Version 4.0 (R13) Fixed-Point Blockset Software Release Notes

Version 4.0 (R13) Fixed-Point Blockset Software

This table summarizes what's new in Version 4.0 (R13):

New features and changes introduced in this version are

• Installation and Licensing

• Unified Simulink and Fixed-Point Blockset Blocks

• Global Data Type Override and Logging Modes

• Shift Arithmetic Block

Installation and Licensing

To support the sharing of models in a large organization, Version 4.0 of the Fixed-Point Blockset product is automatically installed whenever the Simulink software is installed. You can configure models to either take full advantage of all fixed-point features, or to run without a Fixed-Point Blockset software license. Therefore all Simulink software users in your organization can run and work on the same model, regardless of whether they have a Fixed-Point Blockset software license.

You must have a Fixed-Point Blockset software license to run a model if it is configured to log minimums, maximums, or overflows. You control logging with the system-level setting Logging mode. If you turn logging off at the top-level system in a model, then no data is logged for any block in any subsystem of the model, and a Fixed-Point Blockset software license is not required. You also need a Fixed-Point Blockset software license to run a model that uses any nonbuilt-in, fixed-point data types. However, you can use the system-level setting Data type override to force blocks to use doubles or singles instead of fixed-point data types. Therefore, by turning the Data type override parameter on and the Logging mode parameter off at the top level of a model, a Simulink software user without a Fixed-Point Blockset software license can run a model with fixed-point enabled blocks. See Global Data Type Override and Logging Modes for more information on these settings.

If you have a Fixed-Point Blockset software license, you can run bit-true simulations with your models that contain fixed-point enabled blocks. If a Fixed-Point Blockset software license is not available or desired, you can turn logging off and data type override on at the top level of your model and perform idealized floating point-based simulations.

If you have both a Fixed-Point Blockset software license and a Real-Time Workshop software license, you can generate bit-true integer code from your models with fixed-point enabled blocks. If you do not have a Fixed-Point Blockset software license but you do have a Real-Time Workshop software license, you can generate idealized floating-point code from your models with fixed-point enabled blocks.

Unified Simulink and Fixed-Point Blockset Blocks

Many core Simulink and Fixed-Point Blockset blocks with similar functions have been unified in this release. For example, the Sum block in the Simulink Math Operations library and the Sum block in the Fixed-Point Blockset Math library are now the same block. All the functionality from each original block has been maintained in unifying these blocks. Compatibility with fixed-point data types and/or specific fixed-point features are now available with all of these blocks, whether the blocks used are from the Simulink library or the Fixed-Point Blockset library. You do not need to make any changes to your earlier models as a result of this improvement. You can now use any of the unified blocks with either built-in data types or fixed-point data types, which eliminates the need to replace blocks in your models when you want to use different data types. This change does not require Simulink software users to have a Fixed-Point Blockset software license. Refer to Installation and Licensing for more information.

Fixed-Point Blockset blocks that have been unified no longer have an "F" on their block icon. However, not all Fixed-Point Blockset blocks that have counterparts in the Simulink library have been unified. You can still use the fixpt_convert function to replace nonunified Simulink blocks with their corresponding Fixed-Point Blockset blocks in your models.

Nonunified Fixed-Point Blockset blocks have an advantage over their Simulink library counterparts in that they can handle more data types. As discussed above, you can easily switch them between fixed-point data types and singles or doubles using the global data type override setting. However, you may still want to use the Simulink library counterparts of nonunified Fixed-Point Blockset blocks in some cases, because they support faster simulation times for the data types they handle.

The following table lists the unified blocks in this release, and the Simulink and Fixed-Point Blockset libraries in which they are found.

Compatibility Considerations

Breaking library links to Fixed-Point Blockset blocks will almost certainly produce an error when you attempt to run the model. If broken links exist, you will likely uncover them when upgrading to the latest release of the Fixed-Point Blockset software. The fixpt_restore_links command can be used to restore links for Fixed-Point Blockset blocks.

Global Data Type Override and Logging Modes

You can now set data type override and logging modes for systems or subsystems in the Fixed-Point Blockset Interface. The Override data type(s) with doubles and Log minimums and maximums check boxes have been removed from the mask of every Fixed-Point Blockset block.

Compatibility Considerations

The Override data type(s) with doubles and Log minimums and maximums check boxes have been removed from the mask of every Fixed-Point Blockset block. You can now set these parameters on the system or subsystem level.

When you upgrade to Version 4.0, all doubles override and logging information is cleared from your models. You can reset these controls in the Fixed-Point Blockset Interface for any system or subsystem. Access the Fixed-Point Blockset Interface from the Simulink Tools menu, or by typing fxptdlg('modelname') at the MATLAB command line.

If you have been getting or setting the block parameters DblOver or dolog in your M-code, you must now use the system parameters DataTypeOverride and MinMaxOverflowLogging.

Shift Arithmetic Block

The Fixed-Point Blockset software now includes the Shift Arithmetic block in the Bits library. The Shift Arithmetic block shifts the bits or binary point of a signal, or both.

Version 4.0.1 (R13+) Fixed-Point Blockset Software

Compatibility Summary for Simulink Fixed Point Software