Non-linear filter for DML
AssigneeSemtech Canada Corporation
InventorGurpreet S. Bhullar
Life of this patent
Estimating the term…
Abstract
A circuit is disclosed having a component having repeatable distortion characteristics; and a drive circuit for providing a drive signal and comprising a non-linear filter for pre-compensating for distortion introduced by the component having repeatable distortion characteristics in response to the drive signal, the distortion having a non-linear response to the drive signal.
Claims16
A semiconductor analogue circuit for pre-compensating a distortion being introduced by a directly modulated laser (DML), wherein the distortion reduces a signal-to-noise ratio (SNR) and degrades performance of the DML and a data link, wherein the distortion is amplitude dependent and time dependent, wherein the amplitude dependent distortion produces non-linearity, wherein the distortion has different distortion characteristics for a transition to a ones level and to zeros level, wherein the transition is the transition of the DML's response signal, wherein the semiconductor analogue circuit is configured to provide a pre-compensated drive signal to the DML to pre-compensate the distortion being introduced by the DML, wherein the semiconductor analogue circuit comprises:
- an input circuit;
- a non-linear filter coupled to the input circuit;
- and
- a summer coupled to the input circuit and the non-linear filter,
- wherein the input circuit comprises a drive input node and a drive output node,
- wherein the input circuit is configured to receive a drive input signal at the drive input node and configured to provide a drive signal at the drive output node,
- wherein the non-linear filter comprises a plurality of taps arranged in parallel,
- wherein each tap of the plurality of taps comprises a tap output node,
- wherein each tap of the plurality of taps comprises a multiplexer and a multiplier,
- wherein the multiplexer together with the multiplier are configured to generate a non-linear tap output signal that is both dependent on the drive signal's instantaneous amplitude and dependent on time, the non-linear tap output signal being produced at the tap output node,
- wherein the multiplexer is coupled to the drive output node,
- wherein the multiplexer comprises a first multiplexer input node, a second multiplexer input node, a selector node, and a multiplexer output node,
- wherein the multiplexer is configured to receive a first weight signal at the first multiplexer input node, configured to receive a second weight signal at the second multiplexer input node, and configured to receive the drive signal at the selector node,
- wherein the multiplexer is configured to select the first weight signal or the second weight signal based on the drive signal's instantaneous amplitude, and configured to produce, at the multiplexer output node, the selected one of the first weight signal or the second weight signal, wherein:
- the instantaneous amplitude of the drive signal itself is used during an operation of the multiplexer to select the first weight signal or the second weight signal for the multiplexer;
- when the drive signal's instantaneous amplitude is level one, the multiplexer is configured to select the first weight signal, and the first weight signal shapes the non-linear tap output signal at the tap output node;
- when the drive signal's instantaneous amplitude is level zero, the multiplexer is configured to select the second weight signal, and the second weight signal shapes the non-linear tap output signal at the tap output node;
- the level one is different from the level zero;
- and
- the first weight signal is different from the second weight signal, such that the non-linear tap output signal at the tap output node has a first characteristic for the level one instantaneous amplitude of the drive signal, which is different from a second characteristic for the level zero instantaneous amplitude of the drive signal,
- wherein the multiplier is coupled to the drive output node and the multiplexer output node,
- wherein the multiplier comprises a first multiplier node coupled to the drive output node, a second multiplier node coupled to the multiplexer output node, and a multiplier output node,
- wherein the multiplier is configured to receive a tap signal at the first multiplier node, configured to receive the selected one of the first weight signal or the second weight signal at the second multiplier node, configured to scale the tap signal based on the selected one of the first weight signal or the second weight signal, and configured to produce the scaled tap signal at the multiplier output node,
- wherein the tap signal is scaled differently depending on the instantaneous amplitude of the drive signal, generating a non-linear drive signal contribution at the multiplier output node,
- wherein the non-linear tap output signal is the scaled tap signal,
- wherein the tap output node is the multiplier output node,
- wherein the tap signal for a first tap of the plurality of taps comprises the drive signal,
- wherein each of the tap signals for other taps of the plurality of taps comprises a respective delayed version of the drive signal such that the non-linear tap output signal depends on a previous state of the drive signal, and the non-linear tap output signal's non-linear characteristics are time dependent,
- wherein the summer is coupled to the drive output node and the tap output nodes of the plurality of taps, and
- wherein the summer is configured to sum the drive signal from the drive output node and the non-linear tap output signals from the tap output nodes of the plurality of taps, configured to produce the pre-compensated drive signal which is dependent on the instantaneous amplitude of the drive signal and dependent on the previous state of the drive signal, and configured to drive the DML using the pre-compensated drive signal.
A circuit for pre-compensating a distortion being introduced by a directly modulated laser (DML) and for providing a pre-compensated drive signal to the DML to pre-compensate the distortion being introduced by the DML, the circuit comprising:
- an input circuit;
- a non-linear filter coupled to the input circuit;
- and
- a summer coupled to the input circuit and the non-linear filter,
- wherein the input circuit comprises a drive input node and a drive output node,
- wherein the input circuit is configured to receive a drive input signal at the drive input node and configured to provide a drive signal at the drive output node,
- wherein the non-linear filter comprises a plurality of taps arranged in parallel,
- wherein each tap of the plurality of taps comprises a tap output node,
- wherein each tap of the plurality of taps comprises a multiplexer and a multiplier,
- wherein the multiplexer together with the multiplier are configured to generate a non-linear tap output signal at the tap output node,
- wherein the multiplexer is coupled to the drive output node,
- wherein the multiplexer comprises a first multiplexer input node, a second multiplexer input node, a selector node, and a multiplexer output node,
- wherein the multiplexer is configured to receive a first weight signal at the first multiplexer input node, configured to receive a second weight signal at the second multiplexer input node, and configured to receive the drive signal at the selector node,
- wherein the multiplexer is configured to select the first weight signal or the second weight signal based on the drive signal's instantaneous amplitude and configured to produce the selected one of the first weight signal or the second weight signal at the multiplexer output node, wherein:
- the instantaneous amplitude of the drive signal itself is used during an operation of the multiplexer to select the first weight signal or the second weight signal for the multiplexer;
- when the drive signal's instantaneous amplitude is a first level, the multiplexer is configured to select the first weight signal;
- when the drive signal's instantaneous amplitude is a second level, the multiplexer is configured to select the second weight signal;
- the first level is different from the second level;
- and
- the first weight signal is different from the second weight signal, such that the non-linear tap output signal at the tap output node has a first characteristic for the first level instantaneous amplitude of the drive signal, which is different from a second characteristic for the second level instantaneous amplitude of the drive signal,
- wherein the multiplier is coupled to the drive output node and the multiplexer output node,
- wherein the multiplier comprises a first multiplier node coupled to the drive output node, a second multiplier node coupled to the multiplexer output node, and a multiplier output node,
- wherein the multiplier is configured to receive a tap signal at the first multiplier node, configured to receive the selected one of the first weight signal or the second weight signal at the second multiplier node, configured to scale the tap signal based on the selected one of the first weight signal or the second weight signal, and configured to produce the scaled tap signal at the multiplier output node,
- wherein the tap signal is scaled differently depending on the instantaneous amplitude of the drive signal,
- wherein the non-linear tap output signal is the scaled tap signal,
- wherein the tap output node is the multiplier output node,
- wherein the tap signal for a first tap of the plurality of taps comprises the drive signal,
- wherein each of the tap signals for other taps of the plurality of taps comprises a respective delayed version of the drive signal,
- wherein the summer is coupled to the drive output node and the tap output nodes of the plurality of taps, and
- wherein the summer is configured to sum the drive signal from the drive output node and the non-linear tap output signals from the tap output nodes of the plurality of taps, configured to produce the pre-compensated drive signal, and configured to drive the DML using the pre-compensated drive signal.
Dependencies are inferred from the claim text.