Environment chamber monitoring and control

An old environmental monitoring and control chamber system has been replaced with a FieldPoint system that achieves reliable controlled environmental conditions. National Instruments LabVIEW Real-Time and FieldPoint was used to control the system in order to achieve highly accurate steady-state temperature and humidity setpoints.

Reliable controlled environmental conditions

The National Standards Commission (NSC) performs pattern approval type testing on measuring instruments used for trade purposes in Australia. Part of this process involves the use of temperature and humidity controlled environmental chambers. This project relates to a chamber which in extended form has a volume of 34 m3. Achieving reliable controlled environmental conditions has proven difficult in the past, as the previous system would require approximately 24 hours to stabilise to 20°C at 50% relative humidity. The previous system was controlled using a hardwired micro controller, which relied on outdated electronic control systems.

The NSC requested a general improvement to the environmental control parameters, with database control and comprehensive data logging features, which proved impossible with currently available low-cost off-the-shelf solutions. The NVSI environmental control solution was written in NI LabVIEW Real-Time, running in an NI FieldPoint module FP-2010. We achieved control of the system using several LabVIEW Real-Time PID controls to achieve steady-state temperature and humidity setpoints.

Other hardware we used to achieve the solution includes:

• NI FP-RTD-124: 4 wire RTD input module, 8 channel, 16 bit;
• NI FP-RLY-420: 8 channel SPST relay output module;
• NI FP-PG-522: 8 channel pulse generator module;
• NI FP-DO-401: discrete output module;
• NI FP-DI-330: universal discrete input module;
• NI FP-AI-110: analog input module, 8 channel, 16 bit;
• NVSI EnviroMON Ethernet Base, with temperature and pressure sensors.

The solution is database-driven, using an existing NSC SQL database on site - which interfaces to the code using the NI LabVIEW database connectivity toolkit. We created all of the required tables and stored procedures of this project in a generic format to facilitate a simple transition to other projects with NSC. We used a GUI developed with Microsoft Access to create a profile for undertaking tests. These tests typically consist of several legs, with each leg containing one or more physical parameters, such as temperature and relative humidity, along with the required test conditions of wait time, reach time, and stability.

We also defined the parameters we wanted to record using the GUI, and the system dynamically created a database table to store acquired data once we selected a test.

Data pertaining to measurement system components (called tags) are also stored in the database. For example, RTD fit curve types and data and raw and reference calibration data are stored in the database and are called up and read in at start-up.

Furthermore, a separate application used to manage tag calibration is used to write new calibration values to the database after calibration. This system provides easy access to historical calibration data which is used for evaluation of component performance and associated measurement uncertainty.

We attained temperature setpoints to an accuracy of ±0.1 K, an order of magnitude of improvement over the previous system. We included some filtering (modified Savitzky-Golai) to limit the noise produced by the rapid movement of air within the chamber. We used three PID loops to set the temperature - one for overall control of the temperature and two further PID loops required for biasing. These biasing loops ensure that temperature differential existed throughout the physical control system. Biasing of at least 0.5 K existed between the heater and the cooler, and we set up further biasing of at least 2 K between the return temperature and the heater. This biasing ensures a more constant flow of air over the chamber - removing pockets of warm or cool air, while keeping the system loaded to reduce lag time.

We met the relative humidity (RH) setpoint with an accuracy of ±1%, again an order of magnitude improvement. We achieved humidity control using two PID loops - one driving the production of steam, the other controlling the amount of scavenging of water from the chamber using refrigeration. After initial investigation, we found that using the RH as a process variable within the PID loops was far too sensitive to changes in the temperature. A change of 0.1 K equates to a change in relative humidity of approximately 0.5% at room temperature, as demonstrated by the psychometric chart below.

Solving system control problems

The impact of this inherent relationship between temperature and relative humidity resulted in control of the systems being limited to ±0.5 K for temperature, and ±3% for relative humidity. While these values exist within the tolerances required for the project, we found that the system fought itself to maintain control. For example, if we needed a minor increase in temperature, we also would need a corresponding rise in humidity - achieved through the increase in the refrigeration to scavenge water out of the system. We avoided this issue by converting the RH setpoint to absolute humidity (AH), and using this value as the basis for the addition or removal of water into the atmosphere inside the chamber. A change of 0.1 K has an almost negligible effect on the absolute humidity, making it much simpler to control. In the example described above, if the system required an increase in temperature through the use of increased heating, no change would occur to the AH level, so no changes would occur to the level of scavenging required by the refrigeration. Note: If the system required an increase in temperature and the AH was at the required level, then the RH level would fall slightly below the required level and would rise with the temperature as a result of the heating.

Increasing response time from 24 hours to 30 minutes

We replaced the previous NSC solution with a real-time NI FieldPoint-based system, which achieved a significant improvement in both temperature and humidity monitoring and control, and increased response time from 24 hours to 30 minutes. We can easily configure the new system using a database front end, and we can quickly alter it using the NI LabVIEW Real-Time development system and FieldPoint hardware.