Automatic generation of CMSIS DSP code for filters, in ARM toolset

October 12, 2017 // By Graham Prophet
ASN Digital Filter Designer is a code generation tool, by Advanced Solutions Nederland, that enables design of IIR and FIR digital filters from a graphical input. Advanced Solutions together with Keil has notified a webinar to be broadcast on October 19th 2017, highlighting the integration of the filter design facility directly into the ARM CMSIS DSP environment.

In the filter design software, the user defines a filter characteristic graphically, with drag-and-drop manipulation of all key parameters such as corner frequencies, roll-off rates and passband/rejection amplitudes. The software calculates the required coefficient set for a DSP implementation, as well as showing, for example, an interactive pole/zero plot. It then exports the filter definition to Matlab, or C++, or – in the latest release – direct to the ARM context as a CMSIS DSP library function.


Version 4.0 of the ASN Filter Designer package has recently been released. In addition to the ARM integration it will also export to Xilinx Vivado’s FIR compiler for FPGA design. The user interface can now import a variety of CSV and other data files, and allows selection of sub-datasets and editing of entries before saving. An all-pass filter cascade feature allow the user to equalise an IIR's phase with all-pass biquad pairs and fine tune positions interactively. The package’s “FilterScript” feature gets 20 new functions (more digital filter designers); also, it now supports analogue-to-digital filter design (bilinear and matched z-transform), and symbolic Laplace transfer functions.


The webinar aims to show how to exploit the DSP capabilities of Arm Cortex-M based microcontrollers, using the ASN Filter Designer tool to generate CMSIS-DSP compliant code that can be directly imported into µVision. Enhanced features such as the Logic Analyzer display the waveforms of input and output signals.


The webinar will take place at 10:00 CET / 9:00am BST on 19 th October. Registration, via the Keil website, is here.