Mcb trip

[cliffed]

Any ideas on this one,arrived at college been working at,they had a power cut during the night,which messed up a load of things that needed resetting.At one area of the college at a DB a mcb rated at 63 amps had tripped,this supplied another DB supplying ring mains for IT suites,about 12 rings in total.These rings had Rcbo's for each circuit rated at 32amps,30ma.All circuits were on.Tried resetting 63amp mcb,but wouldn't reset.Switch DB off,mcb reset,main switch & rcbo's back on,found fault on one circuit,unplugged all equipment,reset,plugged all back in,everything ok.
Peculiar this also happened in 2 other areas same scenario ,mcb' s there rated at 40 amps & 50 amps,all reset with no problems.No real reason why this should of happened,a surge or spike on incoming mains,any ideas.
cheers

 

[Darkwood]

Inductive spike .... it sounds like the college leave all the computers on standby overnight rather than switch them off, when the power was restored the sub supply mcb's saw all the combined high inductive inrush spikes from several circuits containing many computers...

Tell college to save energy at night or give them a sequence of reset list if it happens again ...mainly to switch off the loads before resetting then sub supply...

TBH having a mcb covering a submains full of mcb's is a poor design as you have no or little discrimination in a fault and a short circuit would be just be as likely to trip the 63amp mcb as it would say a final circuit 20amp.

This is part of the reason why this issue occured in the first place - poor design.

 

[cliffed]

Agree college quite old, & sub mains are all over the place,DB's supplying Db's,ideally new sub mains would be the answer,but doubt that will happen,not yet anyway.Its the first time this as happened so maybe a one off.
Cheers

 

[shanky887614]

Inductive spike .... it sounds like the college leave all the computers on standby overnight rather than switch them off, when the power was restored the sub supply mcb's saw all the combined high inductive inrush spikes from several circuits containing many computers...

Tell college to save energy at night or give them a sequence of reset list if it happens again ...mainly to switch off the loads before resetting then sub supply...

TBH having a mcb covering a submains full of mcb's is a poor design as you have no or little discrimination in a fault and a short circuit would be just be as likely to trip the 63amp mcb as it would say a final circuit 20amp.

This is part of the reason why this issue occured in the first place - poor design.

Agree college quite old, & sub mains are all over the place,DB's supplying Db's,ideally new sub mains would be the answer,but doubt that will happen,not yet anyway.Its the first time this as happened so maybe a one off.
Cheers

the college should get the IT department on the case.

they can set it to WOL and then they only need the servers running.

thats wake on lan btw, if the computers are that old they dont support it (unlikely unless there still using old xp machines) then it can be set in bios.


the problem could even be the psu themselfs and the pc's are not powered up by that point.

have a google if phantom power in electronics and you will get the idea, in other words the psu runs in sleepmode whenever it is plugged into the wall and turns on when the ON button shorts two of its pins together

 

[Darkwood]

the college should get the IT department on the case.

they can set it to WOL and then they only need the servers running.

thats wake on lan btw, if the computers are that old they dont support it (unlikely unless there still using old xp machines) then it can be set in bios.


the problem could even be the psu themselfs and the pc's are not powered up by that point.

have a google if phantom power in electronics and you will get the idea, in other words the psu runs in sleepmode whenever it is plugged into the wall and turns on when the ON button shorts two of its pins together

There was a power failure, regardless how the computers were set they lost all power so when the power came back on its the inductive spike of all the TX - psu in the pc's spiking together hence my post ... I've seen it loads over the years...Wake on Lan will not resolve the problem... its the combined inrush of the supply TX's and the load on the TX has little to do with it.

 

[shanky887614]

There was a power failure, regardless how the computers were set they lost all power so when the power came back on its the inductive spike of all the TX - psu in the pc's spiking together hence my post ... I've seen it loads over the years...Wake on Lan will not resolve the problem... its the combined inrush of the supply TX's and the load on the TX has little to do with it.

fair enough. i wouldnt have thought leakage from it to just output 5v on the secondry would be enough to trip it.

how would you go about fixing it? contactors and timers or similar to stagger the supply?

 

[Darkwood]

fair enough. i wouldnt have thought leakage from it to just output 5v on the secondry would be enough to trip it.

how would you go about fixing it? contactors and timers or similar to stagger the supply?

Its not the output etc that is the issue here, its an inductive load so on power up it will maintain a sharp spike until the field is built up... now on its own not an issue but with 10's of computers trying to switch on together and the sub supply sees the combined spike of several circuits with many computers on hence the final circuits may remain on but the sub supply cannot cope...

As for your question how would I sort... firstly I would never design such a poor set-up, its inherently flawed not only for this reason but also little if any discrimination between the 63amp and the final circuits.. so any short circuit on any of the final circuits is just as likely to take out the 63amp and loose all the sub-main as it would just taking out the final device.

You can only limit this effect now at no cost by telling them to switch of the computers everynight (excluding server) then any power cuts will not be an issue as for what I just mentioned regarding discrim' then a new supply with probably at minimum a 125amp supply but better with the 160 up front.

This is what happens when the domestic brigade tries going Commercial and Industrial they get the jobs because they are a lot cheaper but for the reason they have messed up the design and leave it vunerable to nuisance tripping and in the commercial and industrial sectors this can be very costly and or disruptive...

 

[UKMeterman]

It is actually the capacative inrush of the switch mode power supplies, often 10s of amps each. Take a look at the mag curve of a B32. I have had one PC that managed to spotweld 99p sockets. There is something to be said about BS88 fuses, lots of inrush capability.

 

[Darkwood]

It is actually the capacative inrush of the switch mode power supplies, often 10s of amps each. Take a look at the mag curve of a B32. I have had one PC that managed to spotweld 99p sockets. There is something to be said about BS88 fuses, lots of inrush capability.

I'm a bit dated still got lots of wired wound TX's when I used to build PC as a hobby for gaming ... but yeah switch-mode too, same effect.

 

[shanky887614]

It is actually the capacative inrush of the switch mode power supplies, often 10s of amps each. Take a look at the mag curve of a B32. I have had one PC that managed to spotweld 99p sockets. There is something to be said about BS88 fuses, lots of inrush capability.

fair enough, a lot have say a 20-40amp inrush after looking on one of my fav tech sites.

swapping for a D type would be a good idea as well as upping it.


looking at pfc a D type would allow a higher pfc through it but whether it would allow as much inrush i cant reallt say as i have had very little to do with design

 

[Darkwood]

fair enough, a lot have say a 20-40amp inrush after looking on one of my fav tech sites.

swapping for a D type would be a good idea as well as upping it.


looking at pfc a D type would allow a higher pfc through it but whether it would allow as much inrush i cant reallt say as i have had very little to do with design

Problem here is a 63 amp type 'D' ... check the Zs permitted value then respond on that and you will see the issue there with your suggestion, also look at permitted Zs values of similar size BS88's

 

[cliffed]

Cheers all remarks taken on board,will explain to IT boys,,this is the first time this as happened at the college,& agree on discrimination issues,only to be solved by installing new sub mains protected by either BS 88 or mccb's.
Cheers

 

[shanky887614]

Problem here is a 63 amp type 'D' ... check the Zs permitted value then respond on that and you will see the issue there with your suggestion, also look at permitted Zs values of similar size BS88's

true but i wasnt taking the max zs into account, i was looking more at the higher rating for type D compared to bs88-3 even though bs88 can be used with a higher zs. i can see where your coming from with needing a zs of 0.15 to trip within time where the bs88 is 0.38.