I'm replacing the spotlight in my kitchen (mains type -- no
transformer).

It should be a straight swap out of the existing unit -- but I have
forgotten which wires go where. On the old unit only the L and N
connectors were used. The light can be switched via a switch on the
kitchen wall or via a switch on the hall wall.

Wires dangling are as follows:

2 x red wires crimped together
1 x 3-core wire (I think) with single red wire used, other wires taped
off.

Which of these wires go to the L and which to N?

Thanks
Bruce

wrote:

It should be a straight swap out of the existing unit -- but I have
forgotten which wires go where. On the old unit only the L and N
connectors were used. The light can be switched via a switch on the
kitchen wall or via a switch on the hall wall.

Well, next time you'll remember to label the wires as you disconnect
them, won't you ;-)

Wires dangling are as follows:

2 x red wires crimped together
1 x 3-core wire (I think) with single red wire used, other wires taped
off.

Which of these wires go to the L and which to N?

Ass-U-ming your memory's correct, and there were only two connections
made... zerothly, there should've been three - what about the earth!?
Maybe you've not mentioned that one for simplicity's sake. Maybe you
don't have one - in which case it's time for a rewire.

So, firstly then,(or secondly if we're using awk or Pascal rather than C
or Python...) if these two really are the only non-earth incomers, then
one is neutral and one is switched-live. A moment with a volt-stick,
neon screwdriver, or meter will tell you which is which: one won't light
up regardless of switch position, the other will light your "hello have
you got volts" device when the switches are in one of their On
positions. That's the one you want connected to your L, the other one to
N. (In sheer "will it produce light" terms, it wouldn't matter which way
round the two are connected; for safety and decency, though, we try to
keep L correctly identified, so that (for example) spotlight bases have
the N on the more deeply-recessed base spring, while the skirt is
connected to neutral. Unless we're Continentals, in which case we're too
Hard for any of that pootling about).

Neither the details you've given (all-reds used, "crimped" together -
maybe you mean just twisted round each other? or do you really mean two
wires stuffed, against all the uses of nature, into a single
barrel/bootlace crimp!?) nor your apparent combination of reasoning +
pre-existing knowledge (it's not exactly hard to work out the stuff in
the previous paragraph of mine) fill me with confidence that you're a
Competent Person (in the actual, rather than Part-Pee, sense...) Are you
sure you're not a John Prescott troll?

"Stefek Zaba" wrote
| (In sheer "will it produce light" terms, it wouldn't matter
| which way round the two are connected; for safety and decency,
| though, we try to keep L correctly identified, so that (for
| example) spotlight bases have the N on the more deeply-recessed
| base spring, while the skirt is connected to neutral.

Shall we try that one again after another cup of coffee ...

Owain

Owain wrote:
"Stefek Zaba" wrote
| (In sheer "will it produce light" terms, it wouldn't matter
| which way round the two are connected; for safety and decency,
| though, we try to keep L correctly identified, so that (for
| example) spotlight bases have the N on the more deeply-recessed
L.
I *definitely* wrote L.
Some evil bit of line noise corrupted into an N, *and* automagically
fixed up the link-level and IP-level checksums.
Honest.
| base spring, while the skirt is connected to neutral.

Shall we try that one again after another cup of coffee ...

Owain

15-love ;-)

wrote:


It should be a straight swap out of the existing unit -- but I have
forgotten which wires go where. On the old unit only the L and N
connectors were used. The light can be switched via a switch on the
kitchen wall or via a switch on the hall wall.

Stefek Zaba wrote:

So, firstly then,(or secondly if we're using awk or Pascal rather than C
or Python...) if these two really are the only non-earth incomers, then

snippings

Yup, wot he said....



and, can I have a pint of whatever you are on please?


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

John Rumm wrote:


and, can I have a pint of whatever you are on please?

Wot I'm 'on', matey, is lack of sleep :-(

Y'see, my automagic mail handling (spam filtration - for an entire
short-name domain, dang it; and mail-list recognition) runs on a
d-i-y'ed box of preloved bits in a corner of the cubie at work - a PIII
450MHz or some such, quietly humming away running OpenBSD, which is a
low-resource-consumption excellent server platform for running the
mixture of procmail, an absolutely wonderful Bayesian-multi-word spam
recogniser which takes its name from Peter Seller's B&W WMD fillum, and
Python wot is my "IMAP semi-remote manipulation" mailhandling.

When I say 'quietly humming away', that's what it did for a year or more.

At the weekend, our site people announced a site-wide power shutdown,
and enjoined us to shut down all 'puter kit (which would otherwise
suffer cold-coulomb-turkey) on Friday. Which I did.

Bizarrely, the server woke up all on its ownsome a few minutes later,
and started beeping repeatedly (short beeps, on-off cycles taking oh
maybe 0.6s). The mobo is an Asus P2B-S with an Award bios.

As this was Friday afternoon, I yanked the powercord out and pretended
it hadn't happened. Didn't get into the office till the afternoon today,
after working from home through the morning.

Turned on the highly-reliable server. Beep beep beep beep beep
(actually, first a "long, two shorts" beep, then the repeated pulsing
beeps). Tried to find beep code cheatsheat on t'Interweb, and figure out
wth might be going on. 'One long two shorts is a vidcard fault; repeated
short beeps is memory' seemed to be the consensus vote. But why would
this happen after poweroff?

Went round reseating relevant components. Thought it possible that the
CMOS battery was out to lunch (would explain working OK while powered
up, and a subsequent loss of marbles). Put in new battery. Beep beep
beep beep beep etc. Another round of reseating, reducing config to
minimal, replacing AGP vidcard with old PCI one (which did get rid of
the one-long-two-shorts, at least). But still no booting.

Ended up pulling the two SCSI drives out of that box, and fitting them
into a preloved 300MHz PIII box in t'lab. One trial boot with the OBSD
CD to establish hardware support, one hunt for a better-working LAN
card, eventually succesful. Usual skin-threatening manouvers to fit HDs,
and a hunt for a little more RAM, and and and. Finally booted up off the
old drives (thank the Pope for OpenBSD's insistence on *not* rebuilding
kernels to suit local hardware!). Machine started chewing on 2,500
accumulated email messages, mainly offering *soft* pills to make
genitalia go *hard*. Ahh, the mysteries of marketing...

Leave it all running in t'lab. Arrive home half midnight, log in to
check progress. Discover the site's NFS server where the procmail
scripts live has just issued a Stale NFS Handle, so the mail's being
processed much more quickly than usual :-) by being dumped, unfiltered,
into the default mbox :-( Abort run. umount the NFS server. Try to
remount: refused. Scratch head. Discover this is what's been causing
some historical weirdness, as it happens. Go down to have late supper.
Return. Mmm, NFS server now playing ball again. Pull the 350 or so
unfiltered messages out of the backup box, onto home machine through
Thunderbird mbox 'import' (well, just placing it in the Local Folders
directory), and pushing them all back to the IMAP inbox. Rerun batch;
it's still going. Well, it's only half-two in the morning...

Aaah, puters. You gotta love 'em....

Still want a pint of today's hassles, John? ;-)

Stefek Zaba wrote:

Wot I'm 'on', matey, is lack of sleep :-(

I know that one ;-)

When I say 'quietly humming away', that's what it did for a year or more.

At the weekend, our site people announced a site-wide power shutdown,
and enjoined us to shut down all 'puter kit (which would otherwise
suffer cold-coulomb-turkey) on Friday. Which I did.

Hmm, reminds me why I hate turning mine off!

maybe 0.6s). The mobo is an Asus P2B-S with an Award bios.

Oh, got one of those in my main development platform at the mo...
getting a bit long in the tooth (the PIII 850 is the main problem), but
it is rock solid reliable, which is handy when I am playing "hunt the
obscure cause of data corruption on an ultra SCSI LVD hard drive" on my
more "fun" computer...

As this was Friday afternoon, I yanked the powercord out and pretended
it hadn't happened. Didn't get into the office till the afternoon today,
after working from home through the morning.

Execute the ostrich algorithm ;-)

Turned on the highly-reliable server. Beep beep beep beep beep
(actually, first a "long, two shorts" beep, then the repeated pulsing
beeps). Tried to find beep code cheatsheat on t'Interweb, and figure out
wth might be going on. 'One long two shorts is a vidcard fault; repeated
short beeps is memory' seemed to be the consensus vote. But why would
this happen after poweroff?

Thermal cycle lifting a card out a slot I find is the usual culprit...
el cheapo software modems being No. 1 cause, followed by the dodgy
looking double contact row AGP critters.

Went round reseating relevant components. Thought it possible that the
CMOS battery was out to lunch (would explain working OK while powered
up, and a subsequent loss of marbles). Put in new battery. Beep beep

CR2032?

beep beep beep etc. Another round of reseating, reducing config to
minimal, replacing AGP vidcard with old PCI one (which did get rid of
the one-long-two-shorts, at least). But still no booting.

Not showing any sign of the bulbous capacitor by chance? Had a spate of
those recently causing random reboots, non starts etc. At least they are
easy to spot, just look for the electrolytic with a domed lid ;-)

Ended up pulling the two SCSI drives out of that box, and fitting them
into a preloved 300MHz PIII box in t'lab. One trial boot with the OBSD
CD to establish hardware support, one hunt for a better-working LAN

;-) I think the last batch of 10/100 bargain basement cards I bought
from ebuyer was "one economy too far" (well what do you expect for
£1.80!) Work fine on Win2K up, but lie to you on older platforms (yes
drivers loaded and working gov..... no you can't look at the network)

accumulated email messages, mainly offering *soft* pills to make
genitalia go *hard*. Ahh, the mysteries of marketing...

That one makes me wonder, I get the occasional email offering a larger
member.. which begs the question, how do they know what is "larger"? Is
there some secret government database somewhere with these statistics?
Perhaps this is what Flunket was going on about with his "biometric ID".
Hey that could even be a new chat up line... "want to come over to my
place and see my ID?"

Leave it all running in t'lab. Arrive home half midnight, log in to
check progress. Discover the site's NFS server where the procmail
scripts live has just issued a Stale NFS Handle, so the mail's being
processed much more quickly than usual :-) by being dumped, unfiltered,
into the default mbox :-( Abort run. umount the NFS server. Try to

Not quite the BOFH solution for high speed mail tossing to /dev/null
then ;-)

it's still going. Well, it's only half-two in the morning...

Having an early night then? (Wuss... not even walked the dog yet! and I
have got another quotation to sort and some dodgy Javascript (hey just
thought, I can go try the dalmatian in the snow, now you see her, now
you don't...)) (hope you were counting parenthesise there)

Aaah, puters. You gotta love 'em....

Indeed....

Still want a pint of today's hassles, John? ;-)

Na, sounds too much like "normal", was hoping for something exotic ;-)



--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

John Rumm wrote:

Oh, got one of those in my main development platform at the mo...
getting a bit long in the tooth (the PIII 850 is the main problem), but
it is rock solid reliable, which is handy when I am playing "hunt the
obscure cause of data corruption on an ultra SCSI LVD hard drive" on my
more "fun" computer...

Same here - and the bargain-bucket LAN cards are the Achilees Heel, as
you say...

Thermal cycle lifting a card out a slot I find is the usual culprit...
el cheapo software modems being No. 1 cause, followed by the dodgy
looking double contact row AGP critters.

Yeah, 's wot I suspecturelated. But everything's had at least one lift
out, and the mobo's down to minimal complement now...

CR2032?

right first time ;-) Its replacement reads 3.2V offload, but was no
spring chicken, so I will when the tuits come in try with a fresh CMOS
battery (but it's most unlikely to be the cause...)

Not showing any sign of the bulbous capacitor by chance? Had a spate of
those recently causing random reboots, non starts etc. At least they are
easy to spot, just look for the electrolytic with a domed lid ;-)

Yes, all too likely - could also be the PSU dropping out of decent
regulation (just because the word "power" occurs in both "PSU" and
"power down")... Still with my teeth in "get the server working again",
I was going to raid one of the other boxes for its ATX PSU, when I
realised it'd be simpler to bring the HDs to the previously-working box.

;-) I think the last batch of 10/100 bargain basement cards I bought
from ebuyer was "one economy too far" (well what do you expect for
£1.80!) Work fine on Win2K up, but lie to you on older platforms (yes
drivers loaded and working gov..... no you can't look at the network)

That's the one. Today/yesterday's symptoms were "de0" initialising fine,
having a little conversation with the DHCP server, and then hanging the
OBSD install process after a "unexpected interrupt: 7" message...
whereas the other LAN i/f I scrounged up (the 'minor' function on a
peculiar SCSI/LAN hybrid card!) uses the le driver, and only craps out
with "le1: transmitter disabled" once an hour or so ;-(

Perhaps this is what Flunket was going on about with his "biometric ID".

Most likely. Wonder if he has any marks as Distinctive as the youngest
Jackson brother has (allegedly)?

Having an early night then?

Yup!

(hope you were counting parentheses there)

Always! - even when not writing bitsolisp... Though I'm on the side of
considering the ones embedded in smileys as legit parens - which in
fancypants mailers/newsreaders like Thunderboid which like to replace
them with Cute Little Icons makes the counting go to cock.

Na, sounds too much like "normal", was hoping for something exotic ;-)

Oh well, there's always tomorrow!

G'night - Stefek

Stefek Zaba wrote:

(hope you were counting parentheses there)


Always! - even when not writing bitsolisp... Though I'm on the side of
considering the ones embedded in smileys as legit parens - which in
fancypants mailers/newsreaders like Thunderboid which like to replace
them with Cute Little Icons makes the counting go to cock.

I had a conversation with SWMBO about that, she said it was "not proper"
to "dual purpose" a smiley as a closing bracket with such gay abandon...
I was not in agreement (personally I think it falls into the "object
reuse" argument ;-)


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

"Stefek Zaba" wrote
| John Rumm wrote:
| Wot I'm 'on', matey, is lack of sleep :-(
| Y'see, my automagic mail handling (spam filtration - for an
| entire short-name domain, dang it; and mail-list recognition)
| runs on a d-i-y'ed box of preloved

nice term :-)

snip 70 lines of woe

| Aaah, puters. You gotta love 'em....

I pay Claranet £24 a year for BorderScout antispam and antivirus on all
inbound and outbound mail.

I know I /could/ do it myself (although actually I probably couldn't)

Owain