Saturday, September 25, 2010

What is a PLC?

Typical PLC

A PLC (i.e. Programmable Logic Controller) is a device that was invented to replace the necessary sequential relay circuits for machine control. The PLC works by looking at its inputs and depending upon their state, turning on/off its outputs. The user enters a program, usually via software, that gives the desired results.

PLCs are used in many "real world" applications. If there is industry present, chances are good that there is a plc present. If you are involved in machining, packaging, material handling, automated assembly or countless other industries you are probably already using them. If you are not, you are wasting money and time. Almost any application that needs some type of electrical control has a need for a plc.

For example, let's assume that when a switch turns on we want to turn a solenoid on for 5 seconds and then turn it off regardless of how long the switch is on for. We can do this with a simple external timer. But what if the process included 10 switches and solenoids? We would need 10 external timers. What if the process also needed to count how many times the switches individually turned on? We need a lot of external counters.

As you can see the bigger the process the more of a need we have for a PLC. We can simply program the PLC to count its inputs and turn the solenoids on for the specified time.

This site gives you enough information to be able to write programs far more complicated than the simple one above. We will take a look at what is considered to be the "top 20" plc instructions. It can be safely estimated that with a firm understanding of these instructions one can solve more than 80% of the applications in existence.
 
That's right, more than 80%! Of course we'll learn more than just these instructions to help you solve almost ALL your potential plc applications.

An Example of Animated PLC Working:

counter animation

FIND MORE ON REMAINING POSTS............

PLC History

In the late 1960's PLCs were first introduced. The primary reason for designing such a device was eliminating the large cost involved in replacing the complicated relay based machine control systems. Bedford Associates (Bedford, MA) proposed something called a Modular Digital Controller (MODICON) to a major US car manufacturer. Other companies at the time proposed computer based schemes, one of which was based upon the PDP-8. The MODICON 084 brought the world's first PLC into commercial production.

When production requirements changed so did the control system. This becomes very expensive when the change is frequent. Since relays are mechanical devices they also have a limited lifetime which required strict adhesion to maintenance schedules. Troubleshooting was also quite tedious when so many relays are involved. Now picture a machine control panel that included many, possibly hundreds or thousands, of individual relays. The size could be mind boggling. How about the complicated initial wiring of so many individual devices! These relays would be individually wired together in a manner that would yield the desired outcome. Were there problems? You bet!

These "new controllers" also had to be easily programmed by maintenance and plant engineers. The lifetime had to be long and programming changes easily performed. They also had to survive the harsh industrial environment. That's a lot to ask! The answers were to use a programming technique most people were already familiar with and replace mechanical parts with solid-state ones.

In the mid70's the dominant PLC technologies were sequencer state-machines and the bit-slice based CPU. The AMD 2901 and 2903 were quite popular in Modicon and A-B PLCs. Conventional microprocessors lacked the power to quickly solve PLC logic in all but the smallest PLCs. As conventional microprocessors evolved, larger and larger PLCs were being based upon them. However, even today some are still based upon the 2903.(ref A-B's PLC-3) Modicon has yet to build a faster PLC than their 984A/B/X which was based upon the 2901.

Communications abilities began to appear in approximately 1973. The first such system was Modicon's Modbus. The PLC could now talk to other PLCs and they could be far away from the actual machine they were controlling. They could also now be used to send and receive varying voltages to allow them to enter the analog world. Unfortunately, the lack of standardization coupled with continually changing technology has made PLC communications a nightmare of incompatible protocols and physical networks. Still, it was a great decade for the PLC!

The 80's saw an attempt to standardize communications with General Motor's manufacturing automation protocol(MAP). It was also a time for reducing the size of the PLC and making them software programmable through symbolic programming on personal computers instead of dedicated programming terminals or handheld programmers. Today the world's smallest PLC is about the size of a single control relay!

The 90's have seen a gradual reduction in the introduction of new protocols, and the modernization of the physical layers of some of the more popular protocols that survived the 1980's. The latest standard (IEC 1131-3) has tried to merge plc programming languages under one international standard. We now have PLCs that are programmable in function block diagrams, instruction lists, C and structured text all at the same time! PC's are also being used to replace PLCs in some applications. The original company who commissioned the MODICON 084 has actually switched to a PC based control system.

Inside plc

The PLC mainly consists of a CPU, memory areas, and appropriate circuits to receive input/output data. We can actually consider the PLC to be a box full of hundreds or thousands of separate relays, counters, timers and data storage locations. Do these counters, timers, etc. really exist? No, they don't "physically" exist but rather they are simulated and can be considered software counters, timers, etc. These internal relays are simulated through bit locations in registers. (more on that later)

The  parts inside


What does each part do?
  • INPUT RELAYS-(contacts)These are connected to the outside world. They physically exist and receive signals from switches, sensors, etc. Typically they are not relays but rather they are transistors.
  • INTERNAL UTILITY RELAYS-(contacts) These do not receive signals from the outside world nor do they physically exist. They are simulated relays and are what enables a PLC to eliminate external relays. There are also some special relays that are dedicated to performing only one task. Some are always on while some are always off. Some are on only once during power-on and are typically used for initializing data that was stored.
  • COUNTERS-These again do not physically exist. They are simulated counters and they can be programmed to count pulses. Typically these counters can count up, down or both up and down. Since they are simulated they are limited in their counting speed. Some manufacturers also include high-speed counters that are hardware based. We can think of these as physically existing. Most times these counters can count up, down or up and down.
  • TIMERS-These also do not physically exist. They come in many varieties and increments. The most common type is an on-delay type. Others include off-delay and both retentive and non-retentive types. Increments vary from 1ms through 1s.
  • OUTPUT RELAYS-(coils)These are connected to the outside world. They physically exist and send on/off signals to solenoids, lights, etc. They can be transistors, relays, or triacs depending upon the model chosen.
  • DATA STORAGE-Typically there are registers assigned to simply store data. They are usually used as temporary storage for math or data manipulation. They can also typically be used to store data when power is removed from the PLC. Upon power-up they will still have the same contents as before power was removed. Very convenient and necessary!!

PLC Operation

A PLC works by continually scanning a program. We can think of this scan cycle as consisting of 3 important steps. There are typically more than 3 but we can focus on the important parts and not worry about the others. Typically the others are checking the system and updating the current internal counter and timer values.

A PLC scan

Step 1-CHECK INPUT STATUS-First the PLC takes a look at each input to determine if it is on or off. In other words, is the sensor connected to the first input on? How about the second input? How about the third... It records this data into its memory to be used during the next step.

Step 2-EXECUTE PROGRAM-Next the PLC executes your program one instruction at a time. Maybe your program said that if the first input was on then it should turn on the first output. Since it already knows which inputs are on/off from the previous step it will be able to decide whether the first output should be turned on based on the state of the first input. It will store the execution results for use later during the next step.

Step 3-UPDATE OUTPUT STATUS-Finally the PLC updates the status of the outputs. It updates the outputs based on which inputs were on during the first step and the results of executing your program during the second step. Based on the example in step 2 it would now turn on the first output because the first input was on and your program said to turn on the first output when this condition is true.

After the third step the PLC goes back to step one and repeats the steps continuously. One scan time is defined as the time it takes to execute the 3 steps listed above.

Response Time

The total response time of the PLC is a fact we have to consider when shopping for a PLC. Just like our brains, the PLC takes a certain amount of time to react to changes. In many applications speed is not a concern, in others though...

If you take a moment to look away from this text you might see a picture on the wall. Your eyes actually see the picture before your brain says "Oh, there's a picture on the wall". In this example your eyes can be considered the sensor. The eyes are connected to the input circuit of your brain. The input circuit of your brain takes a certain amount of time to realize that your eyes saw something. (If you have been drinking alcohol this input response time would be longer!) Eventually your brain realizes that the eyes have seen something and it processes the data. It then sends an output signal to your mouth. Your mouth receives this data and begins to respond to it. Eventually your mouth utters the words "Gee, that's a really ugly picture!".

Notice in this example we had to respond to 3 things:

INPUT- It took a certain amount of time for the brain to notice the input signal from the eyes.

EXECUTION- It took a certain amount of time to process the information received from the eyes. Consider the program to be: If the eyes see an ugly picture then output appropriate words to the mouth.

OUTPUT- The mouth receives a signal from the brain and eventually spits (no pun intended) out the words "Gee, that's a really ugly picture!"
Beauty is in the eye of the beholder...

Total response time

Response Time Concerns

Now that we know about response time, here's what it really means to the application. The PLC can only see an input turn on/off when it's looking. In other words, it only looks at its inputs during the check input status part of the scan.

Scan- check input status

In the diagram, input 1 is not seen until scan 2. This is because when input 1 turned on, scan 1 had already finished looking at the inputs.

Input 2 is not seen until scan 3. This is also because when the input turned on scan 2 had already finished looking at the inputs.

Input 3 is never seen. This is because when scan 3 was looking at the inputs, signal 3 was not on yet. It turns off before scan 4 looks at the inputs. Therefore signal 3 is never seen by the plc.

Minimum input time
To avoid this we say that the input should be on for at least 1 input delay time + one scan time.
But what if it was not possible for the input to be on this long? Then the plc doesn't see the input turn on. Therefore it becomes a paper weight! Not true... of course there must be a way to get around this. Actually there are 2 ways.
Pulse stretch function. This function extends the length of the input signal until the plc looks at the inputs during the next scan.( i.e. it stretches the duration of the pulse.)Pulse stretch function
Interrupt function. This function interrupts the scan to process a special routine that you have written. i.e. As soon as the input turns on, regardless of where the scan currently is, the plc immediately stops what its doing and executes an interrupt routine. (A routine can be thought of as a mini program outside of the main program.) After its done executing the interrupt routine, it goes back to the point it left off at and continues on with the normal scan process.Interrupt function
Now let's consider the longest time for an output to actually turn on. Let's assume that when a switch turns on we need to turn on a load connected to the plc output. 

The diagram below shows the longest delay (worst case because the input is not seen until scan 2) for the output to turn on after the input has turned on. 

The maximum delay is thus 2 scan cycles - 1 input delay time.

Maximum delay time


It's not so difficult, now is it ?

Relays

Now that we understand how the PLC processes inputs, outputs, and the actual program we are almost ready to start writing a program. But first lets see how a relay actually works. After all, the main purpose of a plc is to replace "real-world" relays.


We can think of a relay as an electromagnetic switch. Apply a voltage to the coil and a magnetic field is generated. This magnetic field sucks the contacts of the relay in, causing them to make a connection. These contacts can be considered to be a switch. They allow current to flow between 2 points thereby closing the circuit.
Let's consider the following example. Here we simply turn on a bell (Lunch time!) whenever a switch is closed. We have 3 real-world parts. A switch, a relay and a bell. Whenever the switch closes we apply a current to a bell causing it to sound.

Simple relay circuit

Notice in the picture that we have 2 separate circuits. The bottom(blue) indicates the DC part. The top(red) indicates the AC part.

Here we are using a dc relay to control an AC circuit. That's the fun of relays! When the switch is open no current can flow through the coil of the relay.

 As soon as the switch is closed, however, current runs through the coil causing a magnetic field to build up. This magnetic field causes the contacts of the relay to close. Now AC current flows through the bell and we hear it. Lunch time!

A typical industrial relayA typical industrial relay