Contact erosion is mechanical switches?

[Flyback]

Hello,

Can you confirm that the DC voltage level above which contact erosion in cheap mechanical switches becomes very significant is 30VDC?

If you search Switch > Rocker Switch, then above 30VDC rating, there is hardly anything available. And what is available above 30VDC is extremely expensive.



Just quoting from here...
https://www.popularmechanics.com/cars/news/4226979

So what went wrong here? The automotive industry found that switching off 42 volts wasn't that easy. Huh? A mechanical switch for, say, the window motor, has two metal contacts (well, a bunch, actually). When you press a switch, the two move into contact with each other and current flows to the motor. When you let go, they snap apart, interrupting the flow of electricity. But this doesn't happen instantaneously--the surface area of the metal contacts that touch doesn't go from something like 10 square millimeters to nothing in zero time. As the contacts move apart, the current density goes up in inverse proportion to the remaining area. And at the last instant of contact, the current density is high enough to melt and then vaporize the surface of the metal as a short electrical arc bridges the widening gap. Eventually, the contacts wear to the point that they don't work anymore. In a 12-volt system, this phenomenon was controlled by careful contact design and attention to the metallurgy of the contacts themselves. Forty-two volts draws a much longer, hotter arc, and vaporizes more metal.

Contact erosion on 42-volt systems was unacceptable, unless automakers upgraded the switches with more expensive metals and stuff such as spring-loaded contacts that jumped apart faster to reduce the arcing. Would you accept a car that had to have the switches for the power windows replaced every 10,000 miles?


Read more: Whatever Happened to the 42-Volt Car? - Popular Mechanics
Follow us: @PopMech on Twitter | popularmechanics on Facebook
Visit us at PopularMechanics.com

 

[tcmtech]

No I can't confirm it being I work with many devices that switch up to around 300 VDC at 100 - 200 ma with standard off the shelf open frame relays and they stand up to that for years without contact issues.

 

[jpanhalt]

I don't think there is an absolute threshold for voltage. You can weld at very low voltages, for example. Also, from my experience, an ordinary 12V starter relay can fail to release(open) at 12V, if the current being switched is high enough.

I suspect one reason conversion to 42V was delayed is safety. 42V DC is clearly enough to electrocute someone; at 12V, electrocution is much less likely and probably requires some defect in the skin.

John

 

[crutschow]

It's related to the inductance of the load as well as the operating voltage.

 

[panic mode]

there is no magic limit on voltage or current where erosion starts, it is more pronounced with higher voltage and currents but to deal with it components are designed to minimize wear and provide reasonable life. type of load makes difference too (inductive loads are always a problem).we may not know real reason behind 42V car, i've read about many potential ones (dangerously high, too large voltage swigs, transients problem etc.)

 

[tcmtech]

I suspect one reason conversion to 42V was delayed is safety.

Actually it had little do with safety or switch design.

"High-voltage system were slated to appear as secondary systems on cars and trucks, with the normal 12-volt wiring in place until the industry gradually changed over. That was necessitated partly by the need to transition the industry to high-voltage components and practices. Just imagine the changes needed for the aftermarket accessories and electronics industries to redesign all of their components to match, and you can start to explain the death of the superelectric car. Oh, and you can bet it would have been more expensive to have two separate electrical systems on every vehicle. Not to mention it would have made electrical troubleshooting even more of a nightmare.

Innovations such as wire multiplexing have reduced the size and mass of low-voltage harnesses, and electrical power steering, after 10 years of development, now operates well on 12 volts. Meanwhile, manufacturers found other ways to improve fuel economy--idle shutoff being only one of them--without overturning the entire industry. And now we have hybrids with 250-volt-plus batteries and inverters to run the starter motor. Consequently, 42 volts has just dried up and blown away."


https://en.wikipedia.org/wiki/42-volt_electrical_system

https://www.popularmechanics.com/cars/news/4226979

Mostly it was intended to be a stop gap concept until auto tech caught up with the rest of the industrial power electronics design and application which is where the new hybrid vehicles are running 250 - 400+ VDC power systems for the high load stuff.

BTW. This is not his first or only thread relating to this concern that has covered it in some degree of detail. .

https://www.electro-tech-online.com/threads/24vdc-led-lighting-systems.135547/

https://www.electro-tech-online.com/threads/100vdc-bus-and-arcing-contacts-in-connectors.134885/

The thing is there is a whole world wide industry that works just fine on DC voltages well above 24 volts and at currents way past what the OP has issues with.

The OP's issue has nothing to do with safety or design. It's all based on incomplete or improperly educated understandings of things. That and just being cheap.

 

[tcmtech]

we may not know real reason behind 42V car, i've read about many potential ones (dangerously high, too large voltage swigs, transients problem etc.)

Actually it was too low to do any good not too high. To power entire primary drive systems 42 volts is way to low. In fact even at 300 volts the present generation of hybrids can't produce all that much leak DC drive motor power. Most only have 20 - 30 HP available when running on 100% electric power which in most vehicles today makes up for 20 - 30% of the mechanical power the IC engines make.

Even with the high end EV's like the Tesla roadsters they have DC buss voltages working 375 volts range coming from 85 KWH capacity battery packs that drive the massive 300 peak HP drive motors which take some 500+ amps while doing it.

To get level of power on a 42 volt system would take some 4500+ amps!

That's one of the reasons the 42 volt systems didn't catch on. They just didn't give enough gians where raw electrical power was need the most.

To be honest many common electric golf carts are running on 48 volt DC systems and they represent the practical range of what lower voltage DC is good for in people transport applications.

 

[Flyback]

It's related to the inductance of the load as well as the operating voltage.

I agree, though this is easily mitigated (in the case of a discrete inductor) by putting an anti-parallel diode across the inductor. -Come to think of it, many equipments, including SMPS's , do have an input filter inductor, and I am surprised that in many cases they do not have a freewheeling diode across them for this reason.

Also, we all appreciate that contact sparking of switch contacts is a bad thing, but much literature states that this very contact sparking is actually necessary in order to create a valid connection between contacts. That is, in any switch, proper contact is only achieved by a high temperature contact spark which removes the inevitable, thin film of corrosion which covers every metal to varying degrees....even plated metals, and yes, even gold plated metal, even though with gold the thickness of the oxidation is very thin indeed.
I think brass is not quite as good as gold for having resistance to corrosion.

So my question is, is it necessary to put capacitors across switchs that are switching low current so that you get that slight contact spark which correctly, and properly connects the contactors electrically?..............after all, suppose you are only switching say 1mA from a high impedance source, then 1mA will not be enough to give you the sufficient contact spark.........so a capacitor could be needed across the contacts in order to give the contact spark?

 

[kubeek]

Since when does gold make oxides and consequently oxide layers?

 

[Flyback]

..There will be some kind of reaction between gold and something in the atmosphere which causes some kind of "rust" or film to appear on the gold surface, even if only tiny.
Regarding the need for contact sparking in order to get a good connection in switch contacts.....this is a genuine problem....and in vehicles where there is much vibration, often the vibration breaks apart the molecular parts of the switch contacts where the contact fusion has occurred, and the electrical connection can become compromised...therefore in some military vehicles, they get around this by either having clip-on magnetic induction contacts, or obviously by having heavy spring clamps in the switches to hold the contacts tightly together, or as mentioned , using gold or brass plated contacts

 

[ronv]

I think you are talking about fretting corrosion. Usually that shows up when the underlying base metal gets "rubbed" on top of the gold. Of course if you get flux or something inside a non sealed switch you will have problems. Different from oxidation.

 

[ClydeCrashKop]

From another thread:
The power system of the Long Island Railroad consists of three- phase alternating current transmission at 11,000 volts, 25 cycles. Rotary converter substations convert this A.C. power to 700 volt DC third rail power at. 13,000 amps. Because extremely high voltages exist in sub-stations, no entry should be made into the area.
A friend of mine was a big shot electrical engineer and programmer for the Long Island Railroad. He said a big problem that they have is disconnecting that DC when trains are running. He said the arcing wouldn't quit and would melt the contacts, so they use hydraulic pressure to open the contacts and force dielectric oil between them to extinguish the arc.
That is some serious electricity!

 

[tcmtech]

and in vehicles where there is much vibration, often the vibration breaks apart the molecular parts of the switch contacts where the contact fusion has occurred

BS.

Do you have any idea what level of G Forces are required to shake the molecules out of a metal?
Just look at the specs sheets for any switch. If they have any level of vibration ratings they are in the 100's of G or higher.

therefore in some military vehicles, they get around this by either having clip-on magnetic induction contacts, or obviously by having heavy spring clamps in the switches to hold the contacts tightly together,

I am calling BS on that as well..

I have scrapped tons of military electrical gear over the years and I have never seen anything of the sorts.

THe vast majority of military grade switches are simply common designs made weather proof and underrated for service if even that. That is a normal 20 amp switch in a military design might be only rated for 10 amps and its made weather tight.

 

[dr pepper]

I second that, why would you design all your own switchgear from scratch, special purposes maybe but not vehicles, using existing well designed and tested switches makes much more sense.
The switches in my wifes exmil land rover are the same as prod ones, only the ignition circuit is different being watertight.

 

[rumpfy]

Hi Flyback,
The problem is more to do with the amount of energy which has to be dissipated in the switch contacts when the circuit is interrupted. The energy stored in the load device may have nowhere to go when the switch is turned off. Some circuit arrangements dissipate the energy stored in the load by directing it to flow into the load itself. This tends to maintain the energy in the load and cause a slow turn off. In a Kettering type of ignition system, the energy in the coil when the contacts open is transferred to the capacitor (condenser). This reduces the rate of build up of voltage across the contacts so that by the time the voltage across the contacts reaches its maximum value, the contacts are well open and can withstand that voltage without arcing (or breakdown).
The contact material does in fact vaporise in an extreme situation so contacts need to be robust. However, the typical contact materials can oxidise on the surface and can become dirty and make a poor contact, so it useful to have a small amount of arcing; enough to keep the surface oxides burnt off.
Hope this helps.

 

[Flyback]

I am making/breaking the 10mA of coil current in a relay......the coil has a freewheel diode across it. The resistance is some 1950 Ohms all told. 10mA isn't much and I wondered if its enough to give enough arcing, as you appear to recommend is necessary to a degree.

 

[ericgibbs]

hi FB,

Look thru this pdf for relay contact 'wetting'.

E

 

[ronv]

I am making/breaking the 10mA of coil current in a relay......the coil has a freewheel diode across it. The resistance is some 1950 Ohms all told. 10mA isn't much and I wondered if its enough to give enough arcing, as you appear to recommend is necessary to a degree.

You should read the spec.. If it doesn't have a spec. don't use it.

 

[rumpfy]

FB,
In your case, the energy stored in the relay coil is dissipated in the coil itself, due to the diode, when the relay is turned off. The back emf generated by the collapsing coil current changes polarity and the current is caused to flow in the diode. Therefore, the switch contacts will not see any sparking or excessive loading. If your coil current is only 10 mAmp, the amount of energy in the diode is likely to be VERY small indeed. Almost any diode would do the job.
Hope this helps.