Control Platform for High Speed Algorithm Development

Our sponsor MKS instruments deals in developing plasma generators for use in creating silicon computer chips. These generators are controlled by a separate piece of circuitry, which use an output voltage and current in order to ensure safe operation. However, the client's research and development team have run into the problem that any updates to their control software, whether that be a minute bug fix or a major algorithm update, can take hours to rebuild. Our group was tasked with creating a new control platform for MKS that can allow for easy reconfiguration/algorithm replacement, without having to go through extremely long build times. Our group used the "Red Pitaya" FPGA development board as a basis, with the goal of using the board's high-speed ADC/DACs to quickly sample the output of the RF generator, send it into the FPGA fabric for processing, before finally outputting the resulting waveform on the DAC to feed back into the RF generator. This was done using VHDL to model the FPGA hardware, and used the PYNQ interface to control the Red Pitaya's software during operation. The PYNQ interface utilizes Python to grab data from the input sensor, trigger any necessary modules, and output the final results on the DAC. Our client will be able to make any changes based on hardware using the VHDL environment. The Python functions allow for the development of software filters/algorithms as well, without having to rebuild the hardware portion of the design. Additionally, a PCB adapter board was designed to act as an in-between between the Red Pitaya and the RF generator. This board performs the necessary signal filtering/amplification to the main control voltage, as by default, the voltage is too much for the Red Pitaya to handle. It also serves to feed other control signals from the RF generator into the Red Pitaya's GPIO ports, so additional processing can occur based on the status signals.