PLC Basic troubleshooting guidelines

RedRaven,

This is probably a cop out on my part, but if you search this forum and the archives for the word 'troubleshooting' you will find lots of examples of how people troubleshoot on PLC's. Have a read through the posts and then maybe you can write your own basic troubleshooting guide.

Then, if you would still like advice, post what you have written and let the 'experts' have a look and make suggestions.

Paul
 
Get a clear, specific definition of the problem. A statement like "the machine isn't working" tells you nothing about the problem.

A statement like "the conveyor doesn't start when it's supposed to" is a little better. It focuses your attention on the part of the syatem that's acting abnormally.

A statement like "the conveyor doesn't start automatically, but it will run in manual mode" is better still. That tells you that the drive mechanism for the conveyor is functional, but there is something lacking in the automatic control system.

If the person descibing the problem to you doesn't offer information like that, then you need to know enough about how the process works to ask the questions that will narrow down the list of possible causes. Your best source of reliable information is an experienced operator.

If the machine has been operating properly for some period of time, and suddenly starts acting abnormally, the problem is probably not the PLC program. Unfortunately however, people have a tendency to lay the blame on the things they understand the least. Under those circumstances, resist the temptation to fix the problem by tinkering with the program.

In general, approach the problem logically. Consider all of the things that could be causing the abnormal operation, both mechanical and electrical. Things like low hydraulic or pneumatic pressure, reduced or restricted flow, sensors bumped out of alignment, slipping drive belts. Ask if there have been any recent changes in the size, shape, color, weight, or other properties of the product, or in the operation of the machine. If it was working OK yesterday, you're trying to find what's different today.
 
Steve's answer is right on. I always try to explain the procedure to rookies as following a chain link by link. The chain is an event sequence of cause and effects. Start with the desired effect that is not happening, or the undesired efect that is happening. That is the first link. Then look at the event that is supposed to cause that effect, and verify that it is working properly. Then, look at the event that is supposed to cause THIS second effect, and verify it. You woek your way down the links sttep by step until you find the cause of your problem in a link that isn't.

Don't shortcut, don't start in the middle, and don't start chaning parts unless you have verified the part isn't working.

Take good notes of every single step, including things like times, voltage levels, etc. etc.
 
Steve and Tom's answers can probably be summed up in one word.

ISOLATE, ISOLATE, ISOLATE!

(It can't hurt to repeat it a few times)... :p

One more thing I find a lot of people DON'T do...

If you adjust, move, change, etc. something and it doesn't make a difference, PUT IT BACK BEFORE CONTINUING!!!

Most of the time, it's only ONE thing that's causing the problem. If you adjust a bunch of stuff before discovering the REAL problem, the machine may still not work because of the other stuff you changed and failed to put back.

Sorry for the rant... This is one of my pet peeves... :D

beerchug

-Eric
 
Eric,

Your point ought to be a no-brainer, but it's astounding how many people don't bother to undo their tweaks when they don't turn out to improve things. It's not limited to adjustments on the machine. People will install software patches and fail to remove or even document them.

It's equally astounding how many people respond with "gee, I never thought about that..." when I make the same point you did.

It's high on my list of pet peeves too.
 
Steve Bailey said:
Eric,

Your point ought to be a no-brainer, but it's astounding how many people don't bother to undo their tweaks when they don't turn out to improve things. It's not limited to adjustments on the machine. People will install software patches and fail to remove or even document them.

Un-patch software??? What kind of utopia would we be living in if that were as easy as it sounds? I can't remember how many times I've applied a patch to fix a problem, only to find a NEW problem created by the patch.

I'm really glad I stayed out of computer science.
 
Aaron,

I wasn't really referring to things like service packs for commercial software. I was ranting about people who decide that even though they don't quite understand what the original author of the control program was intending, a perfectly valid troubleshooting approach is to jump in and change it. Then, when their changes don't improve things, they make more changes without removing that which didn't work.

Just as the road to hell is paved with good intentions, the road to spaghetti code is full of cul-de-sacs and blind alleys.
 
We provide a trouble shooting guideline for our customers that is available on our web site.

http://www.entertron.com/trouble.htm

This guideline could work for just about any PLC, excluding specific items that relate to Entertron's PLC operations.

If all else fails, they usually contact us. More times than not, it isn't the controller but one of the other devices.

We had a customer who we did a program for. Developed it...simulated it. Everything worked. He installed the controller and was having a problem with it returning to the home position. After many back and forth conversations, and the customer thinking the problem was the PLC, it turned out to be a lose wire on his drive that was causing the problem.

God Bless,

Stephen Luft
 
Far too complicated to get into but I can give you one good pointer.

The first step in your future troubleshooting is to know/understand the process, machine, program, device Etc.....(HOMEWORK)
You will need to know how it works before you can fix it.
Roger
 
Two things you simply MUST know before you succeed:::

How is the darn thing SUPPOSED to act

How the darn thing IS acting. (If possible, see for yourself--four people will describe the same event in four different ways)

Know these, determine the difference, and you've got a darned good chance.
 
good Wiggy? - bad Wiggy?

Yo, everyone,

I’m short on time so my traditional overkill approach isn’t possible today - but something in Vetteboy’s post #11 prompts me to point out some interesting facts about the type of tester he mentioned. These things are commonly called “Wiggins” or “Wiggy” testers.

... show 120v due to leakage current, even when an output is off, so now we carry Voltage and continuity testers in our pockets that have solenoids in them to draw down the leakage.

The “Wiggy” is often used - just as Vetteboy said - to “load” a triac-type output circuit to overcome the false “on” test results which can be caused by leakage current through the solid-state output circuit.

Now I KNOW that many technicians use these things day-in and day-out with no noticeable problems - so I’m not going to debate their use - but I am at least going to offer a little food for thought on the subject.

Here are a few simple questions to get us started. First question: “How many volts does your car battery provide?” Simple answer: “About 12.” Next question: “How many volts does it take to fire your spark plugs?” Answer: “At least 20,000 volts.” Final question: “Where do you think all of that HIGH voltage comes from?” Answer: “From the car’s ignition coil.”

And that’s the problem with the “solenoid-type” of tester that a lot of technicians use. It is essentially the same thing as the ignition coil in your car. All of this goes back to the same phenomenon which many technicians call the “fly-back” effect. In simplest terms, the magnetic force existing in a coil doesn't just simply dissipate (or “go away”) when the service power supply is removed. Instead, the lines of magnetic flux actually COLLAPSE back through the conductors (of the coil) and the resulting system suddenly meets the basic definition of a generator - one capable of generating enough voltage to fry the electronic components of any manufacturer's control system.

Now a lot of guys are thinking, “But I’ve used these things for years with no problems.” Well, maybe so - there are a few people who smoke while they gas up their cars - with no problems. Still ... you might just want to think about this.

So what are the highest acceptable voltages for your AC input and output modules? Most are in the range of 140 volts. Now how many volts could you possibly get out of one of these “solenoid-type” testers? A WHOLE LOT MORE than 140 volts!

And the problem could be extremely intermittent in nature. Consider that the 60Hz sine wave of the service power may be at ANY point in its cycle at the instant that the test probes are removed from the module’s terminals (which causes the field to collapse). It's a totally random thing. If the voltage is at, or near, the zero point in its cycle, you may not see a problem. On the other hand, if the voltage is up near its maximum, the “fly-back” voltage which is generated may be enough to qualify as a "lightning strike" type event. I’ve heard of one plant who (under new management) actually outlawed the “Wiggy” testers that their technicians swore by - and then saw a significantly lower number of output modules “go bad”. The consensus of opinion: “Testing the modules was actually damaging them.” - but the issues were still being debated the last I heard. Some technicians just refuse to believe that their faithful old “Wiggies” had a darker side to their nature and were actually burning out a module from time-to-time.

So, once again, I’m not entering into any debate - just offering some food for thought.

Finally, this “fly-back” effect was once traditionally used in the past to indoctrinate apprentices in electrical repair shops. The apprentice was told to use a Simpson meter to “ohm-out” a transformer coil. Now the young lad KNOWS that the only power in the circuit is the 1.5 volt battery in the meter. So he squeezes one meter probe against one primary lead from the transformer - and squeezes the other meter probe against the other primary lead. Notice that his fingers are in real good contact with what he considers to be a “safe” circuit. And while he’s taking a look at the meter reading, the shop foreman reaches down and pulls one of the test leads out of the meter. The field in the transformer winding collapses - and the apprentice gets a serious jolt to help him remember a valuable lesson. “I TOLD you to ALWAYS keep your fingers off those leads!” says the foreman.

Finally, no offense, Vetteboy - I can assure you that you are not alone in using these things - and I'm not going to try to talk you out of it. But still - just food for thought -
 
Wiggies

I will trust my wiggy any day. DVM's go stupid too many times.

If some weak sister card can't handle that small solinoid, how could it possibly handle a valve solinoid?

If your wiggies are blowing up cards it is time to get better cards

Steve D
 
Like Ron said, from a raw statistical perspective you may be safe using the wiggie. And I think it depends on how you pull the thing off of the test device as to whether it will cause a problem or not (I may be wrong in that one). But the fact that an AC output card can switch a valve solenoid all day long and not have a problem discounts the nature of the output card.
Ron already refered to this. All the AC outputs I have dealt with have used zero cross switching power devices for just the reason Ron talks about. By turning the output off at the zero point in the AC waveform you control the rate of collapse of the magnetic field and prevent the large inductive kickback voltage. So we would expect that the AC output would not need to worry about this case.

Keith
 

Similar Topics

As per the title, does anybody make a basic PLC that directly supports IO-Link without requiring a fieldbus and remote IO modules? I know it goes...
Replies
9
Views
1,588
Create ladder logic program to count 4 on pulses (not using a counter) and then turn the output on if the on pulse input signal can look like the...
Replies
105
Views
36,181
Hi Friends, I'm making a PLC operated chocolate vending machine, using a CLICK Basic PLC (C0-00DD2-D) which has 2 x RS232 ports. I'm also using...
Replies
0
Views
1,518
hi all, which AB PLC or module that runs on BASIC language? thanks
Replies
8
Views
3,534
Hi all, First time posting. I'm an new journeyman electrician from Canada trying to move to the industrial side of things. I have so far written...
Replies
14
Views
5,769
Back
Top Bottom