How to Set Input Filter Value for Accurate Pulse Capture with S7-1200 PLC and Paddle Wheel Flowmeter

Hello everyone, let's delve into the specifications. The maximum flow rate is 211 gallons per minute and the K factor is 38.33 pulses per gallon. By multiplying the flow rate by the K factor, we get 8087.63 pulses per minute. Converting this to pulses per second, we get 134.79 pulses per second. The duration of each pulse is approximately 0.0018548 seconds (or 1.854 milliseconds / 1854.73 microseconds). To ensure accurate pulse counting by the HSC, it is advisable to select the lowest DIG IN input filter value of 0.1 microseconds. This guarantees that all pulses lasting longer than 0.1 microseconds are properly recorded. If you have any additional information to add or correct, please feel free to chime in. Thank you.

• 23-02-2025
• abhmeh1991

abhmeh1991 inquired about choosing the lowest DIG IN input filter value of 0.1 microsec to ensure that all pulses with a duration exceeding 0.1 microsec are registered by the HSC. Can someone please confirm if any information is missing? Great job with the calculations! Based on @BryanG's assumed 25%/75% ON/OFF duty cycle for each pulse, the shortest pulse duration would be 1855µs (1.855ms). It seems that @BryanG is suggesting selecting a filter value that is the largest but still below the 1.855ms threshold.

• 23-02-2025
• drbitboy

When interpreting @BryanG's advice, it is recommended to select the filter value closest to but less than 1.855ms. For example, on the S7-1200, a suitable filter value would be 1.6 milliseconds, but opting for 0.8 milliseconds may be more prudent in case the flow rate unexpectedly increases. However, if screened cable is used with short cable lengths, lower values may suffice. In some cases, flowmeters with varying pulse rates, such as 1000 pulses per litre to 32000 pulses per litre, are utilized. Despite this variability, utilizing a filter value of 0.8 microseconds has not resulted in false counts in my experience.

• 24-02-2025
• BryanG

Hello everyone, I'd like to explain my method for calculating flow rate. I count the number of pulses recorded by the High Speed Counter (HSC) in one minute, and then divide that by the K-factor to determine the flow rate in gallons per minute (GPM). The HSC is reset by a timer every minute. Can anyone confirm if this is the correct approach? Is it possible to monitor the pulse count in real time instead of updating the flow rate value every minute and resetting the HSC if no pulses are detected at the input?

• 25-02-2025
• abhmeh1991

To optimize processing efficiency and capture all pulses without causing interruptions, it is recommended to keep the HSC interrupt at the smallest possible setting. For example, instead of waiting a full minute for an update, consider using a 10-second interval for quicker feedback. One method to achieve this is to store counts in an array every 10 seconds, allowing for a faster update while still obtaining the average packs per minute. This approach ensures a more efficient and accurate running average calculation.

• 25-02-2025
• parky

abhmeh1991 stated: Hello everyone, I wanted to explain my method for calculating flow rate. I count the pulses by using the High-Speed Counter (HSC) every minute and then divide by the K-factor to determine the flow rate in GPM. The HSC is reset by a timer every minute. Can someone confirm if this is the correct approach? Is it possible to count pulses in real time instead of updating the flow rate value every minute and resetting the HSC if no pulse is detected at the input? I strongly recommend against resetting the counter because doing so will result in losing part of a pulse. Instead, consider subtracting the last saved value from the current value at regular intervals (based on a timed interrupt) to calculate the delta over that time period. Then, save the current value for the next calculation. Assuming that the HSC count is a double integer, perform all calculations using double integers to handle the rollover effectively.

• 25-02-2025
• Manglemender

Even though the measurement is in gallons per minute, you don't have to take a reading only once a minute. Similar to how a car speedometer displays miles or kilometers per hour continuously, you can take readings more frequently for better accuracy. For example, if you calculate 134.79 pulses per second, you could take a reading every second and then multiply the result by 60 to get the flow per minute. The timing of the readings depends on the level of precision required and the consistency of the flow being measured. Using a timed interrupt to count pulses may have its challenges, as the interrupt could occur just before or after a pulse. However, a small difference of 134 or 135 pulses per second amounts to less than a 1% discrepancy, which could be acceptable for stable flows. For more erratic flows, averaging or adjusting the frequency of readings may be necessary. In terms of selecting a timer for triggering the count, refer to the Interrupt section in the Extended Instructions. A cyclic interrupt could be a suitable option. Keep in mind that sensor readings always reflect past events rather than the present moment, as there is a delay in outputting readings and converting analog inputs. Regarding the issue of counting over multiple pulses, it is essential to address scenarios where no pulses are received. Employing a Watchdog Timer can help ensure that the system continues to function properly by forcing a count after a set period. However, managing timed and count interrupts simultaneously can be complex. Following Manglemenders' suggestion of not resetting counters or timers can simplify the process, although it adds complexity. In conclusion, sticking with a time-based system can be a practical approach to avoid potential complications in managing interrupts and counts effectively.

• 25-02-2025
• BryanG

Hello BryanG, I want to thank you for your explanation. Currently, I am resetting the HSC every minute as shown in the attached image, but I understand that this may not be the ideal method. Every minute, the value is transferred to a cyclic interrupt, and then I divide the number of pulses by the K Factor to determine the Flow Rate in Gallons Per Minute (GPM). If I change the cyclic interrupt to 1 second and reset the HSC every 1 second, would this logic be accurate? In a paddle wheel flowmeter, if no pulses are detected, the flow rate is 0 GPM. Currently, the PLC in my setup waits for one minute before updating the flow rate value.

• 25-02-2025
• abhmeh1991

One issue with relying on a standard timer is that it is reliant on the PLC's cycle time. While this may not make a difference in certain situations, for greater accuracy, consider utilizing a Cyclic Interrupt. This function, which supersedes the initial line in your screen capture, causes the PLC to pause its current task when an interrupt occurs, execute the corresponding OB, and then resume normal operation. In response to your inquiry about the absence of pulses, the equation is simple: 0 pulses divided by the K factor equals zero. Thus, a lack of pulses will result in a calculated flow of zero.

• 25-02-2025
• BryanG

In the world of PLC programming, relying on a standard timer may not always provide the precision you need due to the PLC's cycle time. Instead, one effective solution is to implement a Cyclic Interrupt. This method allows you to take control of your process at a more precise level, ensuring accurate results. When an interrupt occurs, the PLC pauses its current tasks to execute a designated OB, before seamlessly returning to its regular program cycle. If you're concerned about handling zero pulses, rest assured that the calculation will yield zero when dividing by the K factor. By incorporating a Cyclic Interrupt OB into your programming, you can efficiently reset your High-Speed Counter (HSC) at regular intervals, such as every second. Alternatively, integrating a timer within the Cyclic Interrupt OB can also achieve the same result. Whichever approach you choose, make sure it aligns with your current logic flow for calculating the Flow Rate based on encoder counts. Check out the attached image for a visual representation of the process.

• 25-02-2025
• abhmeh1991

Managing time in the HSC can be challenging due to its asynchronous nature. It is advisable to follow @Manglemender's suggestion of keeping track of the previous last value and calculating the difference with the current value, rather than resetting the counter. Rollover is the main issue that may arise, but it can be easily detected and managed. If resetting the counter is necessary, it is recommended to have the counter reset internally when it reaches a specific value (e.g. 1,000,000) to avoid losing any "partial" pulses. This approach ensures smooth operation without interruptions.

• 25-02-2025
• drbitboy