Contact Resitance

[mityeltu]

During my plant's last pre-CDBI (Component Design Bases Inspection) self assessment (this was performed before I started working here) an issue was discovered in a few of our calcs.

Apparently, it was decided that a statement that contact resistance of hand switches, breakers, limit switches, etc. being negligible was not an acceptable statement without some sort of supporting documentation. The basis for the assumption was that the voltage drop across these contacts would be insignificant when viewed against the voltage drop along the cable connected to these contacts.

These contacts are mostly associated with our 125V vital battery system, so I understand that after a number of hours when the voltage has decreased to near end of life of the batteries, the VD will be higher across the contacts, but this will still be very small in comparison to the VD along the cable.

My question is, how do I justify this statement if my justification as stated above is not acceptable? I have contacted some of the vendors for the SSC's but they are unwilling or unable to supply the information on the contact resistance.

Any thoughts on how to answer this?

 

[zog]

Not sure what you are asking?

 

[mityeltu]

Well, what I'm asking is this: How do I justify saying that contact resistance on relays, hand switches, limit switches, etc. is negligible?

The justification that was used has been rejected. That justification was this: The resistance of the contacts is significantly lower than the resistance of the cables connected to them and therefore, the voltage drop across them will be insignificant compared to the voltage drop due to the cable.

Is there some other way to justify this other than what I have already stated?

 

[petersonra]

Well, what I'm asking is this: How do I justify saying that contact resistance on relays, hand switches, limit switches, etc. is negligible?

The justification that was used has been rejected. That justification was this: The resistance of the contacts is significantly lower than the resistance of the cables connected to them and therefore, the voltage drop across them will be insignificant compared to the voltage drop due to the cable.

Is there some other way to justify this other than what I have already stated?

There is usually published data available from the manufacturer of the switches and relays.

It does usually end up being negligible, HOWEVER, the term negligible is very application dependent. I have run into several cases where it was not negligible.

In any case, you would need to start with a workable definition of what you mean by negligible.

 

[zog]

There is usually published data available from the manufacturer of the switches and relays.

It does usually end up being negligible, HOWEVER, the term negligible is very application dependent. I have run into several cases where it was not negligible.

In any case, you would need to start with a workable definition of what you mean by negligible.

Exactly. We need some values here to be of any help.

 

[templdl]

Contact resistance to me is often like shooting at shadows. First, consider ohms law and the voltage drop that contact resistance can cause. Voltage drop=heat when you figure what the heating watts would be calculated out as when contact resistance is measure with an ohm meter. Please don't forget that an ohm meter uses a very low voltage when it measures resistance.
Exercise the switch, breaker, etc and the resistance measurements will most likely change as contacts normally will rub together when closing.
The best way that I can think of to measure contact resistance is to actually energize them under their operating voltage and load and then measure the voltage between the line and load of the contacts. Knowing the current and voltage drop across the contacts you can calculate what the contact resistance really is.

Using an ohm meter to me doesn't provide a legitimate result. What it does is to indicate that the device should be investigated further and not simply rejected based upon your results. As a former breaker application engineer I have never had any data regarding contact resistance and I knew many old timer breaker engineers many that dated way back to the '50s before they retired. I also was tight with the warranty department having had the opportunity to sit at a test bench to test products that were returned. Contact resistance was never an issue unless they just plain failed. contact resistance under normal load=heating, heating weakens the spring that holds the contact closed, the contact pressure is reduced which increases the resistance resulting in more heating and most likely arcing. Should you run across a possible example of this and you are going do scrap it, take it to the test bench and take it apart and do a post mortem.

 

[Smart $]

...
The best way that I can think of to measure contact resistance is to actually energize them under their operating voltage and load and then measure the voltage between the line and load of the contacts. Knowing the current and voltage drop across the contacts you can calculate what the contact resistance really is.

...

I don't know about best, but will agree to being both the easiest and most accurate method :happyyes:

If your concern is voltage drop, that IS what you are measuring. You don't need to calculate anything, including contact resistance.

 

[mityeltu]

I understand values would be helpful, but understand, this is a "typical" calc for a number of locations throughout the plant. The values are for a particular location and not necessarily all location. They are meant to be a representative sample.

The manufacturer(s) of the equipment will not, for whatever reason, supply me with the information. I have not been able to find it in any of the literature on their web site. GE is not exactly the friendliest company to get information from.

At any rate, here is the circuit layout.
From control power fuse (all cables are 600V, #14, 7strands, PXMJ): 102' from fuse to JB, 634' to next JB, 51' to pentration, 240' to JB, 35' to limit switch (in this case).
This comes out to 1062'.

Unless contact resistance is in the order of .5-1 ohm, I should be able to say it's negligible, but that does not make it worthy of justification based on my previous statements.

According to whomever did the pre-CDBI self-assessment, the statement that it's negligible because the resistance is very small compared to the resistance of the connected cable is somehow not an acceptable justification. What I'm struggling with is how to justify it without saying that. I can't even come up with any good weasle words for this.

 

[gar]

120531-1054 EDT

mityeltu:

You measure contact resistance of a switching device by making a 4 terminal resistance measurement. In this measurement you pass a known current thru the device, then with separate leads (the 3rd and 4th terminals) you measure the voltage drop across the contacts of the switch. The 4 terminal measurement excludes the voltage drop across the connections to the switching device.

Contact resistance may not be constant with applied voltage and current. Mechanical silver-cadmium oxide contacts may be very poor at low voltage. Don't use these contacts at 5 V DC.

Semiconductor switches, Triacs for example, have a volt or more of on-state voltage drop.

Thermal and/or magnetic circuit breakers will have considerably high resistance from terminal to terminal than just the mechanical contact within the breaker.

Is the question being asked by some non-technical administrator? You need to find out what the real questions and goals are.

.

 

[gar]

120531-1202 EDT

mityeltu:

What kind of limit switch is this device? What is the voltage applied to the limit switch when the switch is open? What is your calculated wire resistance? What is your measured wire resistance (put a dead short at the limit switch and measure the voltage at the input to this long cable and the current thru the cable)? You could also measure the limit switch voltage drop at the limit switch and the current thru it when the limit switch was closed.

.

 

[zog]

The best way that I can think of to measure contact resistance is to actually energize them under their operating voltage and load and then measure the voltage between the line and load of the contacts. Knowing the current and voltage drop across the contacts you can calculate what the contact resistance really is.

I disagree, too many variables. Resistance of leads, variations of current during test, etc... Not to mention it is energized and PPE requirements come into play.

You measure contact resistance of a switching device by making a 4 terminal resistance measurement. In this measurement you pass a known current thru the device, then with separate leads (the 3rd and 4th terminals) you measure the voltage drop across the contacts of the switch. The 4 terminal measurement excludes the voltage drop across the connections to the switching device.

Agree, also called a DLRO, microhmeter, Ductor, etc... we use a 100A model for all of our measurements (Typically required in the specs to use 100A)

 

[petersonra]

It might even be that the actual purpose of the switch matters in the analysis.

I would have a hard time seeing how the contact resistance in a typical switch at 125VDC is going to matter if there is more than a few milliamps of load. Even with the 2-3 Ohms of resistance in the wiring.

 

[dpeter]

I gather from this that you would like to eliminate the contact points from consideration in the design? You need to quantify "negligable" and then demonstrate that the contacts are at or below that number.

 

[K8MHZ]

I gather from this that you would like to eliminate the contact points from consideration in the design? You need to quantify "negligable" and then demonstrate that the contacts are at or below that number.

And will stay at that number for the life of the devices?

Would such a consideration take into account the increase in voltage drop as the contacts age? The conductor's resistance won't increase with use, whereas the contact's resistance may. The increase (due to oxidation and arcing) may cause a negligible value to become quite a bit more than that over time.

Just some food for thought.....

 

[templdl]

I don't know about best, but will agree to being both the easiest and most accurate method :happyyes:

If your concern is voltage drop, that is what you are measuring. You don't need to calculate anything, including contact resistance.

You're right if voltage drop is the issue. But voltage drop is a result of contact resistance. As such would not the heating watts be VxI? I would be concerned about the contacts getting hot as the voltage drop increases. Ideally you don't want any contact resistance but that's not the real world. Regarding breakers, switchgear, MCCs, pnlbds, there will be heating watts which some designers want to know what it is so they can consider HVAC requirements. This may not be directly related to the OP but voltage drop is a result of resistance which ends up as heating.
I agree that depending upon the actual application voltage drop may be of a concern but may not involve enough current to result in any significant heating but the loss of voltage because of a voltage drop is. As such all this is guessing unless we know what the goals are.

 

[templdl]

And will stay at that number for the life of the devices?

Would such a consideration take into account the increase in voltage drop as the contacts age? The conductor's resistance won't increase with use, whereas the contact's resistance may. The increase (due to oxidation and arcing) may cause a negligible value to become quite a bit more than that over time.

Just some food for thought.....

I know that with breakers there is no guarantee that contact resistance will remain the same throughout the life of the device. I'm a breaker guy and I know better. In an ideal world we would like to think they would. Opening and closing a breaker numerous times causes the contacts to rub together an reset themselves as well as helping to redistribute the lubrication on the moving parts.
In an ideal work breakers do need to be exercised but they are often out of sight, out of mind, inconvenient to exercise them, etc.
Turning off a breaker means turning off power. I know what a pain it is to even have to go around and reset all of my digital clocks when there is a power outage. In an industrial/commercial facility there are other implications when exercising breakers.

 

[mike_kilroy]

......Apparently, it was decided that a statement that contact resistance of hand switches, breakers, limit switches, etc. being negligible was not an acceptable statement without some sort of supporting documentation. The basis for the assumption was that the voltage drop across these contacts would be insignificant when viewed against the voltage drop along the cable connected to these contacts...........

the proofs in the pudding.....

empirical answer should be enough 'supporting documentation.'

so why not MEASURE actual volt drop across each component in question - the wires, the contacts, etc - and THERE is your documentation. the real McCoy. If your measurements show the drop across the contacts is 'significantly' less then you have your proof. nothing further should be required.

 

[mityeltu]

I agree with all these posts, but I think the topic drifted. I understand that testing these devices would prove definitively whether or not the assumption is valid, thereby removing it as an assumption; however, to then suggest that this empirical evidence gathered on device z also applies to devices a through y is yet another assumption that would then have to be justified by similar means. This is becoming circular in reasoning. That aside, there are hundreds if not thousands of contacts on switches and relays that would then have to be tested individually. This simply will not be a valid option.

I would have imagined there is a way to justify neglecting contact resistance without empirical evidence. I should have thought some form of mathematical proof such as current density given a particular metal content, or considering each contact as a very short cable of some metal with some resistance value of x-ohms per 1000 feet and therefore resistance is negligible.

I don't work in the field, and I can't just go test these things myself. I work in design and I'm not even allowed to turn a screw. Aside from this, our craft folks tend to want to talk about it rather than actually test the equipment. I might ask today and get results next year. If this were my house, it would be a non-issue, but ...

 

[Strathead]

I gather from this that you would like to eliminate the contact points from consideration in the design? You need to quantify "negligable" and then demonstrate that the contacts are at or below that number.

This sounds like a paper pusher, say Government or nuke or something. I feel you have two dierctions you can go. Make up a number that you are comfartable with, such as less than .0001 ohms at normal operating temperature, humdity and conditions, and just live with it. Bottom line, you can't define voltage drop amount or percentage, because that is based on the voltage and the amperage, and I believe you indicated that this is variable. You can't quantify ohms because this will change from heat buildup, outside air temperature and humidity, and the condition and EXACT precision of the contact surfaces every time they close and that will change. Every time the contacts open the arc created will minutely or majorly change the shape of the contact. Every time they close they are made to swipe across each other and self clean. So every time they close the resistance will change.

So either make up the number, or start asking 100 questions of the person who says the answer you gave isn't enough, until they get the idea that either they make up a number, or the live with your answer!

 

[ggunn]

If you are confident in your design and willing to take the risk, make up a number and dare them to prove you wrong.