What's My Electrical Switch Doing?

[Muckin_Slusher]

2 wire non-contact prox switch.

We use these quite a bit, and occasionally get this strangeness.

Pepperl&Fuchs brand.

It's supposed to change state when it detects steel. The "contacts" are supposed to be open when no steel is in front of the target, and the "contacts" are supposed to be closed when steel is in front of the target. The green light simply indicates power. The yellow light illuminates when the "contacts" close. I put contacts in quotes, because this doesn't really have dry contacts since it's a 2 wire powered switch somehow.

We shut power off to this piece of equipment, then once power restored, the switch decides to operate backwards.

Watch the video. You will see me plug in the switch and pass a screwdriver in front of the target, then unplug the switch and plug it in again. The second time the switch is operating opposite how it works the first time. The first time I activate the switch, it's working backwards. The second time is how it's supposed to work.

This system works on 120VAC. This condition could not be rectified by changing out the switch (replacements all acted the same way). Situation was rectified by changing out the hardwired socket.

Explain to me why this is happening because I'm at a loss.

This is how it's wired. It's a 2 wire switch. It can somehow operate with a low enough current to not give input to the PLC.

 

[Muckin_Slusher]

Also note that these switches are used in critical applications.

I sure ****ing wouldn't want this thing on the landing gear of my commercial airliner, telling me the wheels are down when they aren't.

 

[ThePanzerFuhrer]

I always blame a bad ground on this type of problem . Im curious how this works only being 2 wire. For the light to come on it needs to have current. Can’t have that if it’s in the open state. Interesting

 

[AgitatedPancake]

https://files.pepperl-fuchs.com/webcat/navi/productInfo/pds/248047_eng.pdf?v=20241216200401

Looks like that switch can be used as normally open by using wire locations 3 and 4, or normally closed with wire locations 1 and 2, is there any way the unused wire locations are somehow getting random intermittent power?

 

[larboc]

Fuched

 

[Dethmachinefab]

Battery in the plc dead? Program got funky with the loss of power id guess.

 

[grumpy356]

Battery in the plc dead? Program got funky with the loss of power id guess.

The way I understand this , this is a switch problem. It doesn’t have anything to do with the PLC. He’s looking at the indicator of the switch, not the input on the PLC.



He is having this problem independent of the PLC.

I’d reach out to the switch manufacturer.

If it was a DC input, I would be looking at syncing and sourcing. But 120 l VAC…:. Nah.

 

[M92PV4U]

I assume that it has to do with installing and removing the switch with power to it.

 

[Muckin_Slusher]

I assume that it has to do with installing and removing the switch with power to it.

I think that would be exactly what happens every time there is a power outage, or the breaker is cycled. These switches are used on a critical application keeping very expensive stuff from crashing.

I always blame a bad ground on this type of problem . Im curious how this works only being 2 wire. For the light to come on it needs to have current. Can’t have that if it’s in the open state. Interesting

Yeah. I'm pretty sure in the past I've played with these, and they work even if the ground isn't connected, so whatever current the thing needs to work, has to pass into the PLC or relay coil it's operating. The current level must be below what the PLC considers an input, and below what it takes to pull in a relay. The power for an LED is miniscule. So small that we have a lot of trouble with running LED indicators off PLCs. Even when "off" the damn things illuminate, so we either need to add a relay, or something else. I've tried some resistors without luck.

https://files.pepperl-fuchs.com/webcat/navi/productInfo/pds/248047_eng.pdf?v=20241216200401

Looks like that switch can be used as normally open by using wire locations 3 and 4, or normally closed with wire locations 1 and 2, is there any way the unused wire locations are somehow getting random intermittent power?

I guess that's possible, but hard to imagine. The socket had some minor corrosion, but was dry. I plugged that thing in and out a bunch of times, and it was about 50/50 which way it would work. Also, tried with a bunch of new switches, and all behaved the same. Changed out the socket and worked normally, so must have something to do with it.

Maybe next time it happens I can chop off the pins for the N/C operation and see if that helps. If it does, I'll make that standard procedure with these ****ing things.

 

[M92PV4U]

I think that would be exactly what happens every time there is a power outage, or the breaker is cycled. These switches are used on a critical application keeping very expensive stuff from crashing.

It's easy enough to test. I can't remember what, but plugging something in or connection something while energized was giving me problems in the past.

 

[78bronco460]

Late to the topic, so ignore this if it’s not helpful: 2 wire prox switches suck for PLC inputs unless they are type 2 d.i. cards. It’s an impedance/current thing. Best to go to 3 wire prox switches and end the tail chases.

 

[TanTJJim]

There's likely a fair amount of electrical components in that switch. I've seen an issue in the past with a particular integrated circuit, which has a known issue documented in the datasheet from the supplier of the component, where the output of the part can be in the wrong state if the input changes state too quickly. I suspect something similar might be going on here.

The problem likely isn't because power is being turned on or off, the problem is likely that power to the switch is not stable. You'll get this exact condition when "hot swapping" the part as seen in your video. Knowing that the problem was resolved by changing the socket indicates an intermittent connection at the socket probably caused unstable power to the switch which just exposed a possible weakness of the design of the switch. Hope that all makes sense.

 

[Muckin_Slusher]

Late to the topic, so ignore this if it’s not helpful: 2 wire prox switches suck for PLC inputs unless they are type 2 d.i. cards. It’s an impedance/current thing. Best to go to 3 wire prox switches and end the tail chases.

That's a pretty big deal, since this is an engineered machine that hoists men.

There's likely a fair amount of electrical components in that switch. I've seen an issue in the past with a particular integrated circuit, which has a known issue documented in the datasheet from the supplier of the component, where the output of the part can be in the wrong state if the input changes state too quickly. I suspect something similar might be going on here.

The problem likely isn't because power is being turned on or off, the problem is likely that power to the switch is not stable. You'll get this exact condition when "hot swapping" the part as seen in your video. Knowing that the problem was resolved by changing the socket indicates an intermittent connection at the socket probably caused unstable power to the switch which just exposed a possible weakness of the design of the switch. Hope that all makes sense.

I would hope that a big time mining hoist engineering firm would know about that.

 

[78bronco460]

That's a pretty big deal, since this is an engineered machine that hoists men.



I would hope that a big time mining hoist engineering firm would know about that.

Elevator work is for elevator technicians. I’m not that. Good luck.

 

[IGT]

Will the switch still change states on the shelf with no power and monitoring the contacts with your fluke?

 

[Muckin_Slusher]

Will the switch still change states on the shelf with no power and monitoring the contacts with your fluke?

I don't think I've ever tried that. My guess is no.

 

[IGT]

I don't think I've ever tried that. My guess is no.

Do you have an instructions manual on the switch?

First thing that would happen if I called tech support and asked about this switch our folks would ask “did you read the ****in manual”? 😁

 

[TanTJJim]

I would hope that a big time mining hoist engineering firm would know about that.

I wouldn't hold my breath on that one. The example I shared is what we call a corner case as it may be outside or close to the edge of how this thing was meant to operate.

If the design engineer didn't design this to be hot swappable, they'd never ensure the design was robust against this condition. Even if they were, no design engineer ever would ensure the design was robust enough to function correctly if the socket it's attached to has a poor connection. The assumption at that point we be that the socket needs to be replaced to ensure there's a reliable power supply to the switch.


For reference, the design projects I've been around take over 5 years, 2-3 of that is typically just qualification to verify the thing does what it's supposed to do. Even with this level of scrutiny, things like this can (and do) slip through the cracks.

I suspect the switches you're working with took 1-2 months to design. Depending on the safety criticality of the part, maybe a couple months of certification? And was likely designed by engineers who aren't familiar at all with the environment it'll be used in.

All this is to say it's more likely than not there's a potential design issue with this switch itself that's the cause of the behavior you're seeing. Especially since you've seen it on multiple switches right? Only common denominator was the socket?