Standard accessories used with DC

[GarethW]

Have you tried....

http://www.gepowercontrols.com/eu/p...ial/minature_circuit_breaker/mcb_ep100uc.html

Gareth

 

[Monkeh]

apart from arcing at contacts in MCBs (which will also break down the switch contacts) there is an odd effect with DC, where IIRC there is a migration of molecules (or possibly a smaller particle, I can't remember) shifting the metal from one point to the other. I learned about this ages ago and have forgotten. With an AC circuit, the particles move one way, then back again every cycle so it doesn't matter.

http://en.wikipedia.org/wiki/Electrolysis

 

[JohnD]

yes, I know about electrolysis but there is no electrolyte in a switch, so not sure that's the effect.

 

[sparkticus]

there is an odd effect with DC, where IIRC there is a migration of molecules (or possibly a smaller particle, I can't remember) shifting the metal from one point to the other.

Yep that is called the electromigration effect. The transport mechanism does involve ions (which is essentially a molecule with one or more electrons missing from it) The process requires both current & heat - of course you usually get the heat as a result of the current. The interconnecting thin metal (usually Au or Al or Cu) conductors that interconnect the active regions of semiconductors devices suffer from it considerably and it is a recognised failure mode within them especially in power semiconductors such as MOSFETs, Triacs, LEDs, diodes but it is also a failure mode in microprocessors/microcontrollers etc. solder joins are also susceptible to electromigration failure over time.

 

[Monkeh]

Hmm, okay, I forgot about that.

 

[sparkticus]

yes, I know about electrolysis but there is no electrolyte in a switch, so not sure that's the effect.

There is in a room containing a bath or shower

 

[OwainDIYer]

The DC isn't dimmed. It's a fail safe system so if the mains is lost, the house lights come on at full power regardless of what state they were in for show purposes.

Would they come on though if there were downstream faults on the AC circuits, rather than total loss of mains?

 

[IJWS15]

Not an electricial but a railway trained engineer.

Ollder diesel locos used DC up to 1000v and contactors used arc chutes to blow out the arc when switching on load.

Would be very wary or you may find that some of the circuits don't switch off.

As far as migration goes you get both heat and current as you switch a circuit. Contactor tips are designed to roll so the part which experiences the arc (and degrades) is not the part which carries the current when the circuit is made.

 

[bernardgreen]

In terms of switches you should be OK. 230/240v AC-RMS is equivalent to 230/240 DC. It's the arcing that will get them. That's where the suppression caps can help.

NO

In some circumstances 230 volt DC will sustain an arc across the 3mm gap in a 230 volt AC switch.

In this emergency lighting application it is un-likely that the switches are operated very often while on the DC supply so accumulated damage from arcing will be minimal.

It might be cost effective to replace the 230 battery bank with a 24 volt bank and fit a pair of 24 volt to 230 volt invertor of suitable size. If not now then do it when the batteries are due for replacement.

 

[JohnD]

Mr Blobby knows.

View media item 1576

 

[JohnW2]

In terms of switches you should be OK. 230/240v AC-RMS is equivalent to 230/240 DC. It's the arcing that will get them.

I would have said that may be a little contradictory. Insulation-wise, there's obviously no problem in using DC at least as high as the AC RMS rating. However, as everyone says, arcing is the problem, and (although I'm struggling abit with the physics here) I imagine that's probably why components with switching contacts often have a much lower DC rating than AC rating - for example, the unbiquitous octal plug-in relays generally have contacts rated at 230v AC or 28v DC.

Kind Regards, John.

 

[sparkticus]

In terms of switches you should be OK. 230/240v AC-RMS is equivalent to 230/240 DC. It's the arcing that will get them. That's where the suppression caps can help.

NO

In some circumstances 230 volt DC will sustain an arc across the 3mm gap in a 230 volt AC switch.

OK but if you read what I wrote I did say "It's the arcing that will get you"


In some circumstances AC rated switches are fine with DC applications. I have used 6 amp Type B Wylex mcbs for 24 volt DC applications and they have been fine though to be fair not operated much. I have also used standard light switches (approximately 3 years ago) to switch approximately 2amps 24 Volt DC which do get used regularly and have not seen any problems thus far. Obviously as the DC voltage and current increases there is more likely going to be a problem.

In the distant past on semiconductor test rigs we regularly switched over 40 amps DC with no significant switch failures that I remember. In fact on thyristor test rigs the MCBs where always being hammered because it was quite common for a thyristor under test to go short circuit.

 

[sparkticus]

In terms of switches you should be OK. 230/240v AC-RMS is equivalent to 230/240 DC. It's the arcing that will get them.

I would have said that may be a little contradictory. Insulation-wise, there's obviously no problem in using DC at least as high as the AC RMS rating. However, as everyone says, arcing is the problem, and (although I'm struggling abit with the physics here) I imagine that's probably why components with switching contacts often have a much lower DC rating than AC rating - for example, the unbiquitous octal plug-in relays generally have contacts rated at 230v AC or 28v DC.

Yes I could have been more clear and that does sound a little contradictory.

I meant that for lower current/lower inductance DC applications AC switches (in my experience) have a life expectancy not much shorter than in AC applications. In higher current applications especially with inductive loads then of course the contacts will sit there and become a discharge lamp.

Again, I can only speak from experience, I have used AC switches extensively for DC applications and not seen significant problems. Now they may not meet regulations for RF Lightings application and perhaps not not in DC applications for PV panels but that is a different matter. I am only speaking of empirical reliability.

 

[sparkticus]

However, as everyone says, arcing is the problem, and (although I'm struggling abit with the physics here) I imagine that's probably why components with switching contacts often have a much lower DC rating than AC rating - for example, the unbiquitous octal plug-in relays generally have contacts rated at 230v AC or 28v DC.

Yes that is correct. In an AC circuit the voltage goes to zero N times per second based on the AC frequency. For 50 Hertz mains you would see 100 zero crossings per second. When you switch off a switch carrying AC then the arc has a limited lifetime of probably maximum 10 millisconds (for a 50 hertz system) With DC there obviously is no zero crossing point so you essentially get into the dielectric strength of the ionised air/gas between the contacts which sadly is not very strong so the arc persists for quite some time. With very high current or simply the switch contact gap being to small the arc can continue until the gap increases as a function of it burning away (not good of course)

 

[JohnW2]

However, as everyone says, arcing is the problem, and (although I'm struggling abit with the physics here) I imagine that's probably why components with switching contacts often have a much lower DC rating than AC rating - for example, the unbiquitous octal plug-in relays generally have contacts rated at 230v AC or 28v DC.

Yes that is correct. In an AC circuit the voltage goes to zero N times per second based on the AC frequency. For 50 Hertz mains you would see 100 zero crossings per second. When you switch off a switch carrying AC then the arc has a limited lifetime of probably maximum 10 millisconds (for a 50 hertz system) With DC there obviously is no zero crossing point so you essentially get into the dielectric strength of the ionised air/gas between the contacts ....

Yes, I understand that. What I was 'struggling' with is the high degree of 'voltage de-rating' for DC. Once ionisation has occurred, is the arcing not primarily current-, rather than voltage-dependent - or have I got that wrong?

Kind Regards, John.