Numerous discussions have surfaced regarding the integration of analog flow sensor signals to calculate total volume, with some doubts raised about its accuracy and reliability. Some experts suggest utilizing the pulse output feature of flow meters for a more dependable measurement. In an attempt to explore this further, I conducted an experiment using a 25-year-old GE Fanuc 90-30 CPU351 in ladder logic. Rather than using analog signals, I simulated a constant flow rate of 100 (e.g. GPM) with a sinusoidal fluctuation of +/- 10 percent, oscillating between 90 and 110 at a specific frequency. I tested three different integration methods: time-based integration with uniform sample intervals, sample-based integration with consistent intervals, and time-based integration with samples taken during each scan. The results, along with the ladder logic code, can be found in the provided attachment. Feedback and critiques on my experiment are welcome, as I am open to suggestions for alternative approaches or additional tests for future exploration. Feel free to examine the details and provide your insights.
I enjoy conducting simulations and experimenting with various methods. When provided with counts, I opt to simply tally pulses over time. Utilizing the real-time clock is essential for accurately measuring the time period. It is assumed that the flow sensors can effectively respond to changes in flow frequency. Another consideration is the necessity of calculating the total volume. To address issues with analog-only systems, it is crucial to have short intervals between samples to avoid missing brief bursts of flow. Implementing a 10 ms timed interrupt can be beneficial in these cases. Monitoring the bandwidth of the analog flow sensor and input card is key, and using a scope to examine noise levels or data filtration in the analog output of the flow sensor is worth considering.
After setting out to integrate a signal in ladder logic, Steve Bailey achieved excellent results. Click to learn more about his impressive work!
Great job, Mr. Bailey! Your point about achieving the best accuracy with a pulse output is well taken. I recall a project in the fuel sector that utilized a state-of-the-art mass flow meter, which accounted for variables like temperature and pressure. This just scratches the surface of other influencing factors. Regarding analog concerns, I echo Peter's worry about noise interference. It's been my experience that the more precise the measurement, the more variables come into play. Wouldn't scan-based sampling offer a quicker alternative to analog conversion time? There are many valuable insights to be gleaned from your experiment. Thank you for sharing!
Is scan-based sampling a quicker option compared to analog conversion time? The speed will vary based on the analog module and scan time. Specifically, the 90-30 analog module had a conversion time of 2.5 milliseconds for all channels. In my particular application, longer scan times were required, so this wasn't an issue. However, with a newer and faster model PLC, this could potentially pose a problem.
The document referenced in the initial post of this thread is no longer available as it was in MS Word format (.docx), which is not currently supported among the list of acceptable file extensions. It is possible that this has changed since December. Nevertheless, I have converted the document to a PDF for accessibility in case individuals stumble upon this thread at a later date.
Thanks for conducting such a detailed experiment, it's intriguing and inspiring. On first glance, your choice to use a 25-year-old GE Fanuc 90-30 CPU351 certainly got my attention due to its age and the unique challenge that presents. I appreciate your methodical approach to settling this long-debated issue. I'll dig into your attachment in more detail soon, as I'm interested to see the specifics of your integration methods. It's a testament to creative problem-solving. Perhaps additionally, a comparison between your analog sensor outcomes and a modern digital flow sensor may provide valuable context and further enhance this study. Anyway, kudos for the good work, it's a hefty contribution to elucidating this issue!
This is a substantial contribution to the conversation, and I applaud your initiative in conducting your own experiment. Your choice to use the GE Fanuc 90-30 CPU351 provides an interesting case study due to its age. I'm currently examining your ladder logic code and findings. From a cursory view, your three integration methods seem like a solid ground for comparison. I'm eager to dig deeper, particularly into your method of time-based integration with samples taken during each scan, as I believe that could potentially offer a more realistic estimate of flow rate fluctuations. I'll certainly get back after exploring the details more thoroughly. Thanks for stirring up such an engaging discussion!
This is a fascinating experiment, and it's great to see you experimenting with different integration methods! I’d be curious to know how the accuracy of your total volume calculations compared across the various techniques, especially with that sinusoidal fluctuation. Time-based integration often sounds promising, but sometimes it can introduce errors if the sampling isn't perfectly timed with the flow changes. Have you considered using a moving average filter to smooth out the fluctuations before integration? That might help enhance the reliability of your measurements. Also, for future tests, it could be interesting to compare your findings with a known calibration standard to see just how close you can get to the actual volume. Looking forward to your insights!
This is a fascinating experiment, and I appreciate you sharing your methodology and results! Your approach to simulating flow rates with a sinusoidal fluctuation is quite clever; it really helps to highlight potential drifts in accuracy with different integration methods. One area you might consider exploring further is the impact of varying the frequency of your oscillation on the integration results—this could mimic real-world scenarios where flow rates fluctuate unpredictably. Additionally, have you thought about incorporating some drift compensation in your calculations to see how that affects overall accuracy? Looking forward to seeing your results and any adjustments you make in future tests!
✅ Work Order Management
✅ Asset Tracking
✅ Preventive Maintenance
✅ Inspection Report
We have received your information. We will share Schedule Demo details on your Mail Id.
Answer: - Pulse output integration from flow meters can provide a more dependable measurement by utilizing discrete pulses to calculate flow rate and total volume, which may offer better accuracy and reliability compared to analog signals that are prone to noise and interference.
Answer: - Utilizing the pulse output feature of flow meters can simplify the measurement process, improve accuracy, and enhance reliability by providing a digital signal that can be easily integrated into ladder logic programming for precise calculations.
Answer: - The experiment simulated a constant flow rate with sinusoidal fluctuation and tested three integration methods: time-based integration with uniform sample intervals, sample-based integration with consistent intervals, and time-based integration with samples taken during each scan. The results and ladder logic code can be found in the provided attachment.
Answer: - The results of the experiment can offer insights into the effectiveness of different integration methods and help in identifying the most suitable approach for optimizing flow measurement in ladder logic programming. This information can guide future improvements and refinements in flow measurement systems.
Join hundreds of satisfied customers who have transformed their maintenance processes.
Sign up today and start optimizing your workflow.