Instrument integration without FPGA expertise

Research science organisations often purchase or rent dedicated lab equipment and multiple measurement devices for their test environments. As a result, their scientists and engineers will spend a lot of time setting up new experiments, collecting data and analysing results.

To help these organisations boost productivity and reduce costs, Liquid Instruments set out to introduce flexible software-defined instrumentation, Moku:Pro, for easily deploying multiple precision instruments as well as custom signal-processing functionality onto FPGAs.

Of course, programming FPGAs requires specialised knowledge that researchers don’t typically possess. To address this challenge, Liquid Instruments developed a workflow for programming Moku:Pro that combines HDL Coder and Simulink with Zynq UltraScale+ MPSoC and Moku Cloud Compile. With these tools, there is no need to download FPGA programming tools or to have FPGA expertise.

“HDL Coder enabled our team to build an end-to-end capability that gives engineers and scientists a way to customise their systems while maintaining tight control over hardware,” said Daniel Shaddock, CEO of Liquid Instruments. “Using Moku:Pro, customers can swap instruments in and out and change configurations on the fly, which can streamline testing in research and production environments.”

The Liquid Instruments team designed Moku:Pro so that research scientists and engineers can customise the device’s built-in measurement systems using MATLAB or Simulink. To enable this customisation, the team partitioned the onboard Zynq UltraScale+ MPSoC into five distinct regions. Four regions are dedicated to instruments and one to system control. Each region can be independently and dynamically reconfigured to implement its own function and bitstream file.

Liquid Instruments customers can choose from a wide variety of lab instruments to deploy on Moku:Pro, including oscilloscopes, waveform generators, frequency response analysers, PID controllers, data loggers, spectrum analysers, lock-in amplifiers, phase meters and digital filter boxes. Customers who need more advanced or custom processing then use MATLAB or Simulink to develop and model new IP blocks.

Customers use Fixed-Point Designer to convert the algorithms to fixed-point implementation and use HDL Coder to convert the design into low-level VHDL code, then paste the code into a browser and use Moku Cloud Compile to program the FPGA modules.

With subscription-based Moku Cloud Compile, customers design and deploy custom signal processing algorithms through Moku:Pro’s onboard FPGA. The cloud-based tool accepts VHDL code as its input and is fully compatible with the device’s Multi-instrument Mode, enabling custom-built test suites optimised for specific applications and requirements. Instrument space is reduced by up to four times, with Shaddock saying, “Users can dynamically reprogram Moku:Pro with up to four instrument modules from more than 10 precision instruments.”

Furthermore, FPGA programming capabilities can also be extended to non-specialists. “HDL Coder transforms MATLAB and Simulink files into VHDL and eliminates the need for users to compile within an integrated design environment (IDE) such as Xilinx Vivado,” Shaddock said.

Finally, processing time is reduced from hours to minutes. Shaddock explained, “In the new workflow, signal processing is programmed in MATLAB and modelled in Simulink. Using the output from HDL Coder, Moku Cloud Compile builds a bitstream and deploys it to the FPGA from a browser, automating resource configuration.”

Image caption: A researcher uses Moku:Pro’s built-in instruments customised with MATLAB.

Please follow us and share on Twitter and Facebook. You can also subscribe for FREE to our weekly newsletters and bimonthly magazine.