How VAT was able to acquire high-frequency real-time data in the development of an embedded system without DAQ-Hardware
Context
When working with bare-metal embedded systems, understanding what’s happening inside your systems isn’t a luxury. It’s essential. This was precisely the situation faced by a corporate R&D team at VAT Group, a leading supplier to the semiconductor industry. The team is responsible for developing and validating embedded systems ranging from sensor interfaces to motor control, involving microcontrollers as well as high-speed FPGA-based signal processing. Their job is to ensure that everything works together seamlessly, from analog sensor data to high-speed digital control logic. But validating that requires one critical thing: visibility.
They need insight into the system’s internal behavior. They need a way to observe, analyze and adjust system behavior. Internal variables. Real-time states. Data explains not just what happened, but why and how. Quickly, non-invasively and with enough detail to make informed decisions. Preferably, while the system is running.
That’s why the team joined the early access program for es:scope® – a real-time software oscilloscope purpose-built to provide direct access to runtime variables in embedded systems. Their goal was clear: to understand system behavior faster, and debug without guesswork.
Challenge
The development environment comprised a complex combination of microcontroller-based control logic and FPGA-driven, high-speed signal processing, integrated into a real-time system for precision actuation and sensor interaction. However, visibility into the system’s internal state was limited. Many critical variables, such as sensor timings, control loop states, and transient anomalies, could only be observed indirectly or through post-processing. This created three key challenges for the team:
1) Lack of real-time insight into internal variables
Conventional tools like JTAG and serial logging were insufficient for observing high-frequency variables, particularly in interrupt-driven and time-critical subsystems. While oscilloscopes provided access to physical signals, they offered no insight into internal software states. For some cases, such as capturing a single high-speed variable, the team needed streaming rates of 500 kSa/s. In others, they monitored 10 variables at ~20 kSa/s each – a rate that standard debugging tools couldn’t sustain.
2) Challenge to synchronize sensor and control data
The system included sensors and FPGA logic that operated on independent clocks. Aligning their output and correlating them with control decisions was difficult, particularly due to the lack of a shared timebase. However, timing remained critical even within single subsystems, such as the microcontroller’s motor control loop. For instance, when analyzing behavior within a single pulse width modulation (PWM) cycle, the team required high-resolution insight into variables such as current feedback, voltage ripple, and controller-internal states. These dynamics directly impact actuator performance and stability, yet they are often too fast to capture with standard tools. Traditional setups either lacked temporal resolution or required intrusive instrumentation. The missing piece was the ability to observe internal variables with sub-cycle resolution that were tightly correlated with real-world execution and available live, not just after logging.
3) Limitations of traditional setup
DAQ systems were considered because they are highly capable of measuring physical signals, such as current and voltage. However, in this case, the variables of interest were internal and included digital control states, feedback calculations, and state machine transitions. Replicating these signals at an external test point would have required additional firmware logic and pin routing. This would have added both complexity and potential timing interference. Additionally, each test cycle would have required extensive post-processing. For a fast-paced development cycle focused on embedded behavior, this approach was impractical. Tracing tools were also evaluated, but many lacked practical real-time, multi-variable visualization on bare-metal systems. Others introduced significant execution overhead or required proprietary toolchains. Importantly, neither data acquisition systems nor tracers could adjust system parameters, such as control gains, in real time while monitoring the results.
Solution
To overcome these limitations, the development team joined the early access program for a suite of tools developed by es:saar, including:
- es:scope®: A real-time software oscilloscope for embedded systems
- es:prot: An open middleware for efficient signal transport from embedded targets
- es:com: An offloading adapter for high-bandwidth data extraction from FPGAs
First, the team integrated es:prot into their firmware*, which allowed selected internal variables from the bare-metal microcontroller system to be exposed as structured signals. These were then transmitted over standard Ethernet (UDP) with minimal system overhead. This lightweight protocol enabled high-frequency, real-time data streams without interfering with time-critical behavior. For analysis, es:scope® provided an oscilloscope-like interface on a standard laptop. Developers could visualize signals in real time, apply FFT or XY analyses, set triggers and export data in formats such as CSV or MATLAB for further processing. At a later stage, the team extended their setup using the es:com offloading adapter. This allowed internal FPGA signals to be output directly and in parallel to the adapter, which then streamed the data over Ethernet to es:scope®. The result was high-speed visibility into internal FPGA states without occupying processing resources or requiring modifications to the core logic. Together, these tools addressed a key bottleneck in the development of embedded systems: the efficient and non-invasive acquisition of internal runtime data, without the complexity of DAQ hardware or intrusive instrumentation.
* Technical details about the specific setup will not be disclosed due to customer discretion. However, this is not a limiting factor: the open, C-based es:prot protocol can be integrated into any microcontroller or processor architecture – with or without an RTOS. An ARM-based architecture was used in this project.
Results
- Real-time visibility into system behavior
Engineers could monitor internal variables both from microcontrollers and FPGAs in real time, allowing earlier identification of timing issues, unexpected interactions, and edge cases. - Reduced reliance on post-processing
Many verification tasks that previously required multiple test iterations and offline data review were now performed interactively, using conditional recording and real-time visualization. - Improved signal correlation
With support for multi-signal, high-frequency acquisition over standard interfaces, the team was able to correlate internal software states with sensor inputs and actuator responses without the need for synchronized external instrumentation. - Run-Time Tunability
Control parameters could be adjusted during live operation, allowing for rapid calibration and behavioral optimization without firmware rebuilds or system restarts.
- Toolchain adaptability
The open-source nature of the es:prot middleware allowed seamless integration into the project’s bare-metal firmware environment. No vendor lock-in or RTOS dependencies were introduced.
These improvements translated into fewer validation iterations, lower measurement overhead, and faster decision-making during system integration and early testing phases.
Conclusion
For embedded systems teams working at the intersection of sensors, control logic, and high-speed signal processing, internal system visibility is essential- but often hard to achieve without intrusive tools or costly infrastructure. By adopting es:scope®, es:prot, and the es:com offloading adapter early in their development process, the team was able to gain real-time insight into their systems without disrupting existing hardware or firmware designs. This supported their verification and calibration process.
Deutsch 
English