Sticky Relays

[BillRobinson]

I've got a 16 Allen-Bradley 700-TBR24 relays that are "sticking". 24 VDC from a PLC card drives the coil, and 120VAC is on the NO contacts drives a .4A strobe light. At times when the PLC digital output is turned off, the relay indicator goes off, but the 120VAC light remains on.

Removing the 24VDC power source from the card/coil does nothing (the light remains on), and resetting the the 120VAC does noting (the light comes back on once 120VAC is restored). The only way to turn the 120VAC light off is to pull the relay from the terminal block. When replacing the relay back into the terminal block the light stays off (as it should).

This phenomena is seen all 16 relays at time to time, so it's not just one bad relay. What is going on here?

 

[yam785]

This is going to sound really stupid, and everyone will probably think I'm a nut, but I've had issues with strobe lights before. Leakage would make them charge up over time, and flash the strobe, even though the output was off.

I cured the problem by putting a load across the strobe. It was a long time ago, and i foget exactly what I used. Maybe it was a MOV, not sure.

A sure fire way to be sure, would be to disonnect the strobe from the relay contacts, and do an ohms check through it, while toggling the output on and off. If not sticky with ohmeter, figure out the strobe issue.

For you to have the same problem with all of them, and I'm figuring you changed out the card, I just have to think it's a strobe issue.

 

[Ken Roach]

My first guess was leakage from the DC output module, because those are very small relays.

But obviously if you have removed 24V DC power from the output module, there's no way that the relay coil could still be energized.

This genuinely sounds like contact welding to me. Try tapping the relay while it's still in place to see if you can get the contacts to break, rather than removing it from the terminal block.

Post more about this strobe light; I'm not familiar with the inrush or pulse current characteristics of strobe driver circuits.

 

[OkiePC]

Isn't 0.4 amps pretty close to the contact rating of those relays, and driving a strobe, I bet-cha' the peaks are above a 1/2amp. I agree with Ken, that welded contacts are probable.

I am gonna have to look it up, but I suspect you're up against the current limit, At the very least you can expect a reduced MTBF for running at the high end of the contact rating.

What is the duty cycle of the relay?

EDIT: I worked twelve hours today and my units were off by a factor of >10, so I wonder now, if its not magnetizing due to the pulsing of the current and might be better suited for the solid state variety of relay:

700-TBS24

 

[jrwb4gbm]

Here's a link to the relay specifications, maybe some help there.
Link:
http://www.ab.com/en/epub/catalogs/12768/229240/229266/229643/229703/tab5.html

Maybe there is some stray voltage pickup (noise). Try putting a shunt resistor across the relay coil to bleed any stray voltage off.

 

[Gil47]

Just adding a thought here , could you make sure that your output card has a Neutral connected to the card, normally on the bottom terminal of the card on Allen Bradley PLC's

 

[leitmotif]

Something is fishy here.

1. Are you sure you turned off ALL sources of 24 VDC
2. What are the design of PLC output contacts mechanical or solid state?
Removing the relay assures me there is no voltage to the relay. Plugging it back in and having circuit operate OK sure makes me think bleeding OR multiple power source and you just did not get them all turned off.

IF the contacts were welded then I would think they would stay welded.
IF the contact were welded shut then strobe would activate assuming 120 is energized of course.
The next time this acts up put a resistor across the relay coil terminals and see what happens. If bleed thru solid state it should turn off.

Dan Bentler

 

[iant]

Relays dont last forever - Remember the greater damage is caused to the contact when it opens. (Back EMF)
Howmany times would one relay output have switched since you wired it.
if it is greater that 100K I would be recomending they be changed.
With an AC output go to solid state - or higher current relays.
If they are PCB type then replace the card and make an assessment as to when you will change it again - costing etc.

 

[theDave2]

Use a DMM on the coil when this happens and see the voltage, both A/C and D/C. Look across the coil and also from both sides to the P/S hot and also Return and Ground. Only then will you know what you are dealing with. Note the voltage polarity of each measurement.

You have either induced voltage or a loose DC Return wire that allowing reverse feeding of the relay.

 

[BillRobinson]

Notes
- Tapping the relay during a sticky situation, releases the contacts.
- Contacts are new <100 makes/breaks.
- Sticks only occur on relays that contain strobes.

Also isn't using solid state relays a no-no when dealing with lights? Mainly do the leakage.

 

[theDave2]

Please tell the PLC output card model. And strobe.

 

[iant]

Notes
- Tapping the relay during a sticky situation, releases the contacts.
- Contacts are new <100 makes/breaks.
- Sticks only occur on relays that contain strobes.

Also isn't using solid state relays a no-no when dealing with lights? Mainly do the leakage.

Hi Bill
the fact that the contacts are new is a big concern.
1. Tapping the relay - It must be sticking (Welded)- Definately not coil voltage issue
2. with <100 operations, the inductive load must be extreamly high - 0.4 Amp switching current, bugger all impeadance load
3. What type of light is it (Incandescent , LED, Xenon)- R U strobing via the contact or is a Flashing Lamp.

I am suspicious about the relay condition and recommend replacing it / them. - they are already past repairing.

- Try switching half wave - Add a diode to the relay output.
and another diode to the light- not a good practice but worth a test.

Solid state leakage may be an issue - generaly with low loads like Neon's.

The only alternative is to switch an auxillary relay coil and use higher rated contacts - Omron GV2 or similar.
Get that load away from the PLC output card completely.

 

[nwboson]

Lots of great ideas here. Why keep guessing? Simply put a fast digital oscilloscope on the output contact and watch what happens - especially when the contacts shut off.

If the problem is significant enough to cause welding it should be immediately obvious on the oscillocope: presumably a very large negative voltage spike when the contacts open. Once you've verified that is occurring, you can try the half wave diode trick mentioned above and add a reverse clamping diode like you would on a 24V DC output: that will clamp any (V = L di/dt) effects.

 

[iant]

hi nwboson,
Sorry dont know your name,
That is the best Idea for conformation of cause.
Most of the maintenance people don't have the equipment available.
Unfortunately these relays are sealed - But - if Bill has one than you are right

 

[widelto]

I had the same problems with this relays several years ago, since then I never used them no more. In my case any electrical noise triggered them.