Relay dropout time affected by input module

I have a series 110v control circuit that has 40 limit switches connected in series which feed a control relay to allow the machine to run. These limit switchs are normally closed. I have connected an input at the load side of each limit sw to detect when the limit has been activated. By monitoring which limits have power passing through and which do not I am able to determine which limit has droped out causing the control relay to drop out stoping the machine. I have my inputs connected to 1769-IA16 compactlogix I/O cards. The problem is that with my inputs connected when a limit drops out, the control relay takes a sec or two to drop out now causing the machine to run longer than it should. I am not sure what I need to do, I would appreciate any help.
Thanks

What kind of relay takes a second or two to drop out?
You might try replacing it with a solid state relay.

mncarmack said:

I have a series 110v control circuit that has 40 limit switches connected in series which feed a control relay to allow the machine to run. These limit switchs are normally closed. I have connected an input at the load side of each limit sw to detect when the limit has been activated. By monitoring which limits have power passing through and which do not I am able to determine which limit has droped out causing the control relay to drop out stoping the machine. I have my inputs connected to 1769-IA16 compactlogix I/O cards. The problem is that with my inputs connected when a limit drops out, the control relay takes a sec or two to drop out now causing the machine to run longer than it should. I am not sure what I need to do, I would appreciate any help. Thanks

Click to expand...

the control relay takes a sec or two to drop out now causing the machine to run longer than it should.

Are you sure this is not normal machine coast down?
How long does it take to stop when you hit the stop button?
This control relay is not a time delay type is it and is it just a "regular" electromechanical relay ?

I would generously allow .25 to .5 sec for electromechanical relay to drop out.

It is not aged and sticky or has a bunch of dust and **** in it slowing it?

Dan Bentler

Are you implying that the action is different when the monitoring lines are connected to the limit switch nodes than when they are not connected?

If so when a switch opens does the voltage across the control relay drop to zero? If not, what does it drop to? How does this compare between when the monitoring lines are attached and when they are disconnected? If the AC low point on the AC input card the same point as the low side of the control relay?

mncarmack said:

I have a series 110v control circuit that has 40 limit switches connected in series which feed a control relay to allow the machine to run. These limit switchs are normally closed. I have connected an input at the load side of each limit sw to detect when the limit has been activated. By monitoring which limits have power passing through and which do not I am able to determine which limit has droped out causing the control relay to drop out stoping the machine. I have my inputs connected to 1769-IA16 compactlogix I/O cards. Thanks

Click to expand...

So the 40 NC limit switches in series control a relay which stops the machine.

The load side of each switch (say # 18) is connected to
1. the input of the next switch in series ie # 19
2. to an input on the PLC card.

I can see the thinking behind the monitoring - would make finding an open switch a snap and sorta agree.

BUT what you now have is a parallel circuit that may be feeding thru PLC when switch 18 is open and keeping relay energized albeit at a much lower voltage thus delaying opening.

I would consider removing all feeds from the switches to the PLC and control the relay from the switches.
OR
1. making each switch an independent input (you are already hooked up for that)
2. disconnect the series connections of the limit switches and parallel feed them with 120V
3. write logic such that all 40 switches must be shut to allow output #xxx to "energize" and energize the relay.

Finding an open switch should be no more difficult than what you do now ie look at the input lites.

Dan Bentler

Sounds like an e-stop system that is poorly designed. If this is an e-stop system, do not program this through the PLC or even use limit switches and control relays. Use approved force guide switches and a safety controller. These usually have aux outputs for monitoring.

If this is not an e-stop, you can still use a safety controller like the Banner SC-22 for inputs to control an output. I've used the Banner alot and it is so far my favorite for simple e-stop systems. 22 single ended inputs (11 redundant) - 10 aux outputs and 3 safety relays built in for about $500. Hard to beat.

Decades ago I built E-stop pushbutton series circuits like that, with 1771-IAD input modules. But the contacts on the E-Stop pushbuttons were separate from the e-stop series circuit itself.

It does sound like somehow the Compact Input module is "bleeding over" voltage through the last input circuit that is sustaining the relay. I'm not familiar enough with the insides of those AC input modules to figure out how that could happen.

Hmm.

As long as the common on the input modules is the same as the common of the control relay I don't see how a 'bleed over' would happen.

I have seen this type of effect though if the commons are unconnected, then you would have the filter circuitry of two inputs in series with each other by-passing around an open switch. That's why I asked about the stuff I did.

And I agree - if this is only for machine control and is not for personnel safety then have at it. If for safety then I'll stay away from that one thank you.

About ten years ago, my previous employer issued an engineering bulletin which basically stated that any direct, solid state monitoring of the actual e-stop ciruit was to be a no-no. They claimed the possibility of exactly what the OP describes.

I had to go out and audit all of our safety circuits and modify a few of them that violated this new corporate rule.

Monitoring an isolated contact (as Ken described) on each of the devices is the only way to go, but doesn't help when the two sets of contacts don't operate in sync. and the critical one bounces for a few milliseconds killing your process, and leaving no evidence behind.

What we did to troubleshoot them after the elimination of the direct monitoring circuits, was to use what was called a tattle tale relay.

It was just a handful of relays and din rail bases on a stick of din rail with a magnet base.

It had alligator clips on the ends of its leads and we'd tap into various points in the circtui to be monitored like this:

(test point 1)-----] [---(CR1)---
(test point 2)-----] [---(CR2)---
(test point 3)-----] [---(CR3)---

The relays had manual operators, so once connected, you press the operator to seal in the relay, and when the circuit fails, say at test point 2, its relay will drop out. This helps to fault find, is temporary, but not really a great solution

from what's been posted, I'm betting (pocket change only) that the filter circuit on each input is adding enough capacitance to affect the relay's drop out time ... (idea: forty little capacitors in parallel could add up to a significant amount) ...

also ... if I'm reading this correctly, the original poster is trying to set up something like a "first out" indicator to monitor those forty switches ...

at the risk of re-opening a nasty can of worms, I'd suggest a careful reading of this thread:

http://www.plctalk.net/qanda/showthread.php?p=245765&postcount=1


I'd be willing to bet (slightly more than pocket change this time) that careful testing would reveal that the "first out" system isn't nearly as reliable as it might appear ...

so ...

it MIGHT be possible to fix the "delayed dropout" problem - only to find out that the intended "first out" function wasn't really reliable enough to be worth the trouble ...

Ron Beaufort said:

from what's been posted, I'm betting (pocket change only) that the filter circuit on each input is adding enough capacitance to affect the relay's drop out time ... (idea: forty little capacitors in parallel could add up to a significant amount) ...

also ... if I'm reading this correctly, the original poster is trying to set up something like a "first out" indicator to monitor those forty switches ...

at the risk of re-opening a nasty can of worms, I'd suggest a careful reading of this thread:

http://www.plctalk.net/qanda/showthread.php?p=245765&postcount=1


I'd be willing to bet (slightly more than pocket change this time) that careful testing would reveal that the "first out" system isn't nearly as reliable as it might appear ...

so ...

it MIGHT be possible to fix the "delayed dropout" problem - only to find out that the intended "first out" function wasn't really reliable enough to be worth the trouble ...

Click to expand...

this in not a safety issue, although these are jam switches, in that there are 42 pockets that lay a book on a chain, if one pocket lays more than one at a time it stops the line. I am not able to use the other set of contacts on the limit because of other issues. I do believe I am getting feedback though the input modules, because without the terminals connected to the input card, everything works fine. I suppose this is a tattle circuit, but it does seem to work as far as fault finding. I was wondering if I could use a rc suppressor or something across the coil of my stop relay that would cause it to drop out quicker.

Assuming the problem is capacitive then increasing the size of the load will mean that the voltage drains more quickly. Of course increasing the load increases the current the limit switch contacts have to open. Or as well as the standard relay use a voltage sensing relay so that a soon as the voltage drops by say 10% you stop the machine. The smaller the percentage drop you look for the quicker the response but the more likely you will get false trips.

Bryan

Edit:
Thinking again you are using 110V, I guess that is 110V a.c., so rather than detect voltage, detect frequency, the AC component will disappear as soon as a contact opens.