PDA

View Full Version : Calling Electrical Gurus And Practitioners Of PLC Magics...



Liger Zero
07-07-2010, 05:45 PM
This is a post in several parts.

First is a description of the problem:

"The tester started the morning fine, we ran a bunch of parts. Then it stopped functioning... the motor turned off and the machine refused to cycle.

We powered it down, let it sit figuring it was too hot in here. Turned it back on and got a couple of cycles out of it before it quit again.

We opened up the back to check the PLC, the POWER GOOD light is on but the normal PLC blinkenlightz are not blinken nor are they light'n. We can get the motor to cycle if we hold the breaker button in the power cabinet, but as soon as we let go it trips again.

We don't have anyone to troubleshoot this, so we are asking you via Brian ("Liger Zero")."


---

Liger talking now:

The machine is a test rig that pull on a finished product... 20,000 pounds delivered by a hydro-cylinder, other end is a load cell.

All this is coordinated via a PLC, several position sensors, a bank of switches and an activation trigger.

I was off today so I didn't see what happened, but based on the symptoms I'm leaning twords a cooked power-supply that erased/smoked the PLC.


Attached are some pictures. Tomorrow I'll attempt to reload the PLC from my laptop but I want to eliminate other issues first.

Liger Zero
07-07-2010, 06:01 PM
MODICON MICRO
110 CPU 311 03
http://img.photobucket.com/albums/v469/Lizoid/DSC00013.jpg

Allen Bradly Automatic Switch
http://img.photobucket.com/albums/v469/Lizoid/DSC00015.jpg

Controls For Automatic Switch
http://img.photobucket.com/albums/v469/Lizoid/DSC00017.jpg

Power Supply For PLC
http://img.photobucket.com/albums/v469/Lizoid/DSC00014.jpg

Liger Zero
07-07-2010, 06:11 PM
Ok Liger talking again:

I had this issue on another test rig at another job... Those silver thingies with three legs on the top of the power supply were at fault. I had similar symptoms in that case.

On this particular machine, the Automatic Switch failed recently, it was Quite Burnt. There is a possibility that the switch is failed again, but if I recall the PLC stayed lit.

I'll know more when I hook the PC up to the PLC tomorrow, just looking for advice and other possible causes.

john hawkins
07-07-2010, 06:34 PM
Let's start with this. A plc does not usually loose it's mind. First rule. no input no output.. Are any of the inputs led's on? Does the power supply have an output? Most plc's are powered with 24 V DC. Most machine builders use that 24 V DC to provide power to the various sensors, limit switches, &c As the unit resumed working after a cool down, I would be very suspect of the power supply. The 3 leg thingies are usually pass transistors. they are sourced <> 30 volts then other circuits adjust the output voltage to the required 24 volts. If you have a volt ohm meter with diode test it should be easy to t'shoot the electrics. By the way the panel builder didn't do a very neat job of layout/wiring. Does the machine have an E stop? Guarding interlock? HTH John from Dayton

Liger Zero
07-07-2010, 07:03 PM
There are two position sensors. Both position sensors are lit when I power on the machine.

The diagnostic for the load-cell shows no faults there.

The switch that acts as the trigger is also lit.

The PLC however only has ONE light lit and that is "power ok" everything else is dark, and the hydro-pack motor won't start.

In normal operation the PLC is a riot of blinking lights, and the hydro motor makes it's normal "motor sound."

To start the machine, you pull out the Red Button and the PLC lights up and the motor kicks in when it gets the signal from the PLC to start.

Today, you pull the Red Button and Nothing Happens. At all. No motor no PLC light no pull in of the Automatic Switch.


Gut tells me it's the power supply and the PLC is dormant, nothing more. I hope the heck it's not a "dead" PLC as this would be a "major hardship" for the customer.

Weston Bye
07-07-2010, 07:59 PM
How old is that PLC? john hawkins is right that PLC's don't usually loose their mind. But it has been known to happen. Usual cause is the end of life of the memory battery if the PLC is so equipped. Next is a power surge - scrambles the program memory - not generally likely with the PLC insulated from the lines by a good DC power supply - but again, it does happen. Then, there is just plain old natural death. Rare, but it still happens. I've seen all three in 35 years in the trade.

Hopefully, it is a recoverable failure and there is a backup disk of the program. You do have a backup disk, don't you?

Liger Zero
07-07-2010, 08:01 PM
Hopefully, it is a recoverable failure and there is a backup disk of the program. You do have a backup disk, don't you?

FIRST thing I do when I get a PLC driven fixture or begin work on anyone's PLC equipment: I backup the program myself. Have quite the assortment of interconnects AND software thanks to my step-father... who is out of town at the moment. ;)

kf2qd
07-07-2010, 09:13 PM
You pull the red button and the PLC starts flashing. Possibly no power to the I/O rack. Meaning you have lost a fuse to the I/O power supply, or you have lost the I/O power supply itself. You also might have lost the Power OK input and the PLC is not executing its program because of that.

First thing to look at is power supply and proper voltages. Can you disconnect the output of the I/O power supply and see if it works?

Could also be a bad Master Control Relay not pulling in and nothing powering up.


Got any schematics? It is always a lot easier to troubleshoot with the schematic because without it no-one is going to have much clue what is supposed to be happening on the outside of the PLC. Once you have that you can then try to find out what inputs should be active and also see if they have any power.

john hawkins
07-07-2010, 09:20 PM
The red button is the E stop. It seems from your description the E stop interupts power to the PLC which is common practice. I would jump out the E stop button. This is a normmaly closed switch. When you press the E stop it removes power to the PLC. The machine will do nothing in this state. Remove one wire from the E stop and put BOTH wires under one screw. This will jump the E stop, and maybe power the PLC. E stops are noted for failure, as any dirt will hold the contacts open. at least in the failed state the machine will remain stoped. Be sure that you lock out tag out the main power before you do any of this. Also have a helper stand by to turn the power off, as there will be now way to stop the machine when you test the results.

Liger Zero
07-07-2010, 09:21 PM
I will recheck the fuses tomorrow, my meter was giving me "funny" readings because... well I dropped it. :o:rolleyes:

...that's why I called it a day and came home and logged on here... after buying a new one.

john hawkins
07-07-2010, 09:27 PM
I should have said the E stop removes power to all input devices and also to the output relays in the PLC. The PLC is still thinking. But has no input therefore no output. If it were to output there is no power to output.

Liger Zero
07-07-2010, 09:40 PM
Tomorrow I will do the following:

1) Pull all the fuses and check them with the meter. Eliminate the obvious. :)

2) Check the voltages at the power supply.

3) Check the relays, there are six of them that I know of.

4) Put the PC on it and see what the software has to say about the PLC.

I got off the phone with my Step Father his opinion is the PLC is either DOA or not getting the right kind of power, that it is getting SOME power but not enough to "engage..." sort of a low-voltage condition. He says the POWER OK LIGHT is nothing more than a pilot light that comes on when voltage is present and has little to no diagnostic function on this unit.

J Tiers
07-07-2010, 10:36 PM
I'm not seeing any attention being given to the issue of the breaker opening...... but the folks with the problem say they can't get it to stay in.

To me, that is the immediate issue.

See what it supplies power to, and disconnect them one at a time until the breaker does NOT open..... at which point you MAY have found the problem.

And make sure that you are not fooled by taking power off something that merely activates the bad item....... if the contactor for the shorted motor never closes, you don't get an open breaker..... but you have not disconnected the real problem yet.

Once you KNOW what is causing the breaker to open, you can evaluate what otehr things may have happened.

First rule of complex systems is NOT TO GUESS.

Second rule is to FIX WHAT YOU KNOW IS WRONG

Liger Zero
07-07-2010, 10:51 PM
Good advice.

This is the startup sequence:

Pull the red button. PLC comes on, goes through a diagnostic routine then it signals the automatic switch and the motor comes on.

Problem is no sign of life from the PLC. PLC isn't waking up therefore the motor isn't starting.

Pushing the black thing on the automatic switch does no good as the PLC isn't awake and talking to the rest of the machine.

As far as I have been told that auto-switch is JUST for the motor circuit.

However, I will check it tomorrow.

MaxHeadRoom
07-07-2010, 11:20 PM
Used many types of PLC's, but not Modicon, but I would say if the power light is on but the run light is not, it is not going to be external power supplies.
The program sounds like it is in Halt, whether the program is still there is another matter.
Max.

RobbieKnobbie
07-07-2010, 11:32 PM
If you're pushing in the motor starter and the hydraulic pump motor starts - then you have power availableto that circuit, you're just not getting a signal to fire it up. WIthout a signal, of course it will drop right back out again.

Do I read you correctly that you're not seeing ANY of the input or output indicator LEDs come on?
The input LED's should come on as the switches/sensors trigger regardless of the program - or lack thereof.

The fact that you're getting NO input or output LEDs to come on tells me that you either aren't getting enought power to the PLC (easy enough to check) or there's an internal problem with the I/O section of the PLC.

Regardless of the problem, assuming you have the capital availabl I would order a new PLC, load it up with a copy of the program from this one and keep it on the shelf for exactly this sort of emergency.

Liger Zero
07-08-2010, 12:13 AM
Capital is the issue, always is with this kind of company.

If I play this right I should be able to get paid for my hours I put in on this and other projects (he's been good about that) and eat the cost of the PLC and whatever is the cost of the power-supply. If it goes beyond that, he'll have to pay for whatever.

rdfeil
07-08-2010, 01:43 AM
Liger,

One thing that strikes me as suspicious is the fact that they were able to "power it down and restart it" then got a few cycles out before it stopped again. This stinks like a power supply or a switch problem NOT a PLC program problem. It might be a PLC failure but not a program problem, as you know once the program goes away it does not come back on its own :rolleyes: . As was mentioned above, on every PLC I have used the input lights would light based on the inputs regardless of the program in the PLC. If you are not getting any input lights look for a power problem at the inputs. This could be the power supply or fuse or simply the E-Stop (red) button or some other safety switch that feeds the control circuits.
Keep us posted with what you find. This is what I do for a living so I am always interested in what failures other people run into.

Robin

boslab
07-08-2010, 02:02 AM
24 volt failure, check main supply, 110 live to psu primary leg of transformer, spike secondary with meter set to AC, ie before diode/rectifier, or better yet wire in a 24 volt psu direct to the evil modicon, IO strip should light.
Emergency stops over here dont power down a plc, they just stop prime movers from well, moving, it would be pointless pulling power off a plc to stop a machine as there could easily be a suspended load hovering at the whim of a now uncontrolled valve, EM stop function is normally on the second rung of the plc prog [M1 latch link Normally closed to allow the prog to run]
No IO lights then check Digital IO terminals for loose +24 v none found then very odd as the proximity switches/limit switches/reed switches should all still work and light up the IO strip even with a dead modicon [dead meaning plc cooked] look for a loose Neutral or Zero volt line also
good luck
mark
[thank the lord its not a siemens!, i hate s5, s7 and any other one the nutters come out with]

Weston Bye
07-08-2010, 06:05 AM
If you get the POWER OK light and nothing else, the problem is probably the PLC. With a program present and in run mode, there should be lights - even with nothing but power connected to the PLC. None of the rest of the circuitry should matter or have any effect on the status lights (the ones at the right) The input lights *might* function, but without the RUN light the output lights should all be off.

First step should be to connect the programming device and establish that the PLC is communicative. If so, a restart may work, or a reload of the program to memory. Once you get your lights back, then address the rest of the circuitry.

(cue Scarecrow , singing "if I only had a brain...")

[QUOTE=Liger Zero]http://img.photobucket.com/albums/v469/Lizoid/DSC00013.jpgQUOTE]

Liger Zero
07-08-2010, 07:27 AM
So I go in this morning. Pull the red button to apply power to the machine.

PLC lights up, motor goes VOOM! :eek:

I shut it off, and grab my IR thermometer. Those of you who have experience with troubleshooting may skip directly to my victory dance below. :D

I take readings of several thingies in the power supply. The silver thingys, the blue thingy, the black thingy with the screw and the microchip.

Those went on the pad by my hand. I turn the machine back on...

PLC lights up, motor goes VOOM. I put parts in the tester and start cycling it.

Took about an hour for the problem to manifest. I traced it to a crack in the board that opens up when the board expands... I determined this by temp readings... everything else was HOT one group of components wasn't. Figured it was a blown solder joint but it wasn't, it's a crack.

Found the crack, it is quite obvious but would normally be hard to see with the power supply mounted. Crack opens up it drops a group of components out consisting of silver things and a couple of diode-ish things and a small blue thing. Those components if I follow the trace right, have something to do with the wire that feeds into the power terminal on the brain--er PLC.

A temporary repair consisting of epoxy and a gob of soldier and the machine is working along just fine now. I took pictures of both sides of the board, someone who owes me a favor is building us a new one, should be ready Tuesday. ...actually I sent him pictures, a diagram and the voltage at the terminal readings so he can duplicate a power supply for us. ;)



Now excuse me while I take the REST OF THE DAY OFF to do my epic victory dance. :cool: That's right, DAY OFF. Thank you for the suggestions everyone, they are on file now (hardcopy) in case I have any further PLC issues down the road.

Weston Bye
07-08-2010, 08:44 AM
... Then, there is just plain old natural death...

This qualifies. But you cheated death and resurrected Lazarus to live again.

Good work!

john hawkins
07-08-2010, 12:05 PM
Sola/hevi duty makes a nice DIN rail mounted power supply <> 300$

Liger Zero
07-08-2010, 01:58 PM
Sola/hevi duty makes a nice DIN rail mounted power supply <> 300$

Thank you but $25 and a case of Bud is all I'm paying :D

boslab
07-08-2010, 04:07 PM
well done, congrats, its nice when you beat it, next, why did it crack? vibe, thermal or idiot?
mark

kf2qd
07-08-2010, 05:50 PM
Solder a piece of wire across the crack. A glob of solder will hold for a while, but a piece of wire soldered to that trace will outlast every other component on the machine...

Liger Zero
07-08-2010, 07:10 PM
Machine has been cycling 350 times a day for over a decade. When a part fails the test it creates an enormous shock, the entire machine jerks.

Add to that the shop floor is 110 degrees in the summer, 35 degrees in the winter and I am the first repair-man to have to do anything more serious than a minor electrical repair.

It is a very solidly built tester, considering. :)

New power-supply will be ready shortly.