Please help me to understand the meaning of scale type (CIM, CIM_I), and what is the purpose for using contact inversion?

Thanks in advance.

 

CIM means direct contact, and CIM_I is the inverted contact. If you carefully go through the MARK V I/O List you will find that all the critical digital inputs are inverted. More elaborately, all the critical field inputs are normal to close type and they are inverted in MARK V to change this contact to open type. Normal to close is used for safety because if any wire opens you will come to know from alarm immediately.

Otherwise if you use open to normal contact you will never know about the wire break under normal conditions, and machine will not trip even if the critical parameters go out of range. Hope this is what you are asking for.

 

I have heard many theories on the subject of contact inversion. The one I believed was from a GE Field engineer.

GE purchases many devices from many different companies in many different countries. When a turbine and all of the devices are assembled and connected to the Mark V and the generic program that shipped with Mark V, it is easier to invert a contact in IO Configurator than to investigate every instance of that contact in the CSP.

This inversion technique can be a nightmare for a technician. For instance: 4 identical turbines at the same site had 2 turbines with the contact inverted and 2 did not on the same identical switch respectively. It turned out they were wired differently with wiring on the normally open on 2 turbines and the normally closed on the other two. Instead of insisting that the wiring match on all 4 turbines, signals were inverted.

So, the answer is convenience.

I hope to hear some more theories to add to my list.

 

Great question (the second part, anyway).

Many decades ago, the designers of GE heavy duty gas turbine control systems (and possibly even steam turbine control systems) made some basic decisions about how their systems would operate. One of the decisions was that they were going to use a closed circuit to indicate the *absence* of trouble in a system parameter that was being monitored by something like a limit switch, or a temperature switch, or a pressure switch. A closed circuit means that current would be flowing through the contacts of the sensor device when the condition being sensed was "normal" for the given operating condition.

This meant that when the sensing device indicated trouble (valve in a particular position, or high temperature, or low pressure) the contacts of the switch would open. Open sensor contacts would cause current flowing in the circuit to the control system to stop. This also means that a broken wire or a wire that comes loose from a terminal in the circuit, either of which would also result in an open circuit and the cessation of current flow, would "appear" to the control system as a trouble condition. A trouble condition was annunciated as an alarm for the operators.

Back then the method of annunciating an alarm was to illuminate a "trouble lamp" and sound an audible alarm. Lighting an electrical lamp (bulb) or energizing an alarm bell requires current to flow through the filament of the lamp or the coil of the alarm bell mechanism. But, the sensor circuit opens when there's trouble, and it's necessary for a contact in the control system lamp circuit and the alarm bell circuit to close for current to flow to light the lamp or sound the alarm bell. So, the control system designers "inverted" the control system contact input: when the sensor circuit went open, the control system monitoring that circuit would close it's corresponding contacts to cause current to flow to light the trouble lamp and/or sound the alarm bell.

It's as "simple" as that. They decided to use an open sensor contact (circuit) to indicate trouble on the unit, and because they needed a corresponding contact in the control system to close to light a trouble lamp or sound an audible alarm, they "inverted" the corresponding control system input. When the sensor contact and circuit is closed, the corresponding control system contact is open; and when the sensor contact or circuit opens, the corresponding control system contact closes to annunciate a trouble condition.

Why do this? Consider the opposite: using a sensor contact that closes on an abnormal or trouble condition. Doesn't seem like a problem, right? Why do we care if the sensor contact opens or closes when there's a problem--as long as it changes state? Now, consider what happens if a wire in that temperature switch circuit comes loose from it's terminal and there's some problem with the lube oil cooler and the lube oil header temperature increases above the switch setpoint. The temperature switch contact will close but because a wire has come loose in the circuit there will be no current flowing in the loop, and the control system which is expecting current flow to say that a high lube oil temperature exists won't ever know the contact has closed because of the loose wire. And the lube oil header temperature continues to rise, until the temperature exceeds the oil's flash point and the vapors ignite in the bearing housing inside the exhaust area.

So, they decided to use closed sensor contacts to indicate normal, running conditions. And to use open sensor contacts to indicate alarm/trip conditions. To do so, required a method of closing control system contacts to light a trouble lamp or sound an alarm bell--that's the inversion part. In general, if GE finds something that improves reliability and works, they stick with it. And they've done that through several generations of Speedtronic turbine control systems.

Now this all seems overly complicated, and it is until you really stop and consider it. Sometimes it takes a few days, or even weeks, to really comprehend the whole situation considering the pros and cons of using sensor contacts that open on trouble or close on trouble. But, when you finally understand all the implications it's like a million candle-power light being turned on, and you wonder why every control system manufacturer doesn't do the same thing.

And you're asking yourself now, "Why make this so complicated?" It's all about reliability. If a wire in a critical circuit comes loose, don't you want to know about that as soon as possible? Well, if the control system is waiting for a closed sensor contact (circuit) to indicate there's a problem and a wire for that circuit has been loose from it's termination for three minutes, or three hours, or three days, or three weeks, or three months, you're never going to know about it until there's been some kind of equipment failure, maybe catastrophic. But, if the control system interprets that loose wire as a problem and annunciates an alarm as soon as the wire comes loose, well, you're going to know about the problem in the circuit as soon as it occurs. It will be an "erroneous" alarm, because if the condition is okay (verified by looking at some pressure- or temperature gage, or verifying some valve position) you will (should) automatically suspect a wiring problem somewhere in the circuit. (It could be a control system problem, but that's not usually the case.)

Inverting contact inputs is an inexpensive method of something that's commonly referred to as "contact supervision" that's accomplished by modern control systems using "end-of-line" resistors to monitor for a break in the circuit, or loop. Some modern-day control systems use expensive circuits and require lots of end-of-line resistors to monitor for a loose or broken wire in the circuit. (With an end-of-line resistor, there's either "maximum" current flowing when the sensor contact is closed, or some minimal current flowing when the sensor contact is open; no current will flow when there's something wrong with the circuit such as a loose wire or broken terminal, etc.) In this case, it's still good engineering practice to use a sensor contact that opens on trouble, especially for critical conditions, but it's not absolutely necessary.

So, that explains (hopefully) why GE inverts contact inputs on their turbine control systems. There's LOTS more to this whole scheme that, if one understands all the nuances, makes reading relay ladder logic and function block diagrams very, very easy. But, we don't have enough space or time here.

Now, about the CIM/CIM_I scale type thing: They don't mean a thing. You can use either scale type for a contact input in IO.ASG; you can even use LOG (for LOGic) instead of CIM or CIM_I for the scale type for contact inputs in IO.ASG and the compiler and the Mark V will work just fine. The I/O Configurator is where one defines whether or not a contact input is inverted. The value in IO.ASG has absolutely zero effect (none; zippo; nada; niente) on whether or not a contact input is inverted in the Mark V or not. It's a "nice" feature if one can look at the scale type column in IO.ASG and see if the input is inverted in the I/O Configurator or not, but there's no guarantee that it matches what's defined in the I/O Configurator. And, again, changing the type in IO.ASG will not change the "behavior" of the contact input; it's going to behave as defined in the I/O Configurator.

There's no link between changes made to IO.ASG and those made in the I/O Configurator, and vice versa. In other words, changing a contact input's scale type in IO.ASG will not change its inversion mask value in the I/O Configurator (which is where it must be changed to change the behavior of the contact input). And, changing a contact input's inversion mask value in the I/O Configurator will not change the scale type in IO.ASG.

And the same goes for the I/O Report file, TC2KREPT.TXT. All of these files must be manually edited any time a change is made to reflect the change in all three places; there's no automatic function that updates the other files when a change is made in one.

Nowhere in the above did I say *anything* about normally open contacts or normally closed contacts; contacts were referred to only in their open or closed states.

 

CTTech,

GE has defined what the condition of every contact input should be. Many field engineers don't understand contact input inversions and the open to alarm/trip philosophy, and use the inversion mask to "fix" a wiring problem by making a software change. The Device Summary is where the definition is usually made.

You got fed a line of bull. And, you were the victim of poor start-up management. On a multi-unit site, every turbine should be wired similarly, and configured similarly. Unfortunately, there is no review or check of installations to ensure that wiring and software are what they should be. So, people cut corners, and get away with it.

 

In Speedtronic logic, "1" turns on the message on the annunciator panel (in Mark 1 and Mark II systems), on the CRT (on Mark IV systems) or on the <I> or <HMI> in Mark V systems.

If a field contact should initiate the annunciator message in the closed condition, it is not inverted. Else it is inverted. A classic example is the pressure switch 63QT that will be closed in the normal lube oil pressure condition. When the pressure drops, the switch will trip the unit and alarm the condition.

In Mark V systems, in the file IO.ASG, the unit of CIM or CIM_I makes no difference. It is only for information. As mentioned earlier, inversion is enabled or disabled in the IO configuration of the contact input module and then downloaded.

 

Great question.
Great explanation CSA.
Everyone new to process control should read that carefully.

Roy

 

Thanks, Roy; but we don't know if Richard (the originator) got what he was looking for.

Guess we have to presume no news is good news!

Sure wish we could use drawings here; that would have made the description a little easier.

 

In reply to CSA: You are looking for an "image hosting service". There are many such sites on the internet which exist to provide image hosting for discussion forums. Do a search on Google for "image host" and you will turn a great many.

Embedding arbitrary user images directly into forum pages is usually a bad idea from a security and usability standpoint. Instead, you upload your image to an image host which gives you a unique URL directly to that image. You then paste that link into your message. Anyone who wants to see it can use that link to view it with their web browser.

Here is an example that I did a couple of years ago using "tinypic.com" to answer this same question. http://i12.tinypic.com/29fxzs7.png

 

Thanks everybody who provided a lot of experiences to me.

I so appreciate it.

Richard
Beginner in Mark V