Three Parameters Controlling a Single Output

strebej

Member
Join Date
Sep 2002
Location
Neenah,WI
Posts
24
I really need some help on this one. I have three Oxygen / Combustables analyzers. Each one is set up to close a descrete dry contact when an alarm (High / Low Set Point) has been reached. I will progam the PLC to show the alarm status of each independent analyzer. The process is to have the three analyzers online at the same time. There is a main fuel trip (MFT) that will trip the boiler if more than ONE analyzer goes into an alarm status. The problem is I need to design a Logic diagram and produce a generic program, (PLC model not important). This program will need to be able to allow any one of the three analyzers to go into an alarm status and still remain online (No MFT). but! if any one of the other two goes into alarm then set the MFT. I will also need to be able to take any one of the three offline for maintainance. In this case then both of the other analyzers will need to have an alarm before a MFT is activated. Do you see my delema.
 
Please fill in the following information:

This is a complete list of my discrete inputs:


This is a complete list of my discrete outputs:


This is a complete list of my analog inputs:


================

The first discrete output in the above list comes on when:

The first discrete output in the above list goes off when:


- - - - - - -

Repeat the above for each output

=================

Under normal operation, the following sequence occurs:

Step 1 - General Description
Step 1 - What events should the PLC cause to happen in the real world.
Step 1 - What conditions must the real world meet before going on to step 2?


Step 2 - General Description
Step 2 - What events should the PLC cause to happen in the real world.
Step 2 - What conditions must the real world meet before going on to step 3?

- - - - - - - - -

Repeat the above for each step in your sequence.


================

When you do all this, we will be able to help you more. We've seen too many posts such as this one where all the info has been withheld, or the process was poorly defined, so please fill in the above.

Although perhaps when you do, you'll see what needs to be done yourself. It doens't sound that hard.

Good luck.
 
strebej,

I imagine your initial reaction when you saw Allen's post was that he was beating around the bush and not actually answering the question. How come he couldn't just try to help?

Well, I gotta say, I agree with him and his approach to this. The problem you describe SOUNDS like a pretty simple one, but it is NOT one that can safely be answered by somebody here without really seeing and knowing everything about your system. You are working with an inherently dangerous process.

If you follow his approach, you will probably see the answer to you question yourself. It looks to me like once you understand the logic of the system the how-to will be apparent.

Steve
 
three inputs to one output

I thought he made himself clear. Try this.

input 1 input 2 alarm
--| |--------| |----------( )
input 1 input 3 |
--| |--------| |---|
input 2 input 3 |
--| |--------| |---|

input 1 output 1
--| |---------( )

input 2 output 2
--| |---------( )

input 3 output 3
--| |----------( )

end
 
godfrey:

Yeah, but does it satisfy the following:

strebej said:
This program will need to be able to allow any one of the three analyzers to go into an alarm status and still remain online (No MFT). but! if any one of the other two goes into alarm then set the MFT.

Or
I will also need to be able to take any one of the three offline for maintainance.

That's why I want (he needs) a better definition.
 
Outputs 1 thru 3 will indicate individual alarm status. The output "alarm" in my first rung sets the MFT. Rung one will still work if any of the three units is offline.
 
Three Parameters Controlling a Single output

This is a complete list of my discrete inputs:
There are two each descrete input they are as follows: From the analyzer to the PLC is a "digital out fault" and "digital out measuring" are what I call the "watchdog timer alarm" and "process alarm". These are both normally closed dry relay contacts rated at 30 volts DC, 1 amp non-inductive maximum. The watchdog timer is a "hardware" alarm which would transfer in the case of a power failure or software glitch or lockup. The operating software periodically resets the hardware watchdog timer chip. If the programming gets "lost" or locks up,
the timer chip will time out and activate the watchdog timer alarm. This will cause the relay contact to transfer while simultaneously resetting the microprocessor to try to restore normal operation. The process alarm is a software alarm which monitors the operation of the analyzer. This alarm contact will transfer if the software detects abnormal operation, such as a thermocouple failure, over or under temperature, calibration failure, etc.
Essentially, the process alarm will activate any time the microprocessor determines the analyzer readings could be invalid. As noted above, the watchdog timer and process alarms are simple dry relay contacts which are normally closed and will open on alarm.


This is a complete list of my discrete outputs:
There is ONE output from the PLC to the analyzer this is a Calibration initailization contact. The PLC closes this bit on a timed sequence to initailize a calibration sequence that is internal to the analyzer.

Now there are also TWO output contacts on the analyzer going to the PLC. These are the contacts that I'm reffering to. These contacts will close if the rage of the set points have been exceeded. ie.. One contact=High Oxygen levels or Low Oxygen level. The other contact=High Combustables or Low Combustables level. If any ONE of these contacts are activated the PLC will show alarm. If any TWO are activated at the same time or both at any given time, it does not matter when the contacts are set they just have to be on at the same time. Then the PLC will need to initate a MFT and trip the boiler.

This is a complete list of my analog inputs:
There are TWO AI's from the Analyzer to the PLC. The first is 4-20mA measuring Oxygen and the sceond is a 4-20mA measuring Combustables.


================
There are a total of SIX alarm cantacts coming from THREE analyzers. The analyzers have TWO each.

The first discrete analyzer output in the above list comes on when:
The Oxygen level is above or below the set points.

The first discrete analyzer output in the above list goes off when:
The Oxygen level is within the set points.

The Second discrete analyzer output in the above list comes on when:
The Combustables level is above or below the set points.

The Second discrete analyzer output in the above list goes off when:
The Combustables level is within the set points

- - - - - - -

=================

Under normal operation, the following sequence occurs:

Step 1 The boiler is online and the analyzers are monitoring the flue gas. If all is well the PLC will be monitoring the Oxygen and Combustables levels through a 4-20 mA signal being generated in the Analyzer. The analyzer watchdog timer alarms will be internally monitoring is own Fault and measuring status'.

Step 2 - What events should the PLC cause to happen in the real world:
Now let's say that ONE of the analyzers Oxygen/Combustables level go's high or low. This analyzer will close either the "Oxygen out of limits" or the "Combustables out of limits contact". The PLC will see this and iniate an alarm (Oxygen/Combustables out of range). This is OK the boiler will continue to remain online. If we should get either one of the other TWO analyzers to send another "Oxygen/Combutables out of limits" alarm to the PLC then the PLC will need to trip the boiler (MFT).

Of course if all THREE analyzers trip at the same time the PLC will iniate a MFT and trip the boiler.

The first problem is how do I program the PLC to make the determination that at least TWO analyzers have gone into alarm. The PLC will need to determin which analyzers are alarmed. Then the PLC will need to iniate a MFT.

The second problem is if I should need to take one of the analyzers offline for maintanance via. of a remote / local switch. How do I program the PLC to determin that if any one of the analyzers are kicked over to local control that the PLC will now only monitor the two remaining analyzers and that both of them will need to be in alarm before tripping the MFT.


Jim S.
 
Three Parameters Controlling a Single output

godfrey:

Thank you this is the what I need. I think your ladder for the any TWO alarms trips the MFT will work. BUT now I need how to figure out how to put any one analyzer into local control and have the PLC just monitor the other two for alarms. What do you think about building a ladder for each local switch closure.

Jim S.
 
Godfrey's answer is the one you were looking for, but the answer given by Allen is actualy more valuable. If you follow the procedure he identifies clearly and consistently your programs will be better and your job easier.

You tried to answer him, but you gave very contradictory information. You said "There are two each descrete input" but then you said "Now there are also TWO output contacts on the analyzer going to the PLC" which imples additional inputs!

Failure to clearly and concisely communicate is a common failure with engineers - we tend to be much better at number crunching than word smithing. Unfortunately, in my experience many more engineering problems result from poor communications than from poor calculation!

Here is a suggestion. Put the information in tabular form, and use non-interpretive descriptions. An example:

Input
Ch Name Type Voltage Action/Units
I:0.0/0 Anal. 1 NO Dry 24 VDC Close on Fault
Fault contact

I:0.0/1 Anal. 1 NO Dry 24 VDC Close on Normal
Measuring contact Operation

I:0.0/2 Anal. 1 NC Dry 24 VDC Open on Rising
Hi Alarm contact Oxygen Level

I:0.1 Anal. 1 Analog 4-20 mA 0-25% O2




This way there is no interpretation required.

Another tip - if I were doing this, I would use a register and compares. If an alarm contact is ON and the unit is online then add 1 to the register. After you do this for each, if the value in the register is greater than or equal to 2 then set your alarms.
 
Last edited:
Thanks Tom. I was just getting ready to answer and say what ou said, only not as well.

Just to clarify the nomencature. When I referrred to Inputs, I was referring ONLY to inputs into the PLC - things that the PLC knows about the outside world. By Outputs, I was referring ONLY to outputs from the PLC - ways the PLC has to manipulate the outside world. Yes, an input TO the PLC will be an output FROM another device. But since you are writing a program for the PLC, you need to look at the world from inside the PLC - use only what the PLC knows, and do only what the PLC can do.

Part of the exercise was to find out what the PLC knows. You are concerned about the analysers being in mainenance mode, but, from your descriptoin, the PLC has no way to know if they are in such a mode. Therefore, it can't make any decisions based on that mode.

But does it matter? As godfrey alluded to, if one analyser is off-line, you still have the requirement of 2 analysers agreeing to give you an alarm. You will only get into trouble if the off-line analyser is capable of generating an alarm. If it cannot, then the only way to get an alarm is by the two online analysers. If it can, then that signal must be stopped - either hardwired or by letting the PLC know that it is offline, and adding code to take it into account.
 
If I understand your question you want any two inputs to trigger the shut down. Take a look at the following ladder it will do what you want. By using your inputs to turn on the LSB of each interger you get a 1 stored in the integer when your input is on. Then you just add up the numbers then see if they are equal to or greater than 2. It does not matter what sensors are on line and 2 will give you the shutdown. Just dont have two off line....

alarm.jpg


John
 
OK I got a QUESTION

Godfrey stated his ladder would work if any 1 of the devices is offline...as the ladder is done and if devices connected in that matter...yes maybe so.

What I saw was an INPUT enabled from a device when there was a problem, does this mean if the device is disabled or removed then it cant alarm? IF you need an alarm for any 2 devices why would you remove one and be satisfied with an alarm of the other 2 failing...that alarm means the system is down?

What happens if the device has a physical problem then doesnt alarm because it doesnt send the input to the plc?

Personally this is a personal choice and one I made along time ago, I treat alarms similar to e-stops, ...ie they must be enabled ANY time there is a problem ...including device failure of the device that is sending the alarm signal.

Therefore I make their signal ACTIVE at all times and monitor the LOSS of the active signal.

For maintenance purposes there is always a way to setup a bypass.. simple spst will do that.

If the devices require maintenance that often then its best to keep a couple of spares to do a swap out with at pre-set times. Maintain the spares at optimal condition.
 
We think alike!...

I completely agree with you Ron and have followed the same philosophy for years. I apply this to ALL devices, not just "alarm" stuff...

Let's say you're using a photoeye to detect that there are enough parts on an infeed track to allow a machine to run. The parts break the beam when parts are present. If you use the "light-on" output to signal the PLC, then the LACK of signal (input OFF) means that there are enough parts to run. This will work just fine, but what if a wire breaks, or some bozo unplugs the eye? No signal at the PLC, therefore it's okay to run, right? WRONG!...

The correct (or at least BETTER) solution would be to use the "dark-operate" signal from the eye. Then it will give an output to the PLC when the beam is BLOCKED. The lack of signal (input OFF) means that:

1.) There are not enough parts to run (the intended function)
2.) A wire is broken
3.) Some bozo unplugged the eye

or a million other reasons... The point being is that the choice of using INPUT ON or INPUT OFF is important!

beerchug

-Eric
 

Similar Topics

Hello Guys, I have tried the programming but can't twist my mind to make it work correctly. I'm using S71200 and TIA Portal. My system has...
Replies
1
Views
182
Kindly, we connected electrical power, 3x380 vac, between a supply station and an electric cabinet, away 180 meters. The calculation software...
Replies
11
Views
760
im new at Allen bradley rslogix programming; i have something to ask you guys , and i will be happy to get answers so thank you in advance. when...
Replies
1
Views
572
Hi all, We had three drives fail upon power up. The two had F081 fault code. The maintenance personnel attempted to copy VFD parameters using a...
Replies
4
Views
1,466
Monday, we had a VFD that controls the spped of a conveyor belt fail on us. Took the replacement out of the box, installed it and bam, it failed...
Replies
37
Views
16,624
Back
Top Bottom