In-situ gain calibration of radio frequency devices using thermal noise

• US 20060264192A1
• Filed: 05/18/2005
• Published: 11/23/2006
• Est. Priority Date: 05/18/2005
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Keramat, Mansour, Kuo, Timothy C., Liu, Edward Wai Yeung
• Granted Patent: US 7,447,490 B2
• Time in Patent Office: Days
• US Class Current: 455/232.100
• CPC Class Codes: H04B 17/0085 using test signal generators
• CPC Class Codes: H04B 17/21 for calibration; for correcting measurements
• CPC Class Codes: H04W 52/16 Deriving transmission power values from another channel
• CPC Class Codes: H04W 52/367 Power values between minimum and maximum limits, e.g. dynamic range
• CPC Class Codes: H04W 52/40 during macro-diversity or soft handoff

Abstract

An apparatus for calibrating gain of an radio frequency receiver (“Rx”) is disclosed to provide, among other things, a structure for performing in-situ gain calibration of an RF integrated circuit over time and/or over temperature without removing the RF integrated circuit from its operational configuration, especially when the gain of the RF integrated circuit is susceptible to variations in process, such as inherent with the CMOS process. In one embodiment, an exemplary apparatus includes a thermal noise generator configured to generate thermal noise as a calibrating signal into an input of an Rx path of an RF integrated circuit. The apparatus also includes a calibrator configured to first measure an output signal from an output of the Rx path, and then adjust a gain of the Rx path based on the thermal noise. In one embodiment, the thermal noise generator further includes a termination resistance and/or impedance.

Claims

1. An apparatus for calibrating gain of a radio receiver (“

Rx”

) path in a radio frequency (“

RF”

) integrated circuit comprising;

a thermal noise generator configured to generate thermal noise as a calibrating signal into an input of said Rx path; and

a calibrator configured to;

measure an output signal from an output of said Rx path, and

adjust a gain of said Rx path based on said thermal noise.

2. The apparatus of claim 1 wherein said thermal noise generator further comprises a termination resistance or impedance, wherein said thermal noise generated therefrom is constant over frequencies useable by said RF integrated circuit.

3. The apparatus of claim 2 wherein said calibrator adjusts said gain by an amount derived from a noise figure (“

NF”

) of said Rx path, said noise figure being less susceptible to variations over process and temperature than is said gain.

4. The apparatus of claim 1 further comprising a temperature sensor coupled to said calibrator for adjusting a value of said noise figure in accordance with a measured ambient temperature to improve calibration accuracy.

5. The apparatus of claim 1 wherein said calibrator is an in-situ calibrator configured to automatically calibrate said gain while said RF integrated circuit remains configured to receive RF communication signals.

6. The apparatus of claim 1 wherein said calibrator further comprises:

a gain evaluator configured to determine a gain differential between a calculated gain that is based on said output signal and an expected gain; and

a gain adjuster configured to adjust said gain of said Rx path by an amount equivalent to said gain differential, said Rx path including an adjustable amplifier.

7. The apparatus of claim 6 wherein said gain evaluator further comprises:

an output quantifier configured to quantify an output noise power signal as said output signal to determine a measured output noise power signal value; and

a parametric generator configured to provide a plurality of parameters associated with said Rx path, at least one parameter being an input noise power signal value representative of said thermal noise,

wherein said gain evaluator determines said calculated gain as a function of said measured output noise power signal value and said input noise power signal value.

8. The apparatus of claim 7 wherein said plurality of parameters associated with said Rx path further comprises one or more of the following:

a noise figure value representing an equivalent noise figure for said Rx path, and

an equivalent noise bandwidth value representing an equivalent noise bandwidth of said Rx path,

wherein said gain evaluator determines said calculated gain as a difference between said measured output noise power signal value and a collective value representative of contributions by said input noise power signal value, said noise figure value, and said equivalent noise bandwidth value.

9. The apparatus of claim 7 further comprising an analog-to-digital converter for digitizing said output noise power signal to form said measured output noise power signal value.

10. The apparatus of claim 9 wherein said plurality of parameters associated with said Rx path further comprises one or more of the following:

an upper limit of operation of said analog-to-digital converter associated with a first error value representing a clipping error,

a lower limit of operation of said analog-to-digital converter associated with a second error value representing a quantization error, and

a sample size value associated with a sample size for maintaining an estimation error during measurement of said measured output noise power signal value,

wherein said gain evaluator determines said gain differential so that during calibration said output noise power signal does not surpass either said upper limit of operation or lower upper limit of operation,

wherein said gain evaluator quantifies said output noise power signal using said sample size value.

11. A method for calibrating a radio frequency (“

RF”

) integrated circuit comprising;

modeling an effective input noise power signal value from at least one noise figure (“

NF”

) value associated with an RF radio receiver as an input thereto;

measuring an output noise power signal from said RF radio receiver to form a measured output noise power signal value; and

calibrating a gain of said RF radio receiver as a function of said effective input noise power signal value and said measured output noise power signal value.

12. The method of claim 11 wherein measuring said output noise power signal comprises:

generating thermal noise; and

applying said thermal noise as a calibrating signal into said RF radio receiver,

wherein generating thermal noise and applying said thermal noise are repeatable for performing in-situ calibration of said gain.

13. The method of claim 11 further comprising:

measuring ambient temperature to determine a measured temperature for improving calibration accuracy;

modifying a value of said at least one noise figure to form a modified noise figure value based on said measured temperature; and

adjusting said gain in accordance with said modified noise figure value.

14. The method of claim 11 wherein modeling said effective input noise power signal comprises:

determining a first noise figure value for an estimated thermal noise value;

determining a second noise figure value for an Rx path of said RF radio receiver;

determining an equivalent noise bandwidth for said Rx path; and

deriving said effective input noise power signal value from said first noise figure value, said second noise figure value and said equivalent noise bandwidth.

15. The method of claim 11 wherein calibrating said gain comprises:

calculating a calculated gain based on said measured output noise power signal value and said effective input noise power signal value;

determining a gain differential between said calculated gain and a desired gain; and

adjusting said gain of said RF radio receiver by an amount equivalent to said gain differential.

16. The method of claim 11 wherein measuring said output noise power signal comprises:

determining an upper limit of operation associated with a clipping error;

determining a lower limit of operation associated with a quantization error;

determining a sample size value representing a sample size for maintaining an estimation error during measurement of said measured output noise power signal value;

adjusting said gain so that during calibration said output noise power signal does not surpass either said upper limit of operation or lower upper limit of operation; and

using said sample size value during measurement of said measured output noise power signal value.

17. A bi-modal radio receiver (“

Rx”

) integrated circuit formed on a substrate and configured to operate in a receiver mode to receive radio communication signals and to operate in a calibration mode to perform in-situ gain calibration, the bi-modal radio Rx integrated circuit comprising;

an antenna port configured to receive radio communication signals from an antenna;

an in-situ calibration signal generator configured to internally generate a calibration signal;

a mode selector configured to select a signal source from either said antenna port or said in-situ calibration signal generator; and

an amplification path including an amplifier having a gain that is adjustable, said amplifier path coupled to said mode selector for amplifying radio communication signals in said receiver mode and for amplifying said calibration signal in said calibration mode,

wherein said amplifier is configured to receive an adjustment signal representative of an amount to adjust said gain.

18. The bi-modal radio Rx integrated circuit of claim 17 wherein said calibration signal is thermally generated and said amplifier is a radio frequency (“

RF”

) amplifier.

19. The bi-modal radio Rx integrated circuit of claim 17 wherein said adjustment signal is based on a noise figure value of at least a portion of said amplification path.

20. The bi-modal radio Rx integrated circuit of claim 17 wherein said Rx amplification further comprises an analog-to-digital converter, whereby said adjustment signal is configured to modify said gain to minimize measurement errors due to said analog-to-digital converter including clipping errors, quantization noise floor errors, and estimation errors due to sample size.

21. A system for calibrating gain of a radio receiver (“

Rx”

) path comprising;

a radio receiver subsystem having;

a thermal noise source configured to provide a calibration signal,

an Rx path, and

one or more filters, each being modeled to respectively have a thermal noise source value, an Rx path noise figure value, and an equivalent noise bandwidth for said one or more filters, and

a gain adjuster to adjust a gain of said radio receiver subsystem; and

a processor configured to communicate with a central processing unit (“

CPU”

) of a networked computing device, said processor comprising a gain calibrator including;

parameter generator to provide parameter values for deriving an effective input noise power signal value, said parameter values including said thermal noise source value, said Rx path noise figure value, and/or said equivalent noise bandwidth,

an output quantifier to measure an output noise power signal from said Rx path to form a measured output noise power signal value,

a gain evaluator configured to determine a gain differential between a calculated gain and an expected gain, said calculated gain being a function of said measured output noise power signal value and said effective input noise power signal value,

wherein said gain differential is passed to said gain adjuster for calibrating said gain.

22. The system of claim 21 further comprising a temperature sensor coupled to said gain calibrator for modifying either said thermal noise source value or said Rx path noise figure value, or both, based on a measured ambient temperature to improve calibration accuracy.

23. The system of claim 21 wherein said radio receiver is a constituent element of an RF integrated circuit manufactured using a complementary metal oxide semiconductor (“

CMOS”

) processing technology, and said processor is a base band processor including a medium access controller (“

MAC”

) module for operating said system in a wireless local area network (“

WLAN”

).

24. The system of claim 21 wherein said Rx path further comprises an analog-to-digital converter, whereby said gain calibrator is configured to modify said gain differential to minimize measurement errors due to said analog-to-digital converter including clipping errors, quantization noise floor errors, and estimation errors due to sample size.