Strange, Strange fault.

[mikep]

Garlic and holy water,

What about wiring in a time switch to turn the mcb off during the night then back on in the morning.

 

[nsbs]

Ian,What device is it mate?

What does it give you, have you done a Fourier on it?

Can you upload some data catpures as pics?

If I have the software I'll go through the data too.

A Fourier may give you answers as it will show harmonics of the fundemental, I am guessing that this is a single pole device that is tripping?

The recorder is Electrocorder (but I cant remember the model number) I will be back on site either Sunday or Monday, and will take some pics then.

As for the Fourier coefficients, your grasp of calculus & algebra must be a hole lot better than mine. To be honest I have only ever calculated this a couple of times probably 20 odd years ago when I was studying a HND and would not know where to start now. But I seem to remember that the temperature (which I have not recoded accurately) was needed along with the sine & cosine,

 

[Sidewinder]

The software should Fourier this for you?

No you won't need the temp "really".

I am thinking about changing the plots from the time domain to the frequency domain to show up harmonics as well.

These could be stacking up?

How are your power poles fixed to the strucure of the building?

 

[Manator]

I do not think its that complicated Sidewinder, I did think of harmonics and magnetism build up but could not think of a suitable scenario for the delay in the time intervals. I think something unrelated is happening during the night that has an effect on that phase, I did think about any timers on the system for instance, but this would result in the trip time to be the same every night. I am thinking aloud as I type by the way so you will have to bear with me, any DNO switch overs would have been recorded by nsbs, I would however like to see the results and see if there is anything unusual.

If you did an experiment with electro magnetic fields for instance, usually the build up and dissipation could be measured with some accuracy so you could predict the intervals, this prediction would to some extent be constant and have a value, if you get my meaning. This would result in very similar trip times should this be the fault.

From the description of testing so far I would say that I would swap the tower effected for another and see if it occurs again, if only to rule out the tower itself, if it does happen again with a different tower I would investigate anything that occurs during the night, like air con units or lighting systems. I must admit I am stumped, and would like to find out the outcome. Very thought provoking.

 

[androiduk2002]

After reading through all the previous posts and comments , I have a point all though I am probably way off of the mark I will explain my logic as I go.

You say it only happens at night , so I am thinking what might happen at night one thing that happens if little critters come out to play , another is temperature drop.

Could it be little critters interfering in some way ? (sorry if this sounds awful) releaving themself and causing a short ?

Or could it be condensation forming and dripping onto something ?

As its a differant time every night I am ruling out any kind of time clock kicking in as this would be the same time every night so it comes down to something natural I would say and those are my 2 most probable things that might happen at night.

With all the combined brain power thinking on this I am staying away from technical issues as those that have inputed so far know far more than me , so I look at other causes hence my thinking .

Sorry if no real help.

 

[Sidewinder]

What comms gear is in the building?

Servers, incoming/outgoing comms lines etc.

 

[Manator]

Oh, good point Sidewinder, I wonder if any comms gear is linked, server resets on auto update, that would happen at various times through out the night. Could be way off though, I also wonder if any com gear is linked to the MET.

 

[kme]

I know we`ve sort of discounted timed events; but consider a heating system. Although its time switch would bring it online at the same time; the first temp. satisfied shutdown would NOT necessarily be the same time; due to remp. differences internal & external, etc.

Interesting........

Can`t see rodent widdle being the issue - for 2 reasons:

1. They`re more likely to chew on `em than part pee on em.

2. Unless the wiring is damaged, which ins. testing would show, pee on a cable won`t cause it to trip. (not that I`ve tried this!)

KME

 

[nsbs]

Thanks guys

I will try to answer some of the questions that have been posed.

Although there is TP&N supply at the building, its split over 3 floors with each floor on a different phase. There are no servers in this office, but I will be installing one later in the year. There is only one server in the top floor office. But the 1st floor is currently unoccupied and has been isolated at the main switch in the main panel room in the basement. I had thought about comms updating during the night, but discounted that to be honest because its on a different phase etc. I am not sure if the server is connected directly to the MET

 

[Manator]

This board you installed is it a 3ph board? or is it fed from one of the phases.

 

[nsbs]

its a single phase CU, The whole floor is serviced from the same phase, and the other two floors serviced from the other phases. the main switches for the building are in the plant room, with a 35mm - 2 core & seperate CPC run to each of the floors to service there own CU,s. hope that makes sense. each floor has a large cleaner cupboard and there CU's are sited in there.

ian

 

[Manator]

Thanks nsbs, its just that as much info as can be provided helps to eliminate any possible causes, and leaves us with probable solutions we can share and debate. I am still not convinced with thermal or magnetic influences, but I am sure we can get to the bottom of it. The fact that only one breaker is problematic leans me towards a problem on the load side rather than the supply side.

 

[nsbs]

Thanks nsbs, its just that as much info as can be provided helps to eliminate any possible causes, and leaves us with probable solutions we can share and debate. I am still not convinced with thermal or magnetic influences, but I am sure we can get to the bottom of it. The fact that only one breaker is problematic leans me towards a problem on the load side rather than the supply side.

me too, but for the life of me i cant see what else there is to be done.

thanks. ian

 

[Manator]

I think we will have the answer before the weekend ends

 

[Andy™]

I think we will have the answer before the weekend ends

have you also got the lottery number for tomorrow night?

 

[Manator]

If only, no I think after the weekend we will have our answer, I did think about everything scientific about probable causes, and could not come to any conclusion other than that the power tower has a manufacture fault, it is also possible that the mcb has indeed tripped during the day but has gone unreported, only being noticed in the morning on start of work.

 

[androiduk2002]

Have you asked the customer if any other works happened 5 weeks ago ? As you say it worked ok for 7 weeks before any issues showed up , so what happened around that 7 week mark to cause the fault/change ?

 

[binky]

Just as a thought, if circuits and cable are all OK and test accordingly, how about changing MCB to type C - and see if that still trips? If it does it would kind of suggest something is happening somewhere???

Only other thing I can think of is loose connection somwhere either in CU or power pole, which don't always show under test. Somehow though I suspect you've already double checked every connection already

 

[Manator]

Yes binky he has said that he checked and even rewired the connections, I did however miss the part about other works being done 7 or so weeks ago, and that this fault reared up after that. Now the work involved would be my first investigation.

 

[Evans Electric]

Ah yes .......the curse of other trades methinks.