Tired relay?

We are currently experiencing a problem with one our control systems, and I thing the problem might be a few relays going bad, but still functioning everytime I check them. So can anyone shed any light on a "tired" relay. Will an older relay drop out when it shouldn't and then next time it's fired work okay?
Already changed out CPU, input, and output cards etc...

We have a gremlin we can't locate, and any help will be appreciated.

Basically we a have a primary and secondary pump system, where one pump runs all the time and the secondary starts if the flow or pressure drop to low, or if the primary stops running for any reason. Sometimes the primary just stops for no apparent reason, and then when you restart it everything is okay. The plc output if firing a 24VDC relay, that pulls in the control circuit for the motor starter. My next guess is the motor starter itself. All the equipment in this system is about 13-14 years old.

Most coils require about 85% voltage to pick up, but will not drop out until voltage drops to about 60%. So a "tired" relay coil is not likely.

Are the relay contacts worn? Poor contacts could overheat and cause increased resistance thereby lowering the voltage to the coil. After the contacts are turned off and cool down everything tests okay.

Ken,
The only time I'v seen this was one time some bonehead plugged a 24volt relay into a socket wired for 12 volts. The relays look identical except for the printing on the side. I doubt this is your problen if it has been working for that long. Has anyone changed the relays lately? I would suspect a loose wire somewhere. A loose ground wire can be hard to find sometimes. Intermitent problems can be hard to find. Most times, relays either work or they don't. BD

I must agree, I've never heard of tired relays in this sense. The only thing similar I've met was with genuine magnetic core memory in the '70's. We had a case where some memory addresses were not accessed for over a year and in this time the cores slowly lost their magnetism. We finished up having to modify the program so that all addresses were occasionally read so as to refresh the cores. But I've never heard of anything similar for relays.

I have seen a couple of instances where the actual relay socket was bad. Both have been on older machines and just started erratic behavior like you are seeing. These were on heavy loads and were changed often so we just assumed the sockets become loose internally.

Bob

I would look at the contacts rather than the coil.
Most contacts only have a limited life.

Other things to try:
- Loose terminals
- worn insulation

does the primary pump have a pressure switch on it? could be it cuts off if it reads too high or too low. one of those deals where it is right on the cut line.

Ken,
I have seen relays fail to energize because of low voltage, bad plug-in sockets, loose screw terminals, and a broken wire inside the relay. I remember a few cases when the relay was intermittant, sometimes it worked, sometimes it didn't. A broken wire inside the relay can cause this. I encourage the maintenance department to keep spares to do a quick swap to catch those problems.

I encountered a similar problem with two Lead-Lag 25 HP air compressors. It turned out to be a defective low-oil level switch on one of the compressors.

Ken Moore said:

one pump runs all the time and the secondary starts if the flow or pressure drop to low, or if the primary stops running for any reason. Sometimes the primary just stops for no apparent reason.

Click to expand...

So you're certain the second pump is starting because the first one stopped and not one of the other reasons you listed?

You could always go over the overload and contact block for the motor with a fine tooth comb. I had a siemens overload's aux. contact flake out on me once and the control power for the motors contact block ran through it.

I'm just tossing out ideas.

Everyone else covered the intermittent relay thing. I ran into a bad base a couple weeks back causing intermittent behaviour.

Sounds to me like you need to break out a screwdriver and a can of contact cleaner.

Good luck.

This may sound difficult but it shouldn't be that hard to implement, some of it may already be there. If you have a spare input feed the output for that relay back to it. If possible use secondary contacts of the relay and starter to feed to separate inputs to show they are on. Now use these inputs to give you an alarm if either the output, relay or starter drop out WHILE the program "output" is true. NOTE: Do not use the output power to feed the auxillary contacts on the relay or starter.

This should tell you if its "hardware" and which device...ie the plc output itself, the relay or the starter. If it happens and this "alarms" doesnt engage that means the "software output" is going false for some reason.

I'm pretty certain the relays are not the problem. We are changing them all out anyway, the cost is small and the cost of losing this pumping system is large. The most likely suspect is some type of mechanical failure. There has been no PLC programming changes, no new hardware, so we have to find out what is happening.

This system is used on a high pressure reactor vessel's double mechanical seal, the pumping system maintains the gylcol in the seal 500 psi above reactor pressure. If the pumps fail/fault with a lot of pressure on the reactor, the seals blow. The seals are several thousand dollars each, and the system has 4 of them. So right now the cost of fixing this is minor compared to the cost of repeated failures.

The failure usually happens when no ones watching,(doesn't it always) and we cannot reproduce it. So, we are trying anything and everything. We are replacing anything that might be the cause. This system is connected to a SCADA system that historizes the pumping system data, but the event happens too fast to be captured. After this last round of replacing "stuff", if the problem is still present, I'm going to program in a lot of traps into the logic and see if I can deduce what is really happening.

If possible use secondary contacts of the relay and starter to feed to separate inputs to show they are on. Now use these inputs to give you an alarm if either the output, relay or starter drop out WHILE the program "output" is true. NOTE: Do not use the output power to feed the auxillary contacts on the relay or starter.

Click to expand...

We are already picking up the aux. contacts on the starter, that is one of the conditions that starts the second pump, output on, but the starter has dropped out, but I will take a look at connecting the relay output to a plc input. I don' think I have any spare inputs, but it is a large system with several remote bases, and I can add a new input card to one of these.

I would also look at the analog signals that the logic uses to control the relays. Noise or partially plugged orifaces could be giving false signals.

I just isolate

The alarm setup I mentioned would tell which hardware device is "dropping" out first...ie output, relay or starter. If it shows the relay is at fault and you have consistently changed relays, then the base is probably bad. If the starter drops out it could be coil or contacts. I assume the DC output is transitor so its unlikely, but possilbe, that its at fault.

The other aspect is the "holding circuit" for the output, is their an input device (aux contact etc) that could be dropping out intermittently. I can not tell you how many times I have found dust/dirt in a relay/aux contact that caused problems.

This is a common problem, you just have to set a trap to catch the gremlin. Its something you should develop and keep in your tool kit, at times it may require using a hardwired relay ckt as an alarm.

Ugh.. has anyone mentioned the flow switch? I assume that has been checked, or logged at least.

Never seen an "old" relay drop out, but I have seen overloaded control transformers drop out relays.