Interposing Relays on output cards

[theDave2]

Interposing relays have their function. ...

Agree with everything you say, didn't want to quote it all.

I have hardly ever used them. Only when a dry contact is needed or when I have all 24vdc I/O except for one or two 120vac. Then I'll use a relay to save that odd card with only 1 output used. Imagine a panel with +300 relays, eh ?

 

[allscott]

Sounds like a poll is in order. I'm with Panic mode but Hakutsuru made some valid points.

 

[Hakutsuru]

I don't think I've ever seen a poll in this forum. I'm sure the results, as with any opinion poll, can be manipulated according to how the question is asked. I think most of us are smart enough to realize that, and maybe thats why there aren't many done.


-jeff

 

[Erik vdH]

wow - thanks for all the input although opinion seems to be divided. I am actually trying to generate a site standard for a customer. I don't want to remove flexibility for 'one off designs' but want to create a standard that works for most applications - rather than having each contractor do something different based on their particular flavour of the month.

Agreed
1) Use distributed IO

Then - in general

School 1) Use a generic PLC card and interposing relays, bigger box, more wiring, less spares to carry, less slots of PLC required

School 2) Use the correct PLC card smaller box, less wiring, more intrinsic reliability (if stuff is used in spec) but more spares to carry, might need more slots

The other thing that has not been mentioned is that with interposing relays, you can service one output at a time (assuming failure is the relay - not the card), there is more chance that you can keep a process running - even if the card is not hot-swap.


Erik

 

[BobB]

I always use high density input and output cards - input cards have negative common and so do output cards. The only exception is shoe box style PLCs when I use what is available.

I thoroughly agree with the premise that it is far easier to change an external relay in the middle of the night than a PLC output card because a relay has gone missing. External relays also have a larger current carrying capacity and a bigger gap to extinguish the arc.

High density cards save lots of space and money in my view, money particularly if you are using a PLC that requires a rack. I normally use the Omron CJ1 which is rackless (DIN rail mount), small and has 64 bit high density I/O cards the size of a large cigarette packet. Fit an awful lot of I/O in a small space.

Coming away from the output card I use plugs and wiring that I make myself - also saves space and I can make the wiring any length I require to go to doors etc.

The outputs I direct wire to indicating lights - I use IDEC and the lamps are bi-directionally protected, are BA9S and replaceable from the front of the panel unlike a lot of the, quite frankly, pain in the tail European LED indicators. I will not use them.

For driving dry contact outputs and/or contactors I then use external relays. The relays I generally use are the newer Omron G2RV types which are quite high but very narrow (space saving on the DIN rail) and have bridging bars available to cut the heck out of wiring time and costs. The bridging bars can be fitted to the coil connections, the relay common and the normally closed contact connections but not the normally open contact connection.

By the way, I do like using distributed I/O. I am using CompoNet at the present time. It is far faster, more flexible and cheaper than Device Net and is token ring based - works like a charm.

 

[panic mode]

Hakatsuru

i don't agree, some of your points are even contradicting.

lack of plc software or computer is lack of tools. there is no point in standardizing
on interposing relays or type of terminals if one can't standardize on plc platform
and have tools and at least one guy in house 9or available contractor) trained to use
that platform. if there is such company, they should think about replacing management,
and fast...!!!


i have never seen darn good electrician who wasn't familiar with relay logic. and as you know
this is what plcs are tipically programmed in. if he can google and visit this site,
he should be able to open any program. computers are not something invented yesterday,
computer literacy is today's literacy.
it doesn't take much to train on PLC person good with relay logic and computer literate... or replace...

and if they standardize on one plc platform they will need 4-5 cards to cover all I/O for whole plant.
what the heck, let's say you have 20 items (I/O cards, one or two plc processors, rack or two
and one or two power supplies). add software and training plus wages while the top gun is
in training for few weeks.
so let's say hardware would cost some 15k, software 5k, training 5k wages 5k, total 30k....


compare that with 50 (or whatever) machines, each with cabinet packet with different PLCs
but 300 of same cheap relays. what is the cost to buy install and maintain all these relays?
let's say 50x300=15000 relays, add some 50 spares, let's say all of them are socketed
and relay plus socket is only $10. That is $150500.00 just in relays, not to mention labor
to have them wired and labeled, higher cost of cabinets, used floor space etc.


30k versus more than 150k?

30k payed LATEER versus more than 150k payed up front????

i don't know about your neck of woods but to me such economics stink...

are you telling me that you would rather spend money like that and in the maintenace room,
have 50 relays versus paid training, software, and $15k worth of PLC hardware?

which one has better value to practice in spare time and learn something new and even improve
existing process? in later case the top gun can respond and maybe even fix problem without
getting out of his chair. the non-elitist top gun would have to do some walking first
(and pushing of the tool box if he could afford one after all the money was spent on relays...)



elitist? i'm just observant...

the guys using clubs for hunting and finding mates are outdated...

 

[Doug_Adam]

We standardise on interposing relays for many of our output loads.
For us, it allows standardisation of output cards to one type of 24V DC output. Generally, we find that relay cards tend to fail faster than other types, so using external relays of the plug in type allows quick changing. We can also setup our maintenance system to generate work orders to change out the more troublesome relays before their historical failure time.
Generally, we use either Omron G2R or Phoenix contact relays.

This solution works for us.

Recently, I have been looking at ET200S IO from Siemens. This IO seems to combine the benifits of interposing relays with the space and wiring savings of using different, and directly wired IO. We haven't made the move yet, but I think it looks like the way of the future.

 

[Gerry]

I have a contractor trying to convince me to install 'interposing relays' on the output of PLC cards.

-makes it a bigger, more profitable job for the contractor.

One more vote for the contractor. In fact I even do it on inputs for all the same reasons.
__________________
Mickey

Why even have a PLC?

Thank you Gerry M!

YES!!

We can also setup our maintenance system to generate work orders to change out the more troublesome relays before their historical failure time.

This is an advantage??

Low power 24VDC coils are available for most valves and contactors so there's no need to "stress" the outputs.
Why install something that you know is going to fail often and regularly?
Why design a system with a multitude of control voltages and sources?

About 25 years ago I did a PLC-3 system with over 5,000 I/O points. All I/O was 24V and connected directly. There may have been some relays on the larger motors. We did not have I/O module failures, but we did have plenty of issues with field devices and programs. We would have had to build an addition to the factory if we were going to use interposing relays for everything.

why the PLCs ware invented in the first place? (Hint: it's not that there was shortage of relays...)

YES - PLC's were developed to replace relays!

 

[JesperMP]

wow - thanks for all the input although opinion seems to be divided. I am actually trying to generate a site standard for a customer. I don't want to remove flexibility for 'one off designs' but want to create a standard that works for most applications - rather than having each contractor do something different based on their particular flavour of the month.

Agreed
1) Use distributed IO

Then - in general

School 1) Use a generic PLC card and interposing relays, bigger box, more wiring, less spares to carry, less slots of PLC required

School 2) Use the correct PLC card smaller box, less wiring, more intrinsic reliability (if stuff is used in spec) but more spares to carry, might need more slots

The other thing that has not been mentioned is that with interposing relays, you can service one output at a time (assuming failure is the relay - not the card), there is more chance that you can keep a process running - even if the card is not hot-swap.

I agree with everything you wrote. In particular the last sentence. Will be interesting to hear the "School 2" answer to how you can fix a damaged output card while the proces is running. And even if the PLC has hot-swap facility, the problem is always that there are additional outputs on the same card for other parts of the proces that cannot be allowed to be disconnected.

--------------
edit:
In our company there are two departments.

The proces department (where I am). For us the priority is flexibility and the need to keep a proces running even if one item is defective somewhere. And the physical layout is plant-wide. For these reasons distributed i/o is the answer.
In the past we were followers of "School 1".

Then there is the machine-department (the other guys). For them it is not an issue to be able to service a part while the machine is running. Any defective item has to be fixed before the machine is allowed to run. The physical layout stretches max 20 meters. Also, to be able to fit all controls into a panel that is integral on the machine, they want to use the highest density i/o of each type of i/o. So for them it makes sense to follow "School 2".

 

[Sliver]

I don't think anyone has mentioned the fact that for contactors and starters you require interposing relays if you are going to have a hardwired emergency stop. We always use them for this reason among others.

Brian

 

[TConnolly]

I don't think anyone has mentioned the fact that for contactors and starters you require interposing relays if you are going to have a hardwired emergency stop. We always use them for this reason among others.

Brian

That depends on how you design the system. I deprive the entire PLC output card of power through the e-stop string. Interposing relays can introduce their own set of problems should the contacts weld if the system is not designed correctly.

Other than using them only when switching large loads, I have found one advantage to interposing relays in some situations if you use the right relay. Some ice cube relays have a small tab you can pull out that mechanically actuates the contacts. This is a quick and dirty way to force something on. I installed a couple of panels in Europe that were set up that way because they had no PLC people at the site. Other than that I really don't see the need for them, the initial additional cost of interposing relays for every IO point is far greater than the cost of an IO card and its not very likely that a properly designed system will go through several card replacements in its lifetime.

 

[Sliver]

I deprive the entire PLC output card of power through the e-stop string.

That's interesting, never seen it done that way, thanks.

Brian.

 

[72hdflh]

I agree with Alaric, I remove power to my outputs through my E-stop string as well as giving the PLC some input to let it know it has been E-stopped to reset whatever is necessary so the machine will be in a safe state and doesn't start back up where it left off at when power is restored.

 

[Stephen Luft]

I don't know if I have really ever understood the concept of high density modules. Sure, you get many points in one module, but the ratings are next to nothing so you are limited in what you can control with the module directly. Therefore, what you are saving in PLC space and cost, you are giving back in all the additional hardware to make your I/O "real world"

I also don't 100% buy into the notion that it is easier to replace a relay than a module. A great deal will depend on the PLC and the module. Many PLCs now have pluggable terminals, therefore it could be just as easy to replace the module as a relay.

Interposing relays have their function. using them for few selected loads is part of the job
(it's part of design process and involves reading specs and sizing...).
but putting blindly one relay on every single output is rather extreme. I call this
waste of resources and it's not just overkill, it could be very poor design too...

I do agree with Panic Mode in that interposing relays do have their place in applications...but not for every output (unless truly needed). Large inductive devices such as motor starters typically require a much higher rating than any PLC can offer, therefore an interposing relay is most certainly required.

The ultimate problem is that you have very few controllers that are designed with higher current outputs. Sure, some PLCs offer high current modules, so it isn't like they aren't available, you would just require more of them. The bigger problem is in the brick / micro products where the output ratings (relay and solid state) need to improve. Many of the PLC products are getting smaller and cheaper, mainly because they are taking the direct switching capablities out of the controller and making you use additional hardware. The reduced cost isn't really a savings, just a transfer of cost into other external components now required.

School 1) Use a generic PLC card and interposing relays, bigger box, more wiring, less spares to carry, less slots of PLC required

School 2) Use the correct PLC card smaller box, less wiring, more intrinsic reliability (if stuff is used in spec) but more spares to carry, might need more slots

Erik...if you are leaning towards school 2, identify controllers with pluggable terminals. This will make your life much easier if you ever have to replace a module or controller.

The purpose of the PLC is to replace hard wired relays, hopefully making your application less complicated and more reliable.

 

[JesperMP]

The purpose of the PLC is to replace hard wired relays

This nonsense has been stated several times, but that doesnt make it more true.
As if it is written in stone "though shalt not use an interposing relay".

The "purpose" of a PLC is whatever you chose to use it for.