Life is easier when you go with the flow. Chem-Flow, whose customers often need help with a hard-to-solve dispensing problem or an unusual metering application, can attest to that.
Based in Chicago, IL, and in business since 1958, Chem-Flow is an engineering company specializing in manufacturing chemical feed systems. It serves a wide range of chemical processing markets, each with different needs and concerns.
Chem-Flow pays close attention to choosing the right drives when developing the most appropriate solution. As vp of operations Henry Glick explains, "When you''re designing systems whose function is to perform precise metering, transferring, and mixing of multiple chemicals, drives are obviously an intrinsically crucial component, and our customers need to take their performance for granted." Add the fact that the ingredients are often quite expensive and sometimes can deteriorate quickly, and the need for error-free operation becomes even more important. A malfunctioning drive can shut down a production line; an inaccurate drive will ruin a batch of product; a hard-to-operate drive affects efficiency, and the list goes on.
According to Glick, Chem-Flow has been using the entire range of AC Technology Corp''s drives -- "We''ve been working with AC Tech''s Michael Kenney, taking advantage of the company''s willingness to retrofit drives and software in order to meet our needs to the last spec." As far as the technology itself, Glick particularly appreciates the capabilities he gets through the MC 3000 Series drive''s PID (Proportional Integral and Derivative) loop.
"The PID lets our customers keep a tight rein on what''s being pumped," he explains.
What Is a PID Controller?
And what is its role in Chem-Flow''s dispensing and mixing systems? The PID feature allows a drive to hold the desired setpoint based on feedback from the process. Variables such as pressure levels, liquid flow rate, or liquid level, are detected by a transmitter, which sends the data to the variable frequency drive, for comparison to the setpoint.
Process systems often require a system-controlled parameter, such as motor speed, to be able to react to variable situations in order to keep another system attribute constant --pressure, flow, temperature, etc. A simple example is a metering and dispensing system that has multiple discharge valves. For flow to be repeatable at each valve, the pressure in the supply manifold must be held constant. If the pump supplying this system is powered by a drive, the drive speed will need to increase as valves are opened, and the drive speed will need to be reduced as valves are closed, in order to maintain a constant pressure in the manifold.
A means to meet this requirement is to use a "setpoint controller," where the pressure in the manifold is measured with a pressure sensor and this value is compared with a setpoint indicating the value that you want the pressure to be. A setpoint controller compares the setpoint value to the actual value and generates a speed command to the drive to correct the variance or error. The AC Tech MC3000 builds this setpoint controller function into the drive.
One of the most common types of setpoint controller uses a PID algorithm. (See figure) This stands for the 3 types of adjustments (referred to as "gains") that are used to correct for the error: Proportional, Integral, and Derivative adjustments. The Proportional gain is the most basic adjustment, where the speed command is directly proportional to the error. If Proportional gain is used alone, however, there will always be an error in the system -- if Proportional gain is set too low, system response will be quite sluggish; if it''s set too high, the system will oscillate or grow unstable.
To eliminate the error, the Integral gain is used; the Integral adjustment will continue to increase the output speed command based upon the accumulated error over time (or decrease the speed in the event of a negative error). Small amounts of Integral gain can have a significant effect on the setpoint controller''s performance. If set too high, the system will overshoot the setpoint, especially when large step changes occur in the error (i.e. large step changes in the setpoint or in the feedback).
The Derivative gain is used to enhance performance. It basically looks at the rate of change in the error and forces a more dramatic change to the speed command than the one achieved with just Proportional and Integral (PI) alone. While this can be very useful in position control systems, for example, because it can shorten the time required for a drive to respond to a change in the error, it can also lead to a system that overshoots the setpoint or even creates an unstable system. In most cases, the Derivative gain is set to zero or to some very low value to prevent this from happening. (See figure)
Direct and Reverse Acting Setpoint Controllers
Most setpoint controllers are "direct" acting. That is, an increase in the motor speed causes an increase in the process variable you want to change. This is the case on a pump system where pressure is the process variable; increasing the motor speed increases the system pressure. But in some systems an increase in motor speed creates a decrease in the process variable you want to control. Take the case, for example, of a fan blowing air over a heat exchanger, and the temperature of the fluid within the heat exchanger is the process variable you are trying to change. As the motor speed increases, the temperature of the fluid will decrease! In this case, you would need to use a reverse-acting controller in order to achieve the desired change.
Matching the drive technology with the demanding needs of its customers is another way for Chem-Flow to continue being a leading resource for metering and dispensing solutions. Chem-Flow now offers its customers systems featuring variable frequency drives with a 16-character backlit display, overload protection, and a very wide variety of voltages, since AC Technology provides a stainless steel drive in 110-20 V and 208-230 V single phase, 208-230 V and 380-480 V 3-phase, and 590 V option, all programmable in 50/60Hz.
Demystifying PID Control TuningFor motor control systems, tuning the PID gains can be tricky. The steps below can simplify this process:
- Set both Integral and Derivative gains to zero. By making large step changes to the error signal, increase the Proportional gain until the system oscillates around the setpoint. Then decrease this gain until the system is stable on a step change.
- Set the Proportional gain about 15% below this stabilized point.
- Again by making step changes to the error, increase the Integral gain until the system oscillates around the setpoint before stabilizing. Reduce the Integral gain until the system responds to a step change without oscillation or error.
- In most systems, setting the Derivative gain is not required. If the system requires a faster response time, however, it may be possible to achieve this by adjusting the Derivative gain. By making step changes to the error, increase the Derivative gain until the system stabilizes in the minimum time. Increase slowly to avoid an unstable condition. The optimal point will be when there is one overshoot of the setpoint.
- Test the system for stability by making step changes to the setpoint and system variables that affect the error to ensure that worse case scenarios do not create an unstable operating condition.