Introducing Quanta Control: FPGA Feedback at 125 MHz

The Feedback Bottleneck

In precision measurement and quantum technology, the feedback controller is often the critical bottleneck. Whether you're locking a laser to a high-finesse cavity, stabilizing a Kerr microresonator, or reading out a precision sensor, the achievable stability is limited by two primary factors: bandwidth and latency.

Traditional solutions have long been binary:

1. •Analog Controllers: Fixed-function, ultra-fast, but notoriously difficult to reconfigure or integrate with modern data acquisition systems.
2. •Digital Controllers (CPU/DSP): Highly flexible and easy to program, but plagued by non-deterministic latency from operating system jitter, interrupt handling, and slow bus transactions.

Quanta Control was born to break this binary. We believe that researchers should have access to the deterministic performance of analog hardware with the software flexibility of a modern digital instrument.

Our Approach: FPGA-First Architecture

We have built our architecture on the Xilinx Zynq SoC, utilizing the Red Pitaya platform as our primary prototyping target. Our system is designed from the ground up to ensure that the feedback loop never touches a general-purpose CPU.

On the FPGA fabric (Programmable Logic), we implement the entire time-critical signal chain at a 125 MHz clock rate:

• •Multi-tone DDS Signal Generation: Three independent sine wave generators for multi-tone lock-in detection or real-time system identification.
• •6-stage CIC Decimation: Sophisticated filtering that reduces the 125 MSPS ADC stream to a manageable 1 MSPS for the controller, providing significant noise reduction without the latency of complex FIR filters.
• •Fixed-point PI/PID Control: Using Q16.16 arithmetic and hardware-level anti-windup clamping to ensure unconditional stability and predictable performance.
• •Internal Signal Routing (MUX): The ability to route any internal node—pre-filter, post-filter, or error signal—to the output for real-time monitoring on a scope.

Software as an Enabler, Not a Bottleneck

While the logic lives in the FPGA, the control lives in your browser. We leverage the dual-core ARM processor on the Zynq SoC to host:

• •A high-performance Rust-based REST API that manages the memory-mapped register interface.
• •WebSocket streaming for real-time oscilloscope data.
• •A Vue.js + TypeScript web application that provides an intuitive dashboard for parameter tuning, Allan deviation stability analysis, and transfer function visualization.

Because the software layer is decoupled from the signal processing, you can update gains or change setpoints in the middle of a measurement without ever dropping your lock.

Why Open Matters

Precision research requires transparency. A "black box" PID controller with a proprietary binary driver is a liability in a laboratory that needs to publish reproducible results.

Quanta Control is designed to be auditable by design. By using open standards (AXI4), standard HDL (SystemVerilog), and modern web technologies, we ensure that you can verify exactly how your error signal is being processed.

Whether you are in the early stages of a physics experiment or prototyping the control logic for a future commercial optical frequency comb product, our platform provides a robust foundation that scales with your needs.

Looking Ahead

This announcement is just the beginning. In the coming weeks, we will be releasing a series of technical deep-dives into our architecture, including:

1. •Digital Lock-In Amplification on FPGA: Implementing quadrature demodulation for PDH locking.
2. •Fixed-Point PID implementation: The math and the pitfalls of FPGA-based controllers.
3. •Using Allan Deviation: How to interpret stability data directly from your controller.

We invite you to explore our Technology page to see the full architecture in detail, or get in touch if you have a specific feedback control problem that needs a high-performance solution.