcting quadrature phase error between in-phase (I) and quadrature (Q) components of an a.c. signal includes the steps of deriving from each component (OA, OB) a correction signal (AC, BD) and adding the correction signal to the other component, the amplitudes of the correction signals being such as to cause the two resultant vectors (OC, OD) to be in phase quadrature. A feedback arrangement measures the phase difference of the I and Q components, which phase difference is used to derive a phase adjustment signal. The a.c signal may be an r.f. local oscillator signal in a radio transmitter/receiver, or a baseband input signal in a zero IF transmitter or a baseband signal in a zero IF receiver. <IMAGE>

Description

CORRECTION OF QUADRATURE PHASE ERROR This invention relates to a method and means for the correlation of quadrature phase error in a.c.

signals and is of particular relevance to so-called "zero IF" or direct conversion" radio transmitters/receivers which may be incorporated in mobile telephone equipment.

A conventional method of providing quadrature signals, e.g. local oscillator in-phase (I) and quadrature (Q) components, is to feed a single original signal through respective RC and CR networks the outputs of which form the I & Q components. In practice it is difficult to fabricate such circuits in quantity while maintaining an accurate 900 phase shift between the I & Q components.

According to the present invention there is provided a method of correcting quadrature phase error between in-phase (I) and quadrature (Q) components of an a.c. signal including the steps of deriving from each component a correction signal and adding the correction signal to the other component, the amplitudes of the correction signals being such as to cause the two resultant vectors to be in phase quadrature.

Embodiments of the invention will now be described with reference to the accompanying drawings, in which: Fig. 1 is a vector diagram illustrating the method of the invention, Fig. 2 is a block diagram illustrating a quadrature phase error correction arrangement for an IF local oscillator in a zero IF radio equipment, and Fig. 3 is a circuit diagram of an implementation of the arrangement of Fig. 2.

In the diagram of Fig. 1 two r.f. nominal quadrature signals OA and OB of equal amplitude are shown with an initial phase difference of 900-s, where g is the phase angle error. To correct the phase angle of OA-OB we add: to OA a signal AC which is of amplitude OB x sin i/2 = OB tan 6/2 sin(90 +#/2) and to OB a signal BD which is of amplitude OA x sin 6/2 = OA tan 6/2.

cos 6/2 The resultant vectors OC and OD are both of amplitude IOA I cos and have a correct 900 phase cos g /2 difference.

A similar correction procedure can be used in the case of received signals after inaccurate quadrature mixing. The procedure is here represented in symbolic format, the signal being S(t), the non-quadrature multipliers being cos(wt + ~ + 5) and sin(wt + ~ Thus I & Q are: S(t)cos(wt + ~ + 5/2) and S(t)sin(wt + ~ - S/2).

Expansion of these expressions gives: S(t) [ cos(wt + ~)cos#/2-sin(wt + #)sin(#/2)] and S(t) [ cos(wt + #)sin(-#/2)+sin(wt + #)cos(#/2) ] .

Adding tand 2 of each expression to the other expression gives I' and Q' of S(t) [ cos(wt + ~)(cos#/2)+sin(-#/s)tan(#/2)] and S(t) [ sin(wt + ~)(cos(#/2)-sin#/2)tan(#/2)], which is S(t) [ cos # cos(wt + #)] cos(#/2) and S(t) [ cos g sin(wt + ~)].

cos(g/2) These expressions now represent correctly multiplied signals with slightly reduced amplitudes.

Again, a similar procedure can be used to apply a pre-correction to r.f. quadrature signals for mixing within a zero IF transmitter with uncorrected baseband signals.

The r.f. quadrature signals for mixing with the baseband signals are cos(wt + ~ +g/2) and sin(wt + ~ -#/2).

The uncorrected baseband signals are A and B.

To apply correction, we add tan (g/2) of each to the other, to give A + Btan(S/2) and B + Atan(#/2).

[NB. for the case of a phase modulation scheme A+B =1 and the peak amplitude change is the transmitter The mixed signals are added in the transmitter as [A + Btan(#/2)]cos(wt +# +#/2) + [ B + Atan(J/2) ] sin(wt + ~ -#/2) = Acos(wt + #)(cosJ/2-sinJ/2tan & ) + Asin(wt + #)(cos#/2tan#/2-sin#/2) + Bcos(wt + #)(cos#/2tan#/2-sin#/2) + Bsin(wt + #)(cos#/2-tan#/2sin#/2) = [Acos(wt + #) + Bsin(wt + ~)]cos# cos(#/2) i.e. identical to the signal with no phase error, but with an amplitude reduction of xcos#/cos(#/2).

In the arrangement shown in Fig. 2 a local oscillator input signal is fed via an input filter 20 to two phase generating filters 21I and 21Q which are conventional CR and RC networks respectively, producing I and Q components of the input having a nominal quadrature phase relationship. The I and Q components are each fed to a respective limiting amplifier 22I, 22Q and also to a respective differential amplifier 23I, 23Q, for example Gilbert cells. The differential amplifiers are controlled by a phase adjustment signal which is derived by a feedback arrangement (not shown) measuring the phase difference of the output r.f. I and Q signals. The outputs of the differential amplifiers 23I, 23Q are therefore correction signals which are in opposite phase to their respective inputs and whose amplitude is determined by the phase adjustment signal.

The I and Q correction signals are then added to the outputs of the Q and I limiting amplifiers respectively in summing networks 24I, 24Q to form the output r.f. 'I' and 'Q' signals.

In the integrated circuit implementation shown in Fig. 3 the local oscillator input is depicted in the form of LO and LO logical inputs. CR and RC networks form Iot Io, Qo and Qo quadrature components of the local oscillator inputs, which are then fed to main I and Q limiting amplifiers formed by transistor pairs T1-T4. The quadrature components are also fed to correction I and Q limiting differential amplifiers formed by transistors T5-T12 configured as Gilbert cells.Phase correction signals ~ and ~ #are input via transistors T13-T16 to control the derivation of the correction signals which are then added, with appropriate phase opposition, to the outputs of the main limiting amplifiers to form the r.f. outputs I1, Il, Q1 and Ql While the invention has been described in the context of direct correction of the r.f. mixer signal phases, it can also be implemented in other portions of a radio equipment. Thus it can be implemented on baseband signals in the transmit portion of a transceiver. In this case, for the pre-correction for transmit the vector amplitude is increased, not decreased. The invention can also be implemented on the baseband output signals in the receiver portion. For a baseband implementation analogue multipliers and summing networks may be used, but it may be more convenient to implement these functions digitally.

Claims

1. A method of correcting quadrature phase error between in-phase (I) and quadrature (Q) components of an a.c. signal including the steps of deriving from each component a correction signal and adding the correction signal to the other component, the amplitudes of the correction signals being such as to cause the two resultant vectors to be in phase quadrature.

2. A method according to claim 1 wherein the a.c.

signal is an r.f. local oscillator signal in a radio transmitter/receiver equipment.

3. A method according to claim 1 wherein the a.c.

signal is a baseband input signal in a zero IF radio transmitter.

4. A method according to claim 1 wherein the a.c.

signal is a baseband output signal in a zero IF radio receiver.

5. A method of correcting quadrature phase error between in-phase (I) and quadrature (Q) components of an a.c. signal substantially as hereinbefore described.

6. An arrangement for correction of quadrature phase error between in-phase (I) and quadrature (Q) components of an a.c. signal including means for deriving from each component a correction signal of opposite phase to the respective component and means for adding each correction signal to the component from which the other correction signal is derived, the amplitude of the correction signals being such as to cause the two resultant vectors to be in phase quadrature.

7. An arrangement according to claim 6 wherein said means for deriving comprises respective differential amplifiers to which the I and Q components are input, the arrangement including means for applying to the differential amplifiers a phase adjustment control signal to control the amplitudes of the amplifier outputs.

8. An arrangement according to claim 6 or 7 including respective limiting amplifiers to which the I and Q components are input, the I and Q outputs of the limiting amplifiers being added to the Q and I correction signals in said adding means.

9. An arrangement according to claim 6, 7 or 8 wherein the a.c. signal is a local oscillator r.f.

signal.

10. An arrangement according to claim 6, 7 or 8 wherein the a.c. signal is a baseband signal on a radio equipment.

11. An arrangement for correction of phase quadrature error between in-phase (I) and quadrature (Q) components of an a.c. signal substantially as described with reference to the accompanying drawings.