PID control can be tricky and cumbersome; take a look at the two-wire pressure transmitters that come equipped with a universal model regulator function – how simple is it to carry out regulation in this case?
Thread Content
The last edit to this post was made by 1111111 on 2015-8-22 at 17:47. 1. Let’s take a look at a diagram first. This diagram was sent to me by moderator Tianze a few days ago; he asked me how to adjust the three PID parameters. Looking at this serrated waveform, I didn’t reply to Tianze. Why didn’t I reply? Because when such a waveform appears, it means that the PID parameters cannot be adjusted properly. I didn’t want to make Tianze waste his time. Then why can’t the well-set PID parameters be adjusted properly in this case? Well, you see, the PID regulator uses discrete PID formulas in its operations. I can’t go into too much detail on this – first, it would take too much effort to explain, and I don’t get paid for that {:3_51:}. Second, you might find it boring to read. Third, even if you understand the theory, you still can’t modify the PID settings, change its control logic, or adjust the signal sampling frequency. So, just remember: when the waveform is jagged due to parameter issues, there’s no point in trying to fix it; you’ll just have to settle for it. The only possible solution for a jagged waveform seems to be to violate the physical principles behind PID theory by setting the P value to a negative number. When you run out of options, you can try that method, but I can’t guarantee it will work. 2. After talking about PID for so long, let’s return to the topic of today’s post. About 9 years ago, I thought to myself: why not send a linear 4–20mA signal from a pressure transmitter to the PID regulator, so that it can output a nonlinear 4–20mA signal to control things like valve openings? Instead of letting that damned PID generate a jagged waveform and cause problems, why not simply set the relationship between the output current and the adjustment parameters myself? That would solve the problem, right? !
This post was last edited by 1111111 on 2015-8-22 at 18:14. 3. The shortest distance between two points is a straight line, as said by Teacher Fan Tong, who plays the piano with short limbs and fingers from junior high school. So, 9 years ago, I designed a regulator that allows for free programming in 2 to 17 steps. About 5 years ago, I transferred the functionality of this freely programmable regulator to products such as pressure transmitters and level transmitters. In other words, our village’s SC329 pressure transmitter and SC829 level transmitter no longer merely output linear 4–20mA signals over two wires; they can also output nonlinear 4–20mA signals across 2 to 17 steps via those two wires. Therefore, they can also be referred to as universal, freely programmable two-wire super regulators. These devices also support cyclic display of current, pressure, or level, making it easy to determine during debugging whether the current adjustment is too fast or too slow, whether it’s too high or too low, and whether the adjustment range is too wide or too narrow – all at a glance! It’s obvious – if it doesn’t work, just adjust the current value, or the pressure and level values. Isn’t that simpler than randomly fiddling with those three PID values in a guesswork manner? Even if, for the sake of argument, you were that stupid (stupider than pigs and me), I don’t think that using them to regulate things would result in the pressure level fluctuating in a jagged manner and becoming uncontrollable Right? !
The last edit to this post was made by 1111111 on 2016-5-16 at 19:25. 4. This is a fully waterproof integrated type; it allows for on-site programming, and it outputs a nonlinear 4–20mA signal that is sent over long distances to the actuator. 5. This is a separate disc-mounted type, with the sensor located at the measurement point; the signal cable runs for dozens or even hundreds of meters to the control room. You can choose whatever arrangement works best for you. It features a waterproof display enclosure that can be installed within a dashboard, and it sends a nonlinear 4–20mA signal from the enclosure to the actuator. 6. This is a separate standalone type, with a waterproof display enclosure that can be inserted through holes in a panel and secured using nuts; you can place it anywhere you want. It also sends a nonlinear 4–20mA signal from the back of the enclosure to the actuator
The last edit to this post was made by 1111111 on 2015-8-22 at 18:43. 7. A few days ago, a netizen posted a request for help; he said, “The pressure before the valve is 1 kPa, and the pressure after the valve needs to be maintained between 600–800 Pa.” I used his case as an example to explain how to use SC329 and SC829 to achieve reliable pressure regulation. As shown in the diagram: The blue line represents the curve of the regulated current output that I planned. The red dots indicate the set points, That is, the points where the output current corresponds to a specific pressure value. If you look closely at the diagram, you’ll see that I divided the ideal pressure range of 600–800 Pa into three segments: 600–650 Pa, 650–750 Pa, and 750–800 Pa. In particular, for the core segment of 650–750 Pa, I kept the regulated current constant. Why? I have been involved in the design of several ultra-stable and ultra-sensitive pressure calibration pumps, as well as fully automatic pressure controllers. Although these are all small handheld or desktop devices, these experiences have given me a deeper understanding of the static and dynamic behavior of pumps and valves than that of ordinary control engineers. Taking this opportunity, I would like to share this concept of \"signal freezing\" control with everyone for reference. In other words, for any precise control – such as pressure regulation with a fluctuation rate of less than 0.005% FS – it certainly isn’t achieved through continuous adjustment as claimed by PID controllers; the regulator must have periods during which it stops functioning. Especially in micro-pressure systems like those I work with, the more randomly the control valve adjusts, the easier it is for the system pressure to become unstable
This post was last edited by 1111111 on 2016-5-16 at 19:44. It features pressure SC329 and level SC829; it can also be used as a manual 4–20mA signal generator. This feature makes it easy to determine the optimal driving current for a specific pressure value during debugging. The operation and setup methods for these are detailed in the accompanying documentation; see:
This post was last edited by 1111111 on 2015-8-22 at 19:33. 9. If you’re not afraid of high prices of 1300 yuan for products that require no covering, come to me. If Yuandian Euroland can’t handle it, come to me. If screws of the MENG brand or those manufactured in Japan can’t be produced or are of poor quality, come to me. . . For example, a range of ultra-low pressure of ±25 Pa, or a gauge pressure range of 0–100 Pa {:3_51:} At the halfway point of life, I’m just an enthusiast who enjoys working with insects,,, Can work also be considered a hobby? In my eyes, it’s all about skill; there are no masters
Different magnification factors are used for different deviations; this constitutes the basic framework of override control. The approach used is creative; thumbs up.
I also think PID isn’t very effective; why isn’t there any improvement?
:victory::) It’s a really high-end thing; I learned about it*
It is recommended that the original poster learn about fuzzy control