On 24/11/2017 12:39, whisky-dave wrote:
On Thursday, 23 November 2017 21:19:50 UTC, John Rumm wrote:
On 23/11/2017 13:31, whisky-dave wrote:
The only thing you might encounter different is in IT labs
where the large quantities of switched mode PSUs would often
require that high integrity earthing be used as well since
there is often quite high earth leakage when in normal
operation as a result of all the input filters / suppressors.
I'm not sure if 40 or so PCs and dozens of SMPS would count.
Probably would.
So I wouldn't consider it general purpose as in home general
purpose.
its can be general purpose and have high integrity earthing - the two
are not exclusive.
But all the lab sockets did have a filter on them.
Yup, nothing to do with high integrity earthing though.
I wasn't aware they was a connection, and I don't think many homes
have such a filter anmd consoder it general purpose.
Lots of places use mains filters of various types for IT kit. They are
(mostly) useless to be fair.
If you want proper protection for IT stuff, then either a line
interactive or online UPS is the way to go.
Perhaps, but that ain't what they are, and the difference is
significant.
Not to me since we've told this riser has nothing to do with us and
if there is a problem we have to get someone in to sort it.
Well its very relevant to the conversation we have been having and to
some of the questions you asked. It may not be your responsibility to
know about these things as a part of the job, but that does not mean you
have to pretend to be ignorant of the actual details.
Its more likely in general. The design of most flat cables with
the earth in the middle makes it harder to create a LN short
without also creating a LE or NE one.
But every single one of our cables is round, never seen a flat one
used in the lab adn we have 100s of them, I have 50+ unused still
with the twist wore around them, 3 pin plug to IEC (mistakenly called
a kettle lead) and they are all round cables.
Those are flexes not cables. The cables form part of the fixed wiring
between the CU and sockets etc.
Flexes are far more vulnerable to L to N faults than cables, but do have
the advantage of dedicated fault protection from the plug fuse.
The memshield2 breakers are often 10kA rated BTW - so higher
breaking capacity than most domestic stuff.
I wonder why, I'd have thought the cable would have vapourised
long before 10kA .
Nothing to do with the cable - more to do with the "stiffness" of
the supply. A big low impedance mains feed close to a substation
can provide very significant fault current[1]. This can pose a
problem for the MCB because there is an upper limit to the current
flow it can successfully interrupt. A typical domestic MCB will
normally be rated for 6kA. So any fault current less than that it
should be able to open without sustaining damage or welding its
contacts together. More than that and it may fail to disconnect or
get destroyed while trying.
The memshield ones are aimed more at industrial use, and hence many
can cope with higher fault currents.
So this doesn't sound much lioke a genral purpose setup to me.
If you look at that photo of your CU, all the "Ring room nnn" circuits
are general purpose circuits. The ones labelled things like "Door spur",
"door bell", "security panel" etc are not general purpose circuits.
(Note that general purpose has specific meaning in this context)
[1] Although your observations on voltage drop may suggest "big"
and "low impedance" are not words one would use for your supply!
well that I don't know btu I expected a voltage drop as we exceeded
the 32A. I wasn't suprised that my 2KW heater only measured 1.6KW at
202V.
Indeed you would expect a drop, but the amount you got was rather large.
So it looks like what's been installed is what's needed in the
lab and up to standard so everythings OK.
With the exception of the loop impedance and voltage drop. The
voltage drop under load (combined with the voltage reduction
device) means you could damage equipment that is sensitive to
undervolt.
I don't believe we have such equipment, so I'm not particually
worried.
A typical bit of kit to suffer would be something with a compressor
driven by an induction motor. Some fridges / freezers can be quite
sensitive, either having insufficient torque to start and stalling, or
drawing too much current and overheating.
But we have had problems in the PCB room but as yet it;s
rare and hasn't been looked at closely.
So as expected everything seems OK
Its less gloomy than it first appeared, but based on what
currently appears to be the situation, not all "ok".
For me it is.
Well that's ok then... for you.
Threre must be some reason why a B C or D would be
installed.
Yup, B is general purpose and what you see in most domestic
installs (although I usually use type C on lighting circuits
to minimise nuisance trips on filament lamp failures).
That's interesting would the wiring in a typical lighting ring
320A or how about 40A .
The nominal rating for a lighting MCB would be 6A. So the fault
current required to "instantly" trip a type B is 30A, or 60A for a
type C
So would the lighting cable be OK at these currents ?
Yup.
We can do an adiabatic check to be su
S = sqrt( 60^2 x 0.1 ) / 115 = 0.16 mm^2
Since the smallest lighting cable used is 1mm^2 (and 1.5mm^2 being
common in larger installations), you would have loads of headroom.
60A of fault current should be easy to realise on most lighting
circuits if the total circuit impedance is 3.8 ohms or less.
Type C is often used with high inrush loads (large
transformers, induction motors, large banks of strip lights
etc).
I can imagine strip lights of the flourescant kind needing this,
any idea if its true of the recent LED tubes.
Generally less so... the strip lights (especially the older ones
with magnetic ballasts) can take quite a surge on startup and can
present quite and inductive load.
So I would assume if e replaced our 70 odd floursenant tubes with LED
version we wouldn't need to change the woring in the lihting circuit
of teh lab.
The existing wiring as far as the fittings would be fine. With some
replacement tubes you would need to alter the wiring in the lamp fitting
to wire the ballast out of circuit.
(Although you can get some "straight replacement" LED retrofit tubes
that will work with the ballast)
--
Cheers,
John.
/================================================== ===============\
| Internode Ltd -
http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/