Dry Contact Basics

[gar]

090225-0832 EST

The schematic from post #31 shows a relay. A schematic does not always show physical location and/or what a device is.

I have assumed that the Heat Trace is a commericial box with an internal relay with some kind of contact terminating on a terminal pair for connection to external equipment. That would mean the relay is not easily replaced.

On to contact ratings. As a random sample of a small PC board mount relay see:
http://catalog.tycoelectronics.com/catalog/menu/en/16453?BML=10576,16354
http://catalog.tycoelectronics.com/catalog/feat/en/c/16656?BML=10576,16354,16453

http://catalog.tycoelectronics.com/...109941,109949,109950,109951,109947,109948^N~1
Look at the contact voltage ratings. Why would this be unsatisfactory to switch a low current resistive load at 120 V AC?

This one has higher breakdown voltage ratings:
http://catalog.tycoelectronics.com/...109941,109949,109950,109951,109947,109948^N~1

If one wants to have lower voltages at the Heat Trace for safety reasons, then the added relay at the PLC makes sense.

At the PLC input board do you really supply a contact closure, or do you apply a voltage signal? Does the PLC have any 24 V DC inputs?

.

 

[Jraef]

MD88,
Do NOT take the advice of those who are saying "you should just run the 120V through the relay anyway", they are speaking in general terms without specific knowledge of what you are working with. That might be true of many stand-alone relays, but I think I know what you are looking at; it is the on-board alarm output relay of the temperature controller for the heat trace, right? As I see it, you have a PLC with 120V inputs and a dry alarm contact on a temperature controller that is only rated for 24VDC max. This is a common arrangement on temperature controllers, in fact they tend to use 9 or 11 pin DB style connectors on the back to make up the connections. You CANNOT put 120V through those wires and connectors, there is inadequate separation for that voltage, and you SHOULD NOT put any voltage of AC through it either, because they often mix communications within the same wire harness and you can create inductive noise.

Your circuit is the way to do it. Those on-board dry contacts are usually SPECIFICALLY stated to be used with what they refer to as a "line power relay", meaning an external relay that drives another circuit at any other voltage. Another term for that is called an "interposing relay".

The only possible change to look for is that a few of them provide a small on-board 24VDC power source, usually only capable of a few mA, but enough to drive what is called a "low power" ice cube relay. Another possibility is that some of them provide AC alarm relays as well as the DC, but on a different terminal block. Make sure you didn't miss it, because that would make your circuit a little easier. If they have one, but it's being used for something else, it might still be easier to add a 2 pole AC relay to that circuit and use one for the existing circuit, then use the other for your PLC input.

 

[gar]

090225-0954 EST

Jraef:

I do not disagree with what you are saying. You have provided way more information than he has.

There needs to be an adequate definition of what the output of this box is and its real limitations, and why.

.

 

[weressl]

Hello,

I'm having a bit of a hard time getting a truly straight answer on what a dry contact is... I'm under the impression it's switch that closes or opens by some external control source but doesn't have a voltage source itself - it basically just connects to pieces of wire together. So say a control panel is MONITORING a dry contact, what it's doing is trying to pass a current through a circuit and if it successfully passes it's current through the circuit it knows that the contact is closed. Is that correct? So the dry contact is not the source of the current, the thing doing the monitoring is?

I'm looking for a device (hopefully it's a simply type of relay!) that has it's own power supply and monitors a dry contact with a 48V max signal. Depending on whether the dry contact it is monitoring is closed or open, it closes or opens it's own dry contact rated for 120VAC, that in turn a PLC system will be monitoring. Keep in mind that I'm using "dry contact" assuming that my description above is correct! Can anyone tell me what this type of device might be called?

Thanks!

So why do you think you can't connect the 'dry contact' to be monitored directly to the PLC?

 

[MD88]

MD88,
Do NOT take the advice of those who are saying "you should just run the 120V through the relay anyway", they are speaking in general terms without specific knowledge of what you are working with. That might be true of many stand-alone relays, but I think I know what you are looking at; it is the on-board alarm output relay of the temperature controller for the heat trace, right? As I see it, you have a PLC with 120V inputs and a dry alarm contact on a temperature controller that is only rated for 24VDC max. This is a common arrangement on temperature controllers, in fact they tend to use 9 or 11 pin DB style connectors on the back to make up the connections. You CANNOT put 120V through those wires and connectors, there is inadequate separation for that voltage, and you SHOULD NOT put any voltage of AC through it either, because they often mix communications within the same wire harness and you can create inductive noise.

Your circuit is the way to do it. Those on-board dry contacts are usually SPECIFICALLY stated to be used with what they refer to as a "line power relay", meaning an external relay that drives another circuit at any other voltage. Another term for that is called an "interposing relay".

The only possible change to look for is that a few of them provide a small on-board 24VDC power source, usually only capable of a few mA, but enough to drive what is called a "low power" ice cube relay. Another possibility is that some of them provide AC alarm relays as well as the DC, but on a different terminal block. Make sure you didn't miss it, because that would make your circuit a little easier. If they have one, but it's being used for something else, it might still be easier to add a 2 pole AC relay to that circuit and use one for the existing circuit, then use the other for your PLC input.

You seem to have the clearest understanding of the situation! Interestingly enough, we just switched manufacturers and the new one has the same alarm contact rating and they specifically suggest a CRYDOM MS11 series relay in my situation (it's a relay with a 24VDC max coil and 120VAC rated contacts). The panel also has terminal points for 9VDC source, high enough to operate the coil on the CRYDOM relay. So if you look back at the circuit diagram I drew, just replace the 24VDC source with the terminals on the control panel board. This direction came from Raychem themselves, so I imagine it's the proper way! Thanks everybody for your wealth of information! The solution ended up being simple, but I definitely understand the system much better now.

 

[weressl]

You seem to have the clearest understanding of the situation! Interestingly enough, we just switched manufacturers and the new one has the same alarm contact rating and they specifically suggest a CRYDOM MS11 series relay in my situation (it's a relay with a 24VDC max coil and 120VAC rated contacts). The panel also has terminal points for 9VDC source, high enough to operate the coil on the CRYDOM relay. So if you look back at the circuit diagram I drew, just replace the 24VDC source with the terminals on the control panel board. This direction came from Raychem themselves, so I imagine it's the proper way! Thanks everybody for your wealth of information! The solution ended up being simple, but I definitely understand the system much better now.

PLC's also commonly have 24VDC inputs with common source voltage available on the card itself, where such 'dry contact' as you describe can be safely connected.

It saves a lot of mental excercise of "what ifs" when you provide complete information up front.:smile:

 

[MD88]

That would have been ideal! Unfortunately this is a union site and all the discrete inputs are handled by the PLC (elec dept) and thus must be 120VAC. The analog inputs are all handled by a DCS (instrumentation dept) and thus must be 24VDC. It's caused many headaches!

 

[gar]

090225-1338 EST

MD88:

Here is a link to Crydom:
http://crydom.com/en/index.shtml

and to a CS 11:
http://www.crydom.com/en/Products/Catalog/m_s11-cm_x.pdf
You have to scroll down to view the AC output specifications. Two critical ones are maximum off state leakage current, and minimum load current.

You may need additional load at the PLC input besides the PLC input resistance to get good turn-on of the CS 11.

As weressl said more up front information would be a great help.

.

 

[weressl]

That would have been ideal! Unfortunately this is a union site and all the discrete inputs are handled by the PLC (elec dept) and thus must be 120VAC. The analog inputs are all handled by a DCS (instrumentation dept) and thus must be 24VDC. It's caused many headaches!

I think, and always thought, that is criminal.

 

[weressl]

090225-1338 EST

MD88:

Here is a link to Crydom:
http://crydom.com/en/index.shtml

and to a CS 11:
http://www.crydom.com/en/Products/Catalog/m_s11-cm_x.pdf
You have to scroll down to view the AC output specifications. Two critical ones are maximum off state leakage current, and minimum load current.

You may need additional load at the PLC input besides the PLC input resistance to get good turn-on of the CS 11.

As weressl said more up front information would be a great help.

.

Both leakege current and minimum load could create problems for repeateble and reliable operation of the PLC input. You may need to put a loading resistor accross the input.

 

[MD88]

Oh dear I think I'm out of my league now... I'm assuming the leakage current would be a problem because the PLC will read the current and perhaps generate a false ON signal?
I really don't understand the concept of minimum load current... isn't the output side of the relay just a set of contacts, whose operation is independent of what is being sent through them? Oh, is this an SSR issue? I definitely need a refresher on input resistance for the "put a load resistor across the PLC inputs" sentence to make sense! How would that increase the current though the relay contacts?

 

[weressl]

Oh dear I think I'm out of my league now... I'm assuming the leakage current would be a problem because the PLC will read the current and perhaps generate a false ON signal?
I really don't understand the concept of minimum load current... isn't the output side of the relay just a set of contacts, whose operation is independent of what is being sent through them? Oh, is this an SSR issue? I definitely need a refresher on input resistance for the "put a load resistor across the PLC inputs" sentence to make sense! How would that increase the current though the relay contacts?

SS 'relays' do not have a mechanical contact, but an SCR(Silicone Controlled Rectifier) that opens or closes the circuit. They are semi-conductive materials that are controlled by a low level signal to conduct or not to conduct. They always have a small leakage current; they can not be completely 'turned OFF'. They also can not turn ON if the circuit resistance is too high, eg. the load is not 'enough'. The paralleled input resistor across the input terminals provides a shunt for the leakage current and it also provides the minimum load to assure that the SCR will turn on and off reliably and repeatably.

 

[MD88]

Hmmm... I get the shunt of the leakage current, but I don't see how adding that resistor could increase the current through the relay. Are you modelling the PLC as a 120V source with an input resistance in series with it (like a Thevenin)? Then you're putting this load resistor in parallel with the relay? I'm definitely feeling dumb!!

 

[weressl]

Hmmm... I get the shunt of the leakage current, but I don't see how adding that resistor could increase the current through the relay. Are you modelling the PLC as a 120V source with an input resistance in series with it (like a Thevenin)? Then you're putting this load resistor in parallel with the relay? I'm definitely feeling dumb!!

Yes, the resistor connected to the terminals of the input will be a parallel resistor with the input resistance. One leg of the 24VDC or 115VAC power source is connected to inut terminal I1 AND resistor terminal R1. The other leg is connected to the IR1 terminal of the interposing relay. The other side of the realy 'contact' IR2 is connected to the other end of the resistor R2 AND the other input terminal I2.

 

[MD88]

Ah I see what you mean... but if you look at what's between the two terminals of the PLC input, would it not basically be a 120V source in series with an input resistance? Like if you put a voltmeter across I1 and I2, you'd see 120V and it'd have an associated resistance as well. If you put a resistor across I1 and I2, it'd then be in series with the input resistance? I fully believe that you're correct, I just don't really get it haha!

 

[weressl]

Ah I see what you mean... but if you look at what's between the two terminals of the PLC input, would it not basically be a 120V source in series with an input resistance? Like if you put a voltmeter across I1 and I2, you'd see 120V and it'd have an associated resistance as well. If you put a resistor across I1 and I2, it'd then be in series with the input resistance? I fully believe that you're correct, I just don't really get it haha!

OK Let me modify my original one:

Yes, the resistor connected to the terminals of the input will be a parallel resistor with the input resistance. One leg L1 of the 24VDC or 115VAC power source is connected to inut terminal I1 AND resistor terminal R1. The other leg of the power supply L2 is connected to the IR1 terminal of the interposing relay. The other side of the realy 'contact' IR2 is connected to the other end of the resistor R2 AND the other input terminal I2.

When everything is disconnected and you read resistance between I1 and I2, you would be reading what is called the input ressitance. So when you connect I1 to R1 and I2 to R2, you are paralleling those two resistances.

(When the entire circuit is connected as above and you would read voltage accross I1/R1 and I2/R2, you would be reading the voltage drop and you could read another voltage drop across IR1 and IR2. The two readings will add up to the total voltage accross L1 and L2.)

 

[gar]

090225-1757 EST

MD88:

I think in an earlier post I ask what the input of the PLC was like.

To more completely expand on this:

The typical PLC input is a resistor in series with an optical coupler, or something that does the equivalent function. This is a voltage threshold type of input. When the input voltage exceeds a certain level, maybe 80 V it is guaranteed to represent a TRUE signal. When it is below some voltage like 15 V it is guaranteed to be a FALSE input. TRUE and FALSE (terms used in logic theory) can be read as 1 and 0 (terms usually used in computer circuits). HI and LOW are other terms, and sometimes HI means 0 and LOW means 1 (negative logic). Data on an RS232 signal line falls in the negative logic category and actually + corresponds to FALSE and - to TRUE. Here the thresholds are +3 V and -3 V and between these two levels is indeterminent.

Continuing with the PLC input. It usually does not include the voltage source. Thus, it is equivalent to a relay coil. Internally in an input module one side of many inputs are connected together. This common would usually be connected to the common of the 120 V supply. Usually drawn as the right hand rail of a ladder diagram. Somewhere else the hot side of 120 V is provided on a terminal block. This is usually the left hand rail of the ladder diagram. This hot supply then supplies one side of many input devices (limit switches, selector switches, photocells, pressure switches, etc. The other side of all these devices then go to their respective PLC input terminal.

Suppose no shunt resistance is put across the PLC input and that this input has an internal resistance of 30,000 ohms. Most PLC inputs do not have very well defined threshold levels for 0 and 1 and these may be quite temperature sensitive. Suppose 25 V is where switching between 0 and 1 occurs. The leakage resistance from 120 hot to an input to create a 1 state is ((120-25)/25)*30,000 = 114,000 ohms. 95/114,000 = aproximately 800 microamps. If we shunt the PLC input with 2,000 ohms the input resistance is about 1,800 ohms. Now for the same threshold point it will take 13 MA of leakage to create a logic 1 input.

The TRIAC (or back to back SCRs) switch, Crydom CS11, requires 60 MA for its minimum current. At 90 V supply this current requires a load resistance of 1500 ohms. At 120 V it would be 2000 ohms. See why I guessed at a 2000 ohm shunt resistor.

If your PLC is built with an internal 120 hot to one input terminal, and the other input terminal is one end of the input "coil", and somewhere else is the 120 common (the other end of the "coil"), then you need to put the shunt resistor between 120 common and the terminal that is the input "coil". Any PLCs that I have worked with were not built in this fashion.

There is another possibility and that is a pull-up resistor always keeps the input coil on and your external contact shorts the coil. But not a likely input design.

From the start you should have looked at the data sheets for both your sensor device and the PLC to determine exactly how each of these worked, and why things were done the way they were. From our perspective we have had to just guess at what you have.

.

 

[steelersman]

My head is really spinning!

My head is really spinning!

I don't know how you guys understand all of this technobabble. Don't take this the wrong way. I'm not putting anyone down. I'm just amazed at how much you guys know. Even the OP who was the one with the question! I feel like I'm reading something that the mad scientist guy from that movie: BACK TO THE FUTURE would say.

 

[gar]

090225-2034 EST

steelersman:

It is based on an understanding of basics, having experience with many different devices and problems, a broad educational background, interactions with many different people (like Howard H. Aiken, Arthur Burks, Wilson P. Tanner, Harry Goode), great teachers that were true educators (like Max Irland, J. G Tarboux, A. D. Moore, William G. Dow, Mellville B. Stout, Norman R. Scott, Joseph A. Boyd), and many other life experiences.

An example problem. Off the shore of North Koera on the USS Wisconsin our main search radar failed. We generally operated as close as 1000 to 2000 yards from shore. We were down to the last thyratron of a particular part number in the fleet. Also we were the Flagship of the fleet. We had burned out all the other spares. How to solve the problem with only one remaining spare? The service manual provided no solution. The thyratron and magnetron were near the top of the superstructure, about 140 ft above the water. The single shot multivibrator that controlled the pulse duration was in the CIC (Combat Information Center) room roughly about the water line. This was a long climb and no direct communication. Ultimately with only this one last thyratron and by making an arbitrary large change in the pulse duration, shortened it to a minimum, it was possible to get the system working without burn out of the thyratron. Now it was possible to go back and use the manual to correctly adjust the pulse duration.

When 16" guns or for that matter 5" guns are fired there is substantial shock imparted to the ship. This type of shock had probably caused the adjustment pot to change position. Whether this was before or after one day when we shot 300 rounds of 16" shells I do not remember. That was about 1/3 of our stock and amounted to about one shell every 5 minutes. Each weighed about 2000 #.

.

 

[MD88]

Hey I think I get PLC inputs now! That actually makes much more sense than the way I figured it was... once again, thanks everybody it was very informative. Btw, what does OP mean?? I think steelersman called me one, haha.