Project Logs

Collapse

• Designing for real industrial environments

  Alireza Noferesti • 05/09/2026 at 13:12 • 0 comments

  One of the main goals of this project was not just to build another DAQ —  
  but to design a system that behaves reliably in real industrial environments.

  In many practical setups, problems are not caused by software or functionality.

  They come from:

  • Electrical noise  
  • Long cable runs  
  • Ground loops  
  • Shared power domains  
  • External drivers and switching loads  

  These issues become especially noticeable when analog signals and motion systems are combined in the same setup.

  Instead of treating noise as a secondary problem, this design was built around minimizing it from the beginning.

  Some of the key architectural decisions include:

  • Galvanically isolated analog inputs  
  • Separate analog and digital power domains  
  • Isolated communication interfaces  
  • Differential analog inputs for improved noise immunity  
  • Higher-current analog outputs capable of directly driving external circuits  

  The analog input section is fully isolated from the MCU and digital power domain.

  Even the analog ground is not shared with the microcontroller ground or the external communication ground.

  Isolation is achieved using isolated DC-DC converters and isolation stages rated up to approximately 1.5 kV for the analog section, and several kilovolts for communication isolation.

  This significantly reduces the impact of:

  • Ground potential differences  
  • Industrial switching noise  
  • Noise coupling through shared return paths  

  Another important design choice was output drive capability.

  Many DAQ systems provide analog outputs with only a few milliamps of output current, requiring external amplifiers or buffer stages for practical use.

  In this design, each analog output can provide up to approximately 250 mA directly.

  This allows many external devices and control interfaces to be driven without additional amplification hardware.

  The differential analog inputs also make long-distance analog measurements much more practical.

  For example:

  • Remote analog feedback signals  
  • Differential sensor measurements  
  • RTD temperature sensing over long cables  

  Especially with multi-wire RTD configurations, differential measurement greatly improves noise immunity and measurement stability.

  The communication architecture was also designed with industrial environments in mind.

  In addition to USB, the system includes RS485 communication, allowing the DAQ to operate much farther away from the host PC when required.

  This becomes very useful in electrically noisy environments where USB cable distance is limited.

  Another interesting capability enabled by the integrated analog I/O architecture is internal closed-loop analog control.

  For example, an analog output can be connected directly to a differential analog input and controlled through the internal PID system.

  This allows the system to regulate analog setpoints internally without requiring an external PLC or controller for certain applications.

  The goal of these design choices was not simply adding features —  
  but reducing the amount of external hardware normally required in industrial test and control setups.

• Built-in PID: smart integration or bad idea

  Alireza Noferesti • 04/25/2026 at 10:35 • 0 comments

  One of the design decisions in this project was to include PID control directly inside the DAQ.

  At first glance, it seems like a natural extension:

  If the system is already measuring signals, why not close the loop internally?

  This approach offers some clear advantages:

  • Reduced latency (no external communication delays)  
  • Simpler system architecture  
  • Fewer external devices and wiring  

  But it also raises some concerns.

  In more complex systems, control is often handled by a dedicated PLC or controller — not the DAQ itself.

  Reasons for that include:

  • Greater flexibility in control logic  
  • Easier debugging and tuning  
  • Clear separation between measurement and control layers  

  By integrating PID into the DAQ, the system becomes more compact —  
  but potentially less modular.

  So the question becomes:

  👉 Is integrating control into the DAQ a practical simplification?  
  👉 Or does it reduce flexibility in ways that matter in real-world systems?

  In this project, the goal is not to replace full PLC-based control systems,  
  but to explore a middle ground — especially for test setups and compact automation systems.

  I’d really like to hear your thoughts:

  Have you used DAQs with built-in control before?  
  Did it simplify your system — or make it harder to manage?

• Noise is always there (and it never behaves nicely)

  Alireza Noferesti • 04/25/2026 at 10:34 • 0 comments

  One of the earliest challenges in this project wasn’t functionality — it was noise.

  In real-world environments, especially when working with motors, switching power supplies, and long cables, electrical noise is not an exception — it’s the default condition.

  Even simple measurements can become unpredictable:

  • Analog signals fluctuate  
  • Ground references shift  
  • Switching noise couples into sensitive paths  

  At first, the natural instinct is to “filter it out.”

  But in practice, that’s rarely enough.

  This led to a key design decision in this project:

  👉 Separate analog and digital power domains

  The goal was to reduce noise coupling between high-speed digital switching and sensitive analog measurements.

  However, this choice came with its own challenges:

  • Grounding strategy becomes critical  
  • PCB layout requires strict partitioning  
  • Power sequencing and return paths matter much more  

  In other words — it helps, but it’s not a free win.

  In some cases, it simplifies the problem.  
  In others, it introduces new complexity.

  This raises an interesting question:

  👉 Is power domain separation always worth it?  
  👉 Or are there cases where a well-designed single-domain system performs just as well?

  I’d be really interested in hearing your experience:

  Have you seen significant improvements with separate analog/digital domains — or diminishing returns?

• Why combine DAQ, control, and motion?

  Alireza Noferesti • 04/25/2026 at 10:32 • 0 comments

  In many real-world setups, building a complete test or control system usually means combining multiple separate devices:
  

  • A DAQ for measurement  
  • A controller (or PLC) for logic and PID  
  • A motion system interface for actuators  

  This approach works — and it’s widely used.

  But in practice, it often leads to:

  • More wiring  
  • More configuration  
  • More failure points  
  • And in noisy environments, more unpredictable behavior  

  After working with these kinds of setups, a simple question kept coming up:

  How much of this complexity is actually necessary?

  Is it always better to keep everything modular?  
  Or are there cases where integrating these functions into a single system actually improves reliability and simplicity?

  This project started as an attempt to explore that question — not to replace modular systems, but to understand where integration makes sense.

  I’d be really interested to hear your perspective:

  👉 Do you prefer modular systems, or more integrated designs in real-world projects?  
  👉 Where have you seen integration help — or cause problems?

View all 4 project logs