Mains Insulation 'Resistance', Multiple PCs and Load Problems Question

[David Longley]

I had an 'exchange' with a well educated friend the other day - he had
been told by a contractor who was putting in some cabling that they
could not put a room full of PCs on or several room full of PCs on
simultaneously, because there was a problem with computers' wires being
too close together and because of imperfect insulation between live and
neutral (I think this was the reference to the 'wires being too close
together', the cumulative effect would result in overload. (This was not
just an odd way of talking about the combined current exceeding the fuse
on the circuit, or something about surges).

I nearly choked on what I was eating and made some impulsively
regrettable remarks (he is a well educated friend after all). I'm not
sure I have put what he believed he was told accurately, but if anyone
has heard of anything *like* this, I would be grateful for an
elaboration. At the time I put it all down to too much wine and 'folk
physics'.

I've put hundreds of PCs into organisations, worked with electricians
etc and couldn't make any sense of what I heard at all.

 

[tony sayer]

I had an 'exchange' with a well educated friend the other day - he had
been told by a contractor who was putting in some cabling that they
could not put a room full of PCs on or several room full of PCs on
simultaneously, because there was a problem with computers' wires being
too close together and because of imperfect insulation between live and
neutral (I think this was the reference to the 'wires being too close
together', the cumulative effect would result in overload. (This was not
just an odd way of talking about the combined current exceeding the fuse
on the circuit, or something about surges).

I nearly choked on what I was eating and made some impulsively
regrettable remarks (he is a well educated friend after all). I'm not
sure I have put what he believed he was told accurately, but if anyone
has heard of anything *like* this, I would be grateful for an
elaboration. At the time I put it all down to too much wine and 'folk
physics'.

I've put hundreds of PCs into organisations, worked with electricians
etc and couldn't make any sense of what I heard at all.

There are new reg's out that limit the number of things using switch
mode power units, PC's and IT equipment for example, that have
capacitors in them that "connect" the live and earth together so that
there is some current flow. This can, if there are enough units on the
same line, cause RCD's to trip due to the current unbalance exceeding
the rating of the RCD.

There is also a shock risk in that if the earth came undone then the
"earth" wiring could have a potential that could pose a shock risk due
to the leakage in these machines.

I know of a radio station that had a lot of PC's connected to the same
ring main that had a 30 ma trip thereon which tripped quite often until
the system was upgraded.

What you heard was, I suppose, an "interpretation" of these regs!......

 

[Dave Liquorice]

I had an 'exchange' with a well educated friend the other day - he
had been told by a contractor who was putting in some cabling that
they could not put a room full of PCs on or several room full of PCs
on simultaneously,...

As in switch on simultaneously? The inrush is pretty phenonminal
especially if the monitors are CRTs, all those degauss circuits...

...will trip because there was a problem with computers' wires being
too close together and because of imperfect insulation between live
and neutral (I think this was the reference to the 'wires being too
close together', the cumulative effect would result in overload.

Any RCDs? You shouldn't put more than 10 Class 1 devices on a single
30mA RCD because the combined leakage from the suppression components
is likely to trip the RCD.

(This was not just an odd way of talking about the combined current
exceeding the fuse on the circuit, or something about surges).

Oh, well knocks the first one out then... B-)

 

[Jerry G.]

With the proper electrical installation, you should be able to put on as
many PC machines as you want. Do not use ground fault type outlets with
PC's and monitors, because the switching power supplies can cause the ground
fault outlet to trip to off.

The switching supplies can radiate back down the neutral of the AC line,
thus causing some over stressing of the electrical feed. The voltage
difference between the true ground and the neutral can go out of
specifications from this.

There were problems in many older buildings where many devices with
switching power supplies were installed. This ended up over heating the
power company's wiring, and feed transformers. The location where I am, had
this problem. This can all be corrected, but for a cost.

With a proper electrical installation, there should be no problems.

I have worked contracts where we put many servers and PC machines in one
room. I have some clients with as many as 20 machine in one large single
area. On a 15 amp breaker we would put 2 complete machines with monitors,
via a 1000 Watt UPS. Even the UPS has an internal switching power supply.

There are many companies with large offices that have many PC machines in
one area. I have seen large opened office areas in large centres, with a
few hundred machines. There are banks of routers, and sub systems, with a
number of management PC machines in smaller rooms off to the side of the
main areas. This is all workable. If not, there are many companies that
would not be working at large scales of operation today!

I think that this person who has this problem should get a new electrician
who knows about large scale installations.

--

Greetings,

Jerry Greenberg GLG Technologies GLG
=========================================
WebPage http://www.zoom-one.com
Electronics http://www.zoom-one.com/electron.htm
=========================================


I had an 'exchange' with a well educated friend the other day - he had
been told by a contractor who was putting in some cabling that they
could not put a room full of PCs on or several room full of PCs on
simultaneously, because there was a problem with computers' wires being
too close together and because of imperfect insulation between live and
neutral (I think this was the reference to the 'wires being too close
together', the cumulative effect would result in overload. (This was not
just an odd way of talking about the combined current exceeding the fuse
on the circuit, or something about surges).

I nearly choked on what I was eating and made some impulsively
regrettable remarks (he is a well educated friend after all). I'm not
sure I have put what he believed he was told accurately, but if anyone
has heard of anything *like* this, I would be grateful for an
elaboration. At the time I put it all down to too much wine and 'folk
physics'.

I've put hundreds of PCs into organisations, worked with electricians
etc and couldn't make any sense of what I heard at all.

 

[tony sayer]

With the proper electrical installation, you should be able to put on as
many PC machines as you want. Do not use ground fault type outlets with
PC's and monitors, because the switching power supplies can cause the ground
fault outlet to trip to off.

So lets see then. Here we have an office with quite a number of PC's and
other IT stuff and then someone plugs in some portable appliance that
has a fault where that person contacts the live of the mains so no RCD
protection then. Is this what you are suggesting?...

 

[BillR]

So lets see then. Here we have an office with quite a number of PC's
and other IT stuff and then someone plugs in some portable appliance
that
has a fault where that person contacts the live of the mains so no RCD
protection then. Is this what you are suggesting?...

The large installations I've worked at recent years had individual RCDs for
each desk ...

 

[Andrew Gabriel]

So lets see then. Here we have an office with quite a number of PC's and
other IT stuff and then someone plugs in some portable appliance that
has a fault where that person contacts the live of the mains so no RCD
protection then. Is this what you are suggesting?...

Original poster is asking a question about the UK I presume
but failed to say so and cross posted to international newsgroup,
so they got an answer for the US, which is completely wrong for
the UK.

 

[mike ring]

So lets see then. Here we have an office with quite a number of PC's
and other IT stuff and then someone plugs in some portable appliance
that has a fault where that person contacts the live of the mains so
no RCD protection then. Is this what you are suggesting?...

I don't know what the current thinking is, but in my working life I went
through more changes than I can remember with clean mains with walsall
outlets, dirty mains, RcD protected mains, backed up mains, ups mains,
that all I really know is the problem is not as simple as that.

I don't know if there is now peace at last, but considering how many
outages of important services there are, I doubt it.

Whenever I raised the problem for a new installation with the sparks and
users depts, as a humble project engineer, I always invoked codeword PYTHON

(put your tin hat on now)

Doing so.....

mike r

 

[mike ring]

PS. if you've ever sat in a busy office full of pcs, copiers, plugged in a
bit of kit that's just been deliverd to see if the little lights come on,
and heard all the fans spiralling down to a ghastly silence, you'll know
how to sneak out unobserved with the device under your coat while they're
all still wondering what occurred

m r

 

[ANt]

I know of a radio station that had a lot of PC's connected to the same

ring main that had a 30 ma trip thereon which tripped quite often until
the system was upgraded.

I once worked for Manweb (the electricity board as was) and we put a
new digital scanning facility (for digitising the network and
planning) into a building in Wrexham.....when someone had a No 49 from
the Coffee machine all the computers in the room tripped.

Ant.

 

[BigWallop]

I once worked for Manweb (the electricity board as was) and we put a
new digital scanning facility (for digitising the network and
planning) into a building in Wrexham.....when someone had a No 49 from
the Coffee machine all the computers in the room tripped.

Ant.

The number 49 must have been the extra strong coffee then ?

 

[Peter Parry]

I'm not sure I have put what he believed he was told accurately,

Probably not, however www.cda.org.uk/megab2/elecapps/pub142lo.pdf ,
www.cda.org.uk/megab2/elecapps/pub141lo.pdf and
www.iee.org/Publish/WireRegs/article.pdf might explain it all to
you.

 

[Mike Tomlinson]

I had an 'exchange' with a well educated friend the other day - he had
been told by a contractor who was putting in some cabling that they
could not put a room full of PCs on or several room full of PCs on
simultaneously, because there was a problem with computers' wires being
too close together and because of imperfect insulation between live and
neutral

It's not "imperfect insulation", it's surge protection in the switch-
mode power supplies used in computer equipment. These consist of
devices called MOVs (metal oxide varistors) connected between live and
earth and neutral and earth. Normally these devices only conduct when a
surge arrives down the wire; the surge is then carried away to earth.
The devices do have a very small leakage current, which means that in
installations where large numbers of computer equipment are connected to
the same supply, it can cause nuisance tripping of the RCD. Similarly,
if the earth were to become disconnected, the surge protection of all
the connected kit combined could cause the potential of metal-cased
equipment to rise high enough to deliver a nasty shock.

I've put hundreds of PCs into organisations, worked with electricians
etc and couldn't make any sense of what I heard at all.

We had the downstairs office refurbished recently here (uni in NW
England) and as we expected to install a lot of computer kit, the
electricians installed special sockets with two earth tags on the back,
and ran a second earth circuit. Can probably take a photo and put it up
somewhere if it's of interest.

M.

 

[Andrew Gabriel]

We had the downstairs office refurbished recently here (uni in NW
England) and as we expected to install a lot of computer kit, the
electricians installed special sockets with two earth tags on the back,
and ran a second earth circuit. Can probably take a photo and put it up
somewhere if it's of interest.

When a significant earth leakage current is expected, as in this
case, a high integrity earthing system is required. What you
describe is one way of doing this.

 

[George R. Gonzalez]

There could be several things going on:


Computers use switching mode power supplies, which have two quirks:


(1) They draw a big spike of current near the peak of the sine wave. The
average current may be a nice sedate 2 amps or so, but in actuality it may
be drawn as a tall spike, maybe 10 amps for 1/10th of each half-cycle.
Power loss goes up as the square of the current, so all the associated wires
and circuit breakers are going to heat up more than expected.

(2) Each power supply has rather studly line-filter capacitors, to keep
these and other spikes from exiting back thru the power lines. A few of
these on the same circuit may lead to excess ground currents, which might
confuse or trip a GFI.


I've seen this at home where my workshop has like 8 to 10 items of old
electoinic equipment. Put these all on one circuit and the GFI trips.

 

[Jeff Wiseman]

The number 49 must have been the extra strong coffee then ?

I thought that the computers were supposed to be working on the
question who's answer was 42, not 49...

 

[David Longley]

It's not "imperfect insulation", it's surge protection in the switch-
mode power supplies used in computer equipment. These consist of
devices called MOVs (metal oxide varistors) connected between live and
earth and neutral and earth. Normally these devices only conduct when a
surge arrives down the wire; the surge is then carried away to earth.
The devices do have a very small leakage current, which means that in
installations where large numbers of computer equipment are connected to
the same supply, it can cause nuisance tripping of the RCD. Similarly,
if the earth were to become disconnected, the surge protection of all
the connected kit combined could cause the potential of metal-cased
equipment to rise high enough to deliver a nasty shock.


We had the downstairs office refurbished recently here (uni in NW
England) and as we expected to install a lot of computer kit, the
electricians installed special sockets with two earth tags on the back,
and ran a second earth circuit. Can probably take a photo and put it up
somewhere if it's of interest.

M.

Many thanks.

One of the first responses said there were capacitors between the two
wires and that there were new regulations. This, plus your post, plus a
little bit of natural folk physics on the part of my friend, provides a
pretty good idea of what was probably being said or alluded to.

If you have a photo conveniently to hand, by all means send it to me as
an e-mail. If you have one of the MOV inside a PSU, please let me have
one of those too.

Thanks again to you all.