Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

How to connect this built-in oven properly?

788 views
Skip to first unread message

tin...@isbd.co.uk

unread,
Apr 19, 2012, 5:37:08 PM4/19/12
to
I am installing a new built-in oven and hit the usual problem of the
horrible connectors they always seem to come with. However this one
seems even worse than usual and I'm a bit stumped as to how to do a
proper job of connecting it.

I have a dedicated 32 amp MCB for the oven and (electric, induction)
hob and it has a 6sq mm cable running to the kitchen units.

However there is absolutely no way that one can connect 6sq mm cable
to the oven. The connection block is too small. The only cable that
I think I can reasonably connect to it is 2.5sq mm because the screw
terminals are such that only single, solid wire can be reliably
connected to it.

There's not enough room for any sort of termination on the cable.
I've tried and I have a wide variety of crimps in many shapes and
sizes, while you can terminate a cable with a 'ring' that will fit
under the screw the body of the crimp is too fat to fit through the
slot provided for the cable.

The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
is that it's not ideal for allowing the oven to be slid in and out and
it's not properly protected by the 32amp MCB.

What options do I have? I suppose I could have a local FCU for the
oven but it seems a bit overly complex and discrimination might be a
problem. I can't see *any* way to connect flexible cable to the oven
so I think it's stuck with at least a short length of 2.5sq mm. The
best I can do at present is to take the short length of 2.5sq mm to a
junction box (or terminal block) on the back of the oven and run a
more flexible cable from there. Can one get high temperature 2.5sq mm
cable with solid conductors that I could rate at 32 amps? That would
be be the neatest solution I guess.

Or maybe I should take the back off the oven and replace their crap
connector!

--
Chris Green

ARWadsworth

unread,
Apr 19, 2012, 6:13:38 PM4/19/12
to
tin...@isbd.co.uk wrote:
> I am installing a new built-in oven and hit the usual problem of the
> horrible connectors they always seem to come with


> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> is that it's not ideal for allowing the oven to be slid in and out and
> it's not properly protected by the 32amp MCB.

There is a very good chance the 2.5 T&E will be adequately protected by the
32A MCB.


However I now suspect that you have just unleashed the pillock from
Birmingham and we can now expect a load of ranting and bollocks about the
dangers (made up ones) of installing unfused spurs using 2.5 T&E on a 32A
circuit.

Our only hope is that he is in his straight jacket or that nursie has
heavily sedated him tonight to stop him going again out without permission
armed with a hair drier, hi viz jacket and a clipboard. (Last time he went
out it took 3 nurses to pry him off the speed camera he was making love to,
and then they had to presure wash the camera post clean after they had done
so)

PS Are you sure you cannot get 2.5m flex to fit (maybe with a bootlace
ferrule on the end of the flex)
--
Adam


A.Lee

unread,
Apr 19, 2012, 6:23:28 PM4/19/12
to
<tin...@isbd.co.uk> wrote:

> I am installing a new built-in oven and hit the usual problem of the
> horrible connectors they always seem to come with. However this one
> seems even worse than usual and I'm a bit stumped as to how to do a
> proper job of connecting it.
>
> I have a dedicated 32 amp MCB for the oven and (electric, induction)
> hob and it has a 6sq mm cable running to the kitchen units.
>
> However there is absolutely no way that one can connect 6sq mm cable
> to the oven. The connection block is too small. The only cable that
> I think I can reasonably connect to it is 2.5sq mm because the screw
> terminals are such that only single, solid wire can be reliably
> connected to it.

You need to separate the oven and hob, and have 2 switches, one for
each. I presume that 32A is enough for the both?

You spur off the 6mm cable with a 13A FCU to supply the oven. You can
run 4mm cable from a connector box/joint to the FCU if you cannot
terminate the 6mm in a normal FCU.
You dont need heat resistant cable, it doesnt get very hot at the back
of the oven. 2.5mm from the FCU to the oven is fine.

--
To reply by e-mail, change the ' + ' to 'plus'.

John Rumm

unread,
Apr 19, 2012, 7:15:16 PM4/19/12
to
On 19/04/2012 18:37, tin...@isbd.co.uk wrote:
> I am installing a new built-in oven and hit the usual problem of the
> horrible connectors they always seem to come with. However this one
> seems even worse than usual and I'm a bit stumped as to how to do a
> proper job of connecting it.
>
> I have a dedicated 32 amp MCB for the oven and (electric, induction)
> hob and it has a 6sq mm cable running to the kitchen units.

No need if it can only draw 2.3kw peak...

> However there is absolutely no way that one can connect 6sq mm cable
> to the oven. The connection block is too small. The only cable that
> I think I can reasonably connect to it is 2.5sq mm because the screw
> terminals are such that only single, solid wire can be reliably
> connected to it.
>
> There's not enough room for any sort of termination on the cable.
> I've tried and I have a wide variety of crimps in many shapes and
> sizes, while you can terminate a cable with a 'ring' that will fit
> under the screw the body of the crimp is too fat to fit through the
> slot provided for the cable.
>
> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> is that it's not ideal for allowing the oven to be slid in and out and
> it's not properly protected by the 32amp MCB.

It will be adequately protected by the 32A MCB - since in this
application you are only concerned about fault current protection.

> What options do I have? I suppose I could have a local FCU for the
> oven but it seems a bit overly complex and discrimination might be a

You could use a FCU. A 13A fuse would normally discriminate with a 32A
MCB upstream. However the chances of it ever operating are slim anyway.

> problem. I can't see *any* way to connect flexible cable to the oven
> so I think it's stuck with at least a short length of 2.5sq mm. The

If you can get 2.5mm^2 T&E into the terminals then you are likely to be
able to get a 2.5mm^2 heat resistant flex in there as well.

Note however if you were to fit a FCU, then you could reduce your flex
size to 1.5mm^2:

http://www.tlc-direct.co.uk/Products/CA1dot5TQ3slash50.html

> best I can do at present is to take the short length of 2.5sq mm to a
> junction box (or terminal block) on the back of the oven and run a
> more flexible cable from there. Can one get high temperature 2.5sq mm
> cable with solid conductors that I could rate at 32 amps? That would
> be be the neatest solution I guess.

Sounds like you are over thinking this one ;-)


--
Cheers,

John.

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

dennis@home

unread,
Apr 19, 2012, 7:27:15 PM4/19/12
to


"A.Lee" <alan@darkroom.+.com> wrote in message
news:1kitc5y.1wxi34k1qk69a4N%alan@darkroom.+.com...
This sounds much better than what ARW said.
I would go with what Mr Lee says.

js.b1

unread,
Apr 19, 2012, 8:04:19 PM4/19/12
to
On Apr 19, 6:37 pm, tinn...@isbd.co.uk wrote:
> I have a dedicated 32 amp MCB for the oven and (electric, induction)
> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate

Do you have something like an MK 32A cooker switch & 32A cord outlet
behind units, ie, originally for a conventional standalone electric
oven?

32A cord outlet supplies units...
Cord outlet -> 6mm FTE -> Jn-Box
... Jn-Box -> 6mm FTE -> hob
... Jn-Box -> 4mm FTE -> FCU -> 1.5mm Flex -> oven

2.3kW oven cable...
- 1.5mm Flex in Butyl / 90oC PVC
- Commonly used for storage heaters, immersions etc.

NkW hob cable...
- 6.0mm FTE - 60oC PVC (or 90oC 6242BH off Ebay)
- 6.0mm Flex - Butyl, High Temp PVC, Silicone (CPC, Farnell, Conrad)

TLC do a good range of junction boxes, 6mm cut length, 4mm cut length,
1.5mm high temp flex.

Check what spec cable the hob requires, that may make things a bit
more interesting.

Easy. Done. What you moan'n about. Your cat is eating the children
whilst you are thinking about it :-)))

ARWadsworth

unread,
Apr 19, 2012, 8:06:48 PM4/19/12
to
How do you know what I said? You claim you do not read my posts?

Have you invented a time machine so that you can see posts from the future?
That means you can claim that you were replying to a post of mine by seeing
another posters reply (by someone that you do not claim to blacklist) to my
post?

All it takes is someone that you have not blacklisted to reply to my post
with the words "dennis you are a bell end" to make the future come true

Any volunteers to make this post?

--
Adam


tin...@isbd.co.uk

unread,
Apr 19, 2012, 8:56:05 PM4/19/12
to
ARWadsworth <adamwa...@blueyonder.co.uk> wrote:
> tin...@isbd.co.uk wrote:
> > I am installing a new built-in oven and hit the usual problem of the
> > horrible connectors they always seem to come with
>
>
> > The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> > is that it's not ideal for allowing the oven to be slid in and out and
> > it's not properly protected by the 32amp MCB.
>
> There is a very good chance the 2.5 T&E will be adequately protected by the
> 32A MCB.
>
I was wondering if that might be so, it'll be a very short piece of
2.5 T&E.


[snip]
>
> PS Are you sure you cannot get 2.5m flex to fit (maybe with a bootlace
> ferrule on the end of the flex)

Pretty sure yes, for two reasons:-

The bootlace ferrule's "shoulder" almost certainly won't fit
through the space provided.

It won't locate safely under the screw terminal anyway, there's no
washer or anything. The only way one can make 2.5 T&E secure is
by pre-forming a U in the end of the wire and wrapping that round
the terminal screw.

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 19, 2012, 8:57:24 PM4/19/12
to
Yes, that is one option I suppose, just rather 'messy'. (It was a
possibility I mentioned)

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 19, 2012, 9:03:02 PM4/19/12
to
John Rumm <see.my.s...@nowhere.null> wrote:
> On 19/04/2012 18:37, tin...@isbd.co.uk wrote:
> > I am installing a new built-in oven and hit the usual problem of the
> > horrible connectors they always seem to come with. However this one
> > seems even worse than usual and I'm a bit stumped as to how to do a
> > proper job of connecting it.
> >
> > I have a dedicated 32 amp MCB for the oven and (electric, induction)
> > hob and it has a 6sq mm cable running to the kitchen units.
>
> No need if it can only draw 2.3kw peak...
>
The induction hob is 7.5kW (if we get the one we're thinking about
getting).

> > The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> > is that it's not ideal for allowing the oven to be slid in and out and
> > it's not properly protected by the 32amp MCB.
>
> It will be adequately protected by the 32A MCB - since in this
> application you are only concerned about fault current protection.
>
I did wonder about this, maybe I am being a bit paranoid! :-)

> > What options do I have? I suppose I could have a local FCU for the
> > oven but it seems a bit overly complex and discrimination might be a
>
> You could use a FCU. A 13A fuse would normally discriminate with a 32A
> MCB upstream. However the chances of it ever operating are slim anyway.
>
> > problem. I can't see *any* way to connect flexible cable to the oven
> > so I think it's stuck with at least a short length of 2.5sq mm. The
>
> If you can get 2.5mm^2 T&E into the terminals then you are likely to be
> able to get a 2.5mm^2 heat resistant flex in there as well.
>
Yes, but the terminals won't hold flex properly. They are just
round[ish] head screws with a plastic guard around them, no washers or
anything, if you put flex under them it just squeezes out. The only
reason that 2.5sq mm works is because a pre-formed U stays under the
screw.


> Note however if you were to fit a FCU, then you could reduce your flex
> size to 1.5mm^2:
>
> http://www.tlc-direct.co.uk/Products/CA1dot5TQ3slash50.html
>
> > best I can do at present is to take the short length of 2.5sq mm to a
> > junction box (or terminal block) on the back of the oven and run a
> > more flexible cable from there. Can one get high temperature 2.5sq mm
> > cable with solid conductors that I could rate at 32 amps? That would
> > be be the neatest solution I guess.
>
> Sounds like you are over thinking this one ;-)
>
Could be!

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 19, 2012, 9:08:20 PM4/19/12
to
js.b1 <js...@ntlworld.com> wrote:
> On Apr 19, 6:37 pm, tinn...@isbd.co.uk wrote:
> > I have a dedicated 32 amp MCB for the oven and (electric, induction)
> > The oven is rated 2.3kw so 2.5sq mm is perfectly adequate
>
> Do you have something like an MK 32A cooker switch & 32A cord outlet
> behind units, ie, originally for a conventional standalone electric
> oven?
>
> 32A cord outlet supplies units...
> Cord outlet -> 6mm FTE -> Jn-Box
> ... Jn-Box -> 6mm FTE -> hob
> ... Jn-Box -> 4mm FTE -> FCU -> 1.5mm Flex -> oven
>
> 2.3kW oven cable...
> - 1.5mm Flex in Butyl / 90oC PVC
> - Commonly used for storage heaters, immersions etc.
>
As I've said in a previous reply flexible cable simply won't stay
under the terminal screws on the oven.


> NkW hob cable...
> - 6.0mm FTE - 60oC PVC (or 90oC 6242BH off Ebay)
> - 6.0mm Flex - Butyl, High Temp PVC, Silicone (CPC, Farnell, Conrad)
>
I've actually got *loads* of 90 degree flex which I use on our boat.


Thanks for all the advice everyone. I think maybe I'll go for my
current solution:-

Short length of 2.5 T&E connected to oven terminals.

connecting to

Junction box fixed to oven

connecting to

6sqmm cable (which will be properly strain relieved at both ends)

--
Chris Green

rde42

unread,
Apr 19, 2012, 9:29:49 PM4/19/12
to
On Apr 19, 2:13 pm, "ARWadsworth" <adamwadswo...@blueyonder.co.uk>
wrote:
> tinn...@isbd.co.uk wrote:
> > I am installing a new built-inovenand hit the usual problem of the
> > horrible connectors they always seem to come with
> > Theovenis rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> > is that it's not ideal for allowing theovento be slid in and out and
> > it's not properly protected by the 32amp MCB.
>
> There is a very good chance the 2.5 T&E will be adequately protected by the
> 32A MCB.
>
> However I now suspect that you have just unleashed the pillock from
> Birmingham and we can now expect a load of ranting and bollocks about the
> dangers (made up ones) of installing unfused spurs using 2.5 T&E on a 32A
> circuit.
>
> Our only hope is that he is in his straight jacket or that nursie has
> heavily sedated him tonight to stop him going again out without permission
> armed with a hair drier, hi viz jacket and a clipboard. (Last time he went
> out it took 3 nurses to pry him off the speed camera he was making love to,
> and then they had to presure wash the camera post clean after they had done
> so)
>
> PS Are you sure you cannot get 2.5m flex to fit (maybe with a bootlace
> ferrule on the end of the flex)
> --
> Adam

The worst thing is that nursie probably has to clean his horrible
connector.

Mike Tomlinson

unread,
Apr 19, 2012, 9:25:51 PM4/19/12
to
En el artículo <43v569-...@chris.zbmc.eu>, tin...@isbd.co.uk
escribió:

>The oven is rated 2.3kw

At that rating, I would have expected it to come with a flex and plug to
be plugged into a standard 13A socket.

--
(\_/)
(='.'=)
(")_(")

John Rumm

unread,
Apr 19, 2012, 10:14:04 PM4/19/12
to
On 19/04/2012 22:03, tin...@isbd.co.uk wrote:
> John Rumm<see.my.s...@nowhere.null> wrote:
>> On 19/04/2012 18:37, tin...@isbd.co.uk wrote:
>>> I am installing a new built-in oven and hit the usual problem of the
>>> horrible connectors they always seem to come with. However this one
>>> seems even worse than usual and I'm a bit stumped as to how to do a
>>> proper job of connecting it.
>>>
>>> I have a dedicated 32 amp MCB for the oven and (electric, induction)
>>> hob and it has a 6sq mm cable running to the kitchen units.
>>
>> No need if it can only draw 2.3kw peak...
>>
> The induction hob is 7.5kW (if we get the one we're thinking about
> getting).

The hob won't be on the reduced capacity cable feeding the oven though...

Still if you do the sums:

7.5 + 2.3kW = 9,800W = theoretical max combined load of 42A

Now apply diversity: 10A + 30% of remaining 32A = 19.6A

(assuming no socket on the cooker switch, around 25A if there is).

In reality, not a big load.

>>> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
>>> is that it's not ideal for allowing the oven to be slid in and out and
>>> it's not properly protected by the 32amp MCB.
>>
>> It will be adequately protected by the 32A MCB - since in this
>> application you are only concerned about fault current protection.
>>
> I did wonder about this, maybe I am being a bit paranoid! :-)

When you think that you routinely run more powerful devices, on longer
and thinner flexes without concern...

>>> What options do I have? I suppose I could have a local FCU for the
>>> oven but it seems a bit overly complex and discrimination might be a
>>
>> You could use a FCU. A 13A fuse would normally discriminate with a 32A
>> MCB upstream. However the chances of it ever operating are slim anyway.
>>
>>> problem. I can't see *any* way to connect flexible cable to the oven
>>> so I think it's stuck with at least a short length of 2.5sq mm. The
>>
>> If you can get 2.5mm^2 T&E into the terminals then you are likely to be
>> able to get a 2.5mm^2 heat resistant flex in there as well.
>>
> Yes, but the terminals won't hold flex properly. They are just
> round[ish] head screws with a plastic guard around them, no washers or
> anything, if you put flex under them it just squeezes out. The only
> reason that 2.5sq mm works is because a pre-formed U stays under the
> screw.

Either twist the strands together and loop it round, or stick a bootlace
ferrule on the end, which will make it substantially the same as T&E
solid core to terminate. (on 1.5mm^2 there ought to be space to spare!)

>> Note however if you were to fit a FCU, then you could reduce your flex
>> size to 1.5mm^2:
>>
>> http://www.tlc-direct.co.uk/Products/CA1dot5TQ3slash50.html
>>
>>> best I can do at present is to take the short length of 2.5sq mm to a
>>> junction box (or terminal block) on the back of the oven and run a
>>> more flexible cable from there. Can one get high temperature 2.5sq mm
>>> cable with solid conductors that I could rate at 32 amps? That would
>>> be be the neatest solution I guess.
>>
>> Sounds like you are over thinking this one ;-)
>>
> Could be!


--

Rod Speed

unread,
Apr 19, 2012, 11:38:52 PM4/19/12
to


<tin...@isbd.co.uk> wrote in message news:65b669-...@chris.zbmc.eu...
I'd solder the stranded copper solid where it goes under the screw.

Corse you may not have access to a decent soldering iron.

tin...@isbd.co.uk

unread,
Apr 20, 2012, 8:00:06 AM4/20/12
to
John Rumm <see.my.s...@nowhere.null> wrote:
> >>> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
> >>> is that it's not ideal for allowing the oven to be slid in and out and
> >>> it's not properly protected by the 32amp MCB.
> >>
> >> It will be adequately protected by the 32A MCB - since in this
> >> application you are only concerned about fault current protection.
> >>
> > I did wonder about this, maybe I am being a bit paranoid! :-)
>
> When you think that you routinely run more powerful devices, on longer
> and thinner flexes without concern...
>
But not with just 32 amp circuit protection surely? There will be a
13 amp fuse in the plug to protect the cable.

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 20, 2012, 8:02:36 AM4/20/12
to
Rod Speed <rod.sp...@gmail.com> wrote:
> >>
> > Yes, but the terminals won't hold flex properly. They are just
> > round[ish] head screws with a plastic guard around them, no washers or
> > anything, if you put flex under them it just squeezes out. The only
> > reason
> > that 2.5sq mm works is because a pre-formed U stays under the screw.
>
> I'd solder the stranded copper solid where it goes under the screw.
>
> Corse you may not have access to a decent soldering iron.
>
I certainly could do that (and I'm reasonably skilled at soldering
having started my career in the electronics industry in the late
1960s). However there are lots of dire warnings hereabouts that the
solder creeps and the wire fails at the solder/non-solder interface.

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 20, 2012, 8:03:36 AM4/20/12
to
Mike Tomlinson <mi...@jasper.org.uk> wrote:
> En el artículo <43v569-...@chris.zbmc.eu>, tin...@isbd.co.uk
> escribió:
>
> >The oven is rated 2.3kw
>
> At that rating, I would have expected it to come with a flex and plug to
> be plugged into a standard 13A socket.
>
Quite, but it doesn't, it's designed for permament connection.

--
Chris Green

Adam Funk

unread,
Apr 20, 2012, 9:04:52 AM4/20/12
to
On 2012-04-19, ARWadsworth wrote:

> tin...@isbd.co.uk wrote:
>> I am installing a new built-in oven and hit the usual problem of the
>> horrible connectors they always seem to come with
>
>
>> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
>> is that it's not ideal for allowing the oven to be slid in and out and
>> it's not properly protected by the 32amp MCB.
>
> There is a very good chance the 2.5 T&E will be adequately protected by the
> 32A MCB.
>
>
> However I now suspect that you have just unleashed the pillock from
> Birmingham and we can now expect a load of ranting and bollocks about the
> dangers (made up ones) of installing unfused spurs using 2.5 T&E on a 32A
> circuit.

I'm not looking for an argument (or even a good session of
contradiction), but I was under the impression that 2.5 T&E wasn't
rated for 32 A in any situation, and that you ought to use
heat-resistant (rather than "ordinary") cable coming into an oven.
I'd like to know why those notions are wrong.

Andrew Gabriel

unread,
Apr 20, 2012, 9:33:35 AM4/20/12
to
In article <9vbm0h...@mid.individual.net>,
"Rod Speed" <rod.sp...@gmail.com> writes:
>
> I'd solder the stranded copper solid where it goes under the screw.

In which case, you'd generate an overheating join as the solder
creeps and the connection pressure is lost, solder creeps more,
etc, which is why this is forbidden.

If the terminal is not designed to take stranded copper,
the right way to do it is to crimp on a bootlace ferrule,
or an appropriate insulated crimp connector to mate with
the terminal (such as a ring, open ring, flat tab, i.e.
whatever matches the terminal design).

> Corse you may not have access to a decent soldering iron.

Soldering is also a skilled job (including knowing when you
should, or shouldn't use it). Crimping is less skilled
provided you have a reasonable quality ratchet crimper
and the right size crimps for it ans the conductors.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

Rod Speed

unread,
Apr 20, 2012, 9:57:18 AM4/20/12
to
Andrew Gabriel <and...@cucumber.demon.co.uk> wrote
> Rod Speed <rod.sp...@gmail.com> wrote

>> I'd solder the stranded copper solid where it goes under the screw.

> In which case, you'd generate an overheating join as the solder
> creeps and the connection pressure is lost, solder creeps more, etc,

Fantasy.

> which is why this is forbidden.

Fantasy.

> If the terminal is not designed to take stranded copper,
> the right way to do it is to crimp on a bootlace ferrule,

He already said that wont work and that
only a U in the solid conductor works.

> or an appropriate insulated crimp connector to mate
> with the terminal (such as a ring, open ring, flat tab,
> i.e. whatever matches the terminal design).

He did say that that isnt viable and I'm
too lazy to check why he said it wasn't.

>> Corse you may not have access to a decent soldering iron.

> Soldering is also a skilled job

Not when doing something as simple as that it isnt.

> (including knowing when you should, or shouldn't use it).

You clearly don't.

> Crimping is less skilled provided you have
> a reasonable quality ratchet crimper and
> the right size crimps for it ans the conductors.

He did say that that isnt viable and I'm
too lazy to check why he said it wasn't.

Rod Speed

unread,
Apr 20, 2012, 10:00:57 AM4/20/12
to
Rod Speed <rod.sp...@gmail.com> wrote
> Andrew Gabriel <and...@cucumber.demon.co.uk> wrote
>> Rod Speed <rod.sp...@gmail.com> wrote

>>> I'd solder the stranded copper solid where it goes under the screw.

>> In which case, you'd generate an overheating join

Fatasy. The connection isnt going to overhead any more
than it would with a solid conductor, in fact less than it
would with a solid conductor, just because there is rather
more contact area than with a solid conductor.

> as the solder creeps

That isnt going to creep in that situation.

tin...@isbd.co.uk

unread,
Apr 20, 2012, 9:59:03 AM4/20/12
to
Andrew Gabriel <and...@cucumber.demon.co.uk> wrote:
> > I'd solder the stranded copper solid where it goes under the screw.
>
> In which case, you'd generate an overheating join as the solder
> creeps and the connection pressure is lost, solder creeps more,
> etc, which is why this is forbidden.
>
Exactly my thoughts about tinning stranded wire to do this.

> If the terminal is not designed to take stranded copper,
> the right way to do it is to crimp on a bootlace ferrule,
> or an appropriate insulated crimp connector to mate with
> the terminal (such as a ring, open ring, flat tab, i.e.
> whatever matches the terminal design).
>
I don't think the terminal is 'designed' to take anything! :-(

Any sort of bootlace ferrule or insulated crimp on the end of the wire
to be connected prevents the wire from entering the terminal correctly
as there's a very narrow slot where the wire goes in. As I said I
think the only wire which can be reliably fixed under these terminal
screws is solid 2.5 sq mm (1.5 sq mm would probably be OK too).


> > Corse you may not have access to a decent soldering iron.
>
> Soldering is also a skilled job (including knowing when you
> should, or shouldn't use it). Crimping is less skilled
> provided you have a reasonable quality ratchet crimper
> and the right size crimps for it ans the conductors.
>
I have all sorts of insulated crimps and decent ratchet crimp tools, I
also have a load of ferrules and uninsulated crimps for sizes from 0.5
sq mm up to much larger than needed for this oven. I can solder too.
However this oven really stumped me. I hate to think what horrible
bodges get hidden inside oven housings with connection blocks like
this one.

--
Chris Green

Rod Speed

unread,
Apr 20, 2012, 10:27:53 AM4/20/12
to
<tin...@isbd.co.uk> wrote
Cant see that that is going to happen in this particular situation.

> and the wire fails at the solder/non-solder interface.

And that isnt going to happen either.

You arent using normal flex, just a few strands of copper
with the terminal not stopping the strands from escaping
from under the head as the screw is tightened.

And why, if the terminal does allow you bend the
copper in a U works fine, why cant you use a crimp
terminal on the wire, which goes where the U goes ?

John Rumm

unread,
Apr 20, 2012, 12:21:14 PM4/20/12
to
It would also not meet the wiring regs requirement of selection of
appropriate termination techniques, since it would not allow for the
creep of the solder under screw tension.

(it might be ok if you could actually solder the wire to the terminal)

A.Lee

unread,
Apr 20, 2012, 12:33:19 PM4/20/12
to
As Adam has pointed out to me, it is a known load, so for that oven,
there is no need for any overload protection.
Clearly, it is sensible to have o/l protection, but not necessary.

The wire input is at the bottom corner (generally) on built-in ovens, so
the cable is unlikely to get much over 30-40 degrees, hence no need for
heat resistant cable.

Alan.

sm_jamieson

unread,
Apr 20, 2012, 12:58:11 PM4/20/12
to
What make / model is the oven, and what does the installation guide say ?
I think you need to find from the manufacturer what cable it is designed to be connected with.
It would certainly be interesting to see what they say.

Simon.

tin...@isbd.co.uk

unread,
Apr 20, 2012, 1:13:28 PM4/20/12
to
Rod Speed <rod.sp...@gmail.com> wrote:
>
> And why, if the terminal does allow you bend the
> copper in a U works fine, why cant you use a crimp
> terminal on the wire, which goes where the U goes ?
>
It's difficult to describe the problem.

There's space around the terminal screw itself for the ring part of a
crimp terminal, *but* there isn't space for the insulator part of the
terminal. The terminal screw is in a sort of plastic box (with an
open top of course) and there's a narrow slot through which the cable
runs, this slot is wide enough (just) for a single 2.5sq mm conductor
with its insulation but not for anything fatter.

--
Chris Green

tin...@isbd.co.uk

unread,
Apr 20, 2012, 1:31:22 PM4/20/12
to
sm_jamieson <sm_ja...@hotmail.com> wrote:
> > >
> > I have all sorts of insulated crimps and decent ratchet crimp tools, I
> > also have a load of ferrules and uninsulated crimps for sizes from 0.5
> > sq mm up to much larger than needed for this oven. I can solder too.
> > However this oven really stumped me. I hate to think what horrible
> > bodges get hidden inside oven housings with connection blocks like
> > this one.
> >
>
> What make / model is the oven, and what does the installation guide say ?
> I think you need to find from the manufacturer what cable it is designed to be connected with.
> It would certainly be interesting to see what they say.
>
It's a Beko OIF22300. The connection information in the manual is as
follows:-

Open the terminal block with screwdriver

Pass the mains lead through the cable clamp under the terminal and
fix the mains lead to main body with supplied screw via cable
tightener.

Connect the cables according to the supplied diagram.

Close the cover of the terminal block once the wiring is complete.

I've put up a copy of the diagram at:-

http://zbmc.eu/beko001.jpg

The actual width of the box is around 25mm. The screws are quite
small, M4 I'd guess or even M3.5, the square box around each terminal
is probably 7 or 8mm square. The position of the blue wire on the
diagram is impossible as the slot through which it has to feed is
downwards rather than to the side.

I don't think contacting the manufacturer is really going to help all
that much.

I do have an acceptable connection now.

--
Chris Green

John Rumm

unread,
Apr 20, 2012, 1:40:21 PM4/20/12
to
On 20/04/2012 10:04, Adam Funk wrote:
> On 2012-04-19, ARWadsworth wrote:
>
>> tin...@isbd.co.uk wrote:
>>> I am installing a new built-in oven and hit the usual problem of the
>>> horrible connectors they always seem to come with
>>
>>
>>> The oven is rated 2.3kw so 2.5sq mm is perfectly adequate, the problem
>>> is that it's not ideal for allowing the oven to be slid in and out and
>>> it's not properly protected by the 32amp MCB.
>>
>> There is a very good chance the 2.5 T&E will be adequately protected by the
>> 32A MCB.
>>
>>
>> However I now suspect that you have just unleashed the pillock from
>> Birmingham and we can now expect a load of ranting and bollocks about the
>> dangers (made up ones) of installing unfused spurs using 2.5 T&E on a 32A
>> circuit.
>
> I'm not looking for an argument (or even a good session of
> contradiction), but I was under the impression that 2.5 T&E wasn't
> rated for 32 A in any situation,

You are right its not...

However in this case, we are not designing that part of the installation
to carry 32A, or anything close to it.

This is one of those situations where the responsibility for fault and
overload protection can be split. The fault protection must always be at
the origin of the circuit, and must in the event of a fault ensure that
the circuit is disconnected quickly enough to avoid (possibly further)
cable damage (5 secs typically on a fixed appliance). Overload
protection is also usually at the origin of the circuit and provided by
the same protective device. However in some cases the overload
protection can be placed elsewhere in the circuit, or for that matter
omitted altogehter, if it can be imposed by some other design criterion.

In this circumstance the fault protection for the 2.5mm^2 cable will be
adequately provided by the 32A MCB (much the same situation as applies
with a spur in 2.5mm^2 from a 32A radial wired in 4mm^2 for example).
There is however no need for explicit overload protection since in this
case it is imposed by the design i.e. the oven can only draw a maximum
of 10A, and there is no way for someone to unwittingly add additional
load to the end of its cable.

> and that you ought to use
> heat-resistant (rather than "ordinary") cable coming into an oven.

Ideally it should be - and high temperature flex would be the cable of
choice. However there are two common problems with that, its not always
easy to buy in the required size[1], and second it won't always fit in
the connector on the oven. This means that in practice many ovens /
cookers etc get wired in T&E. By selecting a size significantly in
excess of that indicated just by the current demand, you also allow for
significant de-rating due to the high ambient temperature. While it gets
warm in the oven enclosure, its unlikely to actually go above the
insulation temperature limit of the "normal" PVC.

[1] Most wholesalers have stock of 2.5mm^2 which will cover pretty much
any oven, however many do not carry the 4.0mm^2 that would be required
for many hobs.

> I'd like to know why those notions are wrong.

Your notions are not actually wrong as rules of thumb, its just there is
a little more design finesse one can apply to take into account the
specifics of the situation.

sm_jamieson

unread,
Apr 20, 2012, 1:54:40 PM4/20/12
to
Interesting.
I shall not buy a Beko oven.

I have a zanussi oven to fit soon, rating 5.3kW.
The manual, in the installation sections says:
... the appliance should be operated using at
least 6mm² twin core and earth PVC insulated multicore
cable.
I'll be interested to see what the terminals look like.

Simon.

tin...@isbd.co.uk

unread,
Apr 20, 2012, 1:57:30 PM4/20/12
to
John Rumm <see.my.s...@nowhere.null> wrote:
>
> In this circumstance the fault protection for the 2.5mm^2 cable will be
> adequately provided by the 32A MCB (much the same situation as applies
> with a spur in 2.5mm^2 from a 32A radial wired in 4mm^2 for example).
> There is however no need for explicit overload protection since in this
> case it is imposed by the design i.e. the oven can only draw a maximum
> of 10A, and there is no way for someone to unwittingly add additional
> load to the end of its cable.
>
Thanks for this detailed clarification, just what I needed to feel
that the way I have wired it is satisfactory.

--
Chris Green

Adam Funk

unread,
Apr 20, 2012, 2:17:08 PM4/20/12
to
On 2012-04-20, A.Lee wrote:

> Adam Funk <a24...@ducksburg.com> wrote:

>> I'm not looking for an argument (or even a good session of
>> contradiction), but I was under the impression that 2.5 T&E wasn't
>> rated for 32 A in any situation, and that you ought to use
>> heat-resistant (rather than "ordinary") cable coming into an oven.
>> I'd like to know why those notions are wrong.
>
> As Adam has pointed out to me, it is a known load, so for that oven,
> there is no need for any overload protection.
> Clearly, it is sensible to have o/l protection, but not necessary.

Right, but I was thinking more of fault protection (see John Rumm's
reply & my reply to that).


> The wire input is at the bottom corner (generally) on built-in ovens, so
> the cable is unlikely to get much over 30-40 degrees, hence no need for
> heat resistant cable.

Good point.

Adam Funk

unread,
Apr 20, 2012, 2:15:55 PM4/20/12
to
On 2012-04-20, John Rumm wrote:

> On 20/04/2012 10:04, Adam Funk wrote:

>> I'm not looking for an argument (or even a good session of
>> contradiction), but I was under the impression that 2.5 T&E wasn't
>> rated for 32 A in any situation,
>
> You are right its not...
>
> However in this case, we are not designing that part of the installation
> to carry 32A, or anything close to it.
>
> This is one of those situations where the responsibility for fault and
> overload protection can be split. The fault protection must always be at
> the origin of the circuit, and must in the event of a fault ensure that
> the circuit is disconnected quickly enough to avoid (possibly further)
> cable damage (5 secs typically on a fixed appliance). Overload
> protection is also usually at the origin of the circuit and provided by
> the same protective device. However in some cases the overload
> protection can be placed elsewhere in the circuit, or for that matter
> omitted altogehter, if it can be imposed by some other design criterion.
>
> In this circumstance the fault protection for the 2.5mm^2 cable will be
> adequately provided by the 32A MCB (much the same situation as applies
> with a spur in 2.5mm^2 from a 32A radial wired in 4mm^2 for example).
> There is however no need for explicit overload protection since in this
> case it is imposed by the design i.e. the oven can only draw a maximum
> of 10A, and there is no way for someone to unwittingly add additional
> load to the end of its cable.

Right, it's the fault protection that I was thinking of, but (a) I
didn't realize a 32 A MCB was low enough for fault protection on
2.5 mm^2, and (b) what about a fault inside the oven that takes it
over 10 A? Do we assume the oven has internal fault protection?


>> and that you ought to use
>> heat-resistant (rather than "ordinary") cable coming into an oven.
>
> Ideally it should be - and high temperature flex would be the cable of
> choice. However there are two common problems with that, its not always
> easy to buy in the required size[1], and second it won't always fit in
> the connector on the oven. This means that in practice many ovens /
> cookers etc get wired in T&E. By selecting a size significantly in
> excess of that indicated just by the current demand, you also allow for
> significant de-rating due to the high ambient temperature. While it gets
> warm in the oven enclosure, its unlikely to actually go above the
> insulation temperature limit of the "normal" PVC.

Good point.

> Your notions are not actually wrong as rules of thumb, its just there is
> a little more design finesse one can apply to take into account the
> specifics of the situation.

Thanks for the explanation.

ARWadsworth

unread,
Apr 20, 2012, 3:29:22 PM4/20/12
to
http://wiki.diyfaq.org.uk/index.php?title=MCB

and

http://wiki.diyfaq.org.uk/index.php?title=Calculating_A_Cable_Size

gives more detail.

It needs 160A to trip a 32A MCB quickly enough to meet fault protection
requirements.

But 160A is NOT the maximum fault current ir is a mimimum current. The fault
current will be much higher than 160A in most cases. A house close to the
substation could have a very low ELI and well over 1000A could pass through
the cable in the event of a fault (even on a 6A lighting circuit).

If the manufacturer of the oven specifies a 13A fuse needs to be fitted then
you must fit one.

Can you suggest a way that a 3kW oven could fail and allow greater than a
27A current (the maximum 2.5 T&E can carry) for any length of time?


--
Adam (But not Adam Funk)


ARWadsworth

unread,
Apr 20, 2012, 3:34:13 PM4/20/12
to
sm_jamieson wrote:
> On Friday, April 20, 2012 10:59:03 AM UTC+1, (unknown) wrote:
> > Andrew Gabriel <and...@cucumber.demon.co.uk> wrote:
> > > > I'd solder the stranded copper solid where it goes under the
> > > > screw.
> > >
> > > In which case, you'd generate an overheating join as the solder
> > > creeps and the connection pressure is lost, solder creeps more,
> > > etc, which is why this is forbidden.
> > >
> > Exactly my thoughts about tinning stranded wire to do this.
> >
> > > If the terminal is not designed to take stranded copper,
> > > the right way to do it is to crimp on a bootlace ferrule,
> > > or an appropriate insulated crimp connector to mate with
> > > the terminal (such as a ring, open ring, flat tab, i.e.
> > > whatever matches the terminal design).
> > >
> > I don't think the terminal is 'designed' to take anything! :-(
> >
> > Any sort of bootlace ferrule or insulated crimp on the end of the
> > wire to be connected prevents the wire from entering the terminal
> > correctly as there's a very narrow slot where the wire goes in. As
> > I said I think the only wire which can be reliably fixed under
> > these terminal screws is solid 2.5 sq mm


> (1.5 sq mm would probably be OK too).

When dennis reads that he will have a heart attack:-)

--
Adam


John Rumm

unread,
Apr 20, 2012, 3:35:37 PM4/20/12
to
Yup, fault protection (i.e. short circuits L to N or L to E) is a *very*
different ball game from overload protection. You are dealing with very
high instantaneous currents way in excess of the normal duty rating of
the conductors. Hence you need enough fault current to cause the
protective device to open quickly, and you need the cable conductors to
survive long enough to ensure that happens when subjected to adiabatic
heating (i.e. so fast you can't take into account any losses to the
surroundings).

So with a 32A MCB:

http://wiki.diyfaq.org.uk/images/d/d4/Curve-MCBTypeB.png

The table shows that a current of 160A or more is needed to cause it to
operate in its "instantaneous" mode (i.e. its magnetic rather than its
thermal response).

Say you take a fault scenario of the insulation on the oven element
breaking down completely and the element shorts to earth. Say this
presents a total Earth Loop Impedance of 0.8 ohms at the fault position.
That ELI would give rise a fault current of around 230/0.8 = 290A. So
enough to open the MCB in a nominal 0.1 secs. The weak link of the chain
with T&E is the earth conductor which is only 1.5mm^2 in this case. You
can check that it will survive long enough to clear the fault with the
adiabatic check:

s = sqrt( i^2 x t ) / k

Where s is the cross sectional area of earth conductor required, i the
fault current, t the time to clear the fault, and k a factor empirically
derived to factor in the physical response of the cable.

So you get:

s = sqrt( 290^2 x 0.1 ) / 115

(where 115 is the k factor for PVC cable, see table [1])

s = 0.8mm^2

So we known that the 1.5mm^2 in the T&E is >= s, hence it is adequate.

> (b) what about a fault inside the oven that takes it
> over 10 A?

Here is where you need to be careful with language... if you really mean
fault, then its covered by the above - as in my example.

However you also need to assess the risk of overload - i.e. a sustained
period drawing more current than the circuit was designed for.

In this case, while there are a number of fault current scenarios that
may occur in the oven, there are no realistic scenarios where the oven
could start drawing current significantly in excess of its design
current for an extended period of time. There is also no way for a user
to easily do anything that could add to its load.

(there are additional failure modes for things like ovens that result in
excess earth leakage, but only at very low current levels rather than
fault levels)

> Do we assume the oven has internal fault protection?

No, not in general. (Although as a general rule, if an appliance needs
additional protection beyond that expected for its supply conductors
then it should include it).

[1]
http://wiki.diyfaq.org.uk/index.php?title=Calculating_A_Cable_Size#Adiabatic_Check

Mike Tomlinson

unread,
Apr 20, 2012, 4:16:03 PM4/20/12
to
En el artículo <sph769-...@chris.zbmc.eu>, tin...@isbd.co.uk
escribió:

> However there are lots of dire warnings hereabouts that the
>solder creeps and the wire fails at the solder/non-solder interface.

Yes, it's a bad idea to tin wire ends where they will be secured by a
screw. The solder, being soft, "creeps" and a loose connection
eventually develops.

Woddles is talking shit, as usual. http://tinyurl.com/883xp7v

Mike Tomlinson

unread,
Apr 20, 2012, 4:20:18 PM4/20/12
to
En el artículo <65b669-...@chris.zbmc.eu>, tin...@isbd.co.uk
escribió:

>Yes, but the terminals won't hold flex properly. They are just
>round[ish] head screws with a plastic guard around them, no washers or
>anything, if you put flex under them it just squeezes out.

When that has happened to me in the past, I've added another washer
under the screw head. That prevents the stranded wire from squeezing
out as the additional washer doesn't turn with the screw head.

Mike Tomlinson

unread,
Apr 20, 2012, 4:20:58 PM4/20/12
to
En el artículo <12249059.764.1334930080342.JavaMail.geo-discussion-
forums@vbuo17>, sm_jamieson <sm_ja...@hotmail.com> escribió:

>I shall not buy a Beko oven.

They all come out of the same factory, so you could have the same
problem as the OP with any make.

js.b1

unread,
Apr 20, 2012, 4:38:46 PM4/20/12
to
If you want a high temperature 2.5mm solid core (Flat Twin & Earth).
- Item 220592367957 on www.ebay.co.uk
- 2.5mm LSF BS7211 6242BH, 90oC XLPE insulation.
- 99p a metre.

They have miss-listed it BS6004 (PVC).
However white sheath 6242 in Harmonised colours can only be BS7211
LSOH LSF LSZH 6242BH.

You do not need it though, think someone higher up just did the
adiabatic calculation on the minimum required CPC.


Absolutely do not "solder flex ends" to make them solid.
- It is against BS7671 17th.
- Solder cold flow reduces the screw clamping force
- Thermal cycling exacerbates solder cold flow
- Joint becomes slightly loose, solder melts, loose connection

Now, that loose connection will heat the supply conductors back along
the cable orange. Hopefully creating a short circuit where the cable/
flex bends and disconnecting the fault. Otherwise it can create a
fire. The cooker's internal wiring will be high temperature glass
fibre sheathed so will not usually be the the "fuse" in this
situation. Obviously the cable connector block and internal wiring may
still be a write-off.

This is exactly how a lecturer in 1998 stuffed a Miele Pyrolytic oven
that was over £1200. He tinned the copper fine strands "because that
is how they used to be supplied by manufacturers of appliances and you
fitted a plug yourself". The supply cable was charcoal, the units
scorched, the oven internal wiring stuffed.

Big ovens use metric M5 M6 M8 studs with the fine strand or 7-strand
conductors wrapped around between two washers squeezed together by a
nut, or a ratchet crimped ring terminal used instead. For the oven
design, use a bootlace ferrule - that provides a rigid cold-welded
assembly. Might be worth waiting to see what connection the hob
presents :-))

ARWadsworth

unread,
Apr 20, 2012, 4:39:59 PM4/20/12
to
tin...@isbd.co.uk wrote:
> sm_jamieson <sm_ja...@hotmail.com> wrote:
> > > >
> > > I have all sorts of insulated crimps and decent ratchet crimp
> > > tools, I also have a load of ferrules and uninsulated crimps for
> > > sizes from 0.5 sq mm up to much larger than needed for this oven.
> > > I can solder too. However this oven really stumped me. I hate to
> > > think what horrible bodges get hidden inside oven housings with
> > > connection blocks like this one.
> > >
> >
> > What make / model is the oven, and what does the installation guide
> > say ?
> > I think you need to find from the manufacturer what cable it is
> > designed to be connected with. It would certainly be interesting to
> > see what they say.
> >
> It's a Beko OIF22300. The connection information in the manual is as
> follows:-
>
> Open the terminal block with screwdriver
>
> Pass the mains lead through the cable clamp under the terminal and
> fix the mains lead to main body with supplied screw via cable
> tightener.
>
> Connect the cables according to the supplied diagram.
>
> Close the cover of the terminal block once the wiring is complete.
>
> I've put up a copy of the diagram at:-
>
> http://zbmc.eu/beko001.jpg
>

http://fiftyeggs.blob.core.windows.net/bekoupload/manuals/OIF22300.pdf

tells you to use HOVV-FG 3x1.55^2 cable with a mimimum fuse of 13A

Twisting the strands on the flex usually makes it solid enough to uses as
solid cable.

--
Adam


Rod Speed

unread,
Apr 20, 2012, 7:35:00 PM4/20/12
to
<tin...@isbd.co.uk> wrote
> Rod Speed <rod.sp...@gmail.com> wrote:

>> And why, if the terminal does allow you bend the
>> copper in a U works fine, why cant you use a crimp
>> terminal on the wire, which goes where the U goes ?

> It's difficult to describe the problem.

You did fine.

> There's space around the terminal screw itself for the ring part of a
> crimp
> terminal, *but* there isn't space for the insulator part of the terminal.

You can obviously cut the insulator part off trivially.

And they arent all insulated either.

> The terminal screw is in a sort of plastic box (with an open
> top of course) and there's a narrow slot through which the
> cable runs, this slot is wide enough (just) for a single 2.5sq
> mm conductor with its insulation but not for anything fatter.

Sounds like a real abortion of a design.

I'd solder the loop. That’s MUCH less likely to fail than using
a single solid conductor if you move the oven in and out much.

Corse you arent likely to move it much so a solid conductor should be fine.

Whats the brand and model so others can avoid it ?


Adam Funk

unread,
Apr 20, 2012, 8:33:25 PM4/20/12
to
On 2012-04-20, sm_jamieson wrote:

> On Friday, April 20, 2012 2:31:22 PM UTC+1, (unknown) wrote:

>> I've put up a copy of the diagram at:-
>>
>> http://zbmc.eu/beko001.jpg
>>
>> The actual width of the box is around 25mm. The screws are quite
>> small, M4 I'd guess or even M3.5, the square box around each terminal
>> is probably 7 or 8mm square. The position of the blue wire on the
>> diagram is impossible as the slot through which it has to feed is
>> downwards rather than to the side.
>>
>> I don't think contacting the manufacturer is really going to help all
>> that much.

> Interesting.
> I shall not buy a Beko oven.

I'd be wary of buying one now. That diagram makes it look very tight
--- can there be a good reason for making it so awkward?


> I have a zanussi oven to fit soon, rating 5.3kW.
> The manual, in the installation sections says:
> ... the appliance should be operated using at
> least 6mm² twin core and earth PVC insulated multicore
> cable.
> I'll be interested to see what the terminals look like.

I recently installed a Zanussi electric fan oven (lower power,
mentioned in a previous thread with helpful advice from the other
Adam). The terminal block itself was generously spaced & easy to use.
I only noticed 3 weird things about it:

1. I had to take the whole back panel off (with torx screws) to get to
the terminal block;

2. the terminal block itself had torx screws;

3. the internal live wire was orange instead of brown (the others
were blue & yellow/green, & I did manage to find "L", "N", & "E"
on the plastic by shining a bright light at it).

Adam Funk

unread,
Apr 20, 2012, 8:32:21 PM4/20/12
to
On 2012-04-20, ARWadsworth wrote:

> Adam Funk wrote:

>> Right, it's the fault protection that I was thinking of, but (a) I
>> didn't realize a 32 A MCB was low enough for fault protection on
>> 2.5 mm^2, and (b) what about a fault inside the oven that takes it
>> over 10 A? Do we assume the oven has internal fault protection?
>>
>
> http://wiki.diyfaq.org.uk/index.php?title=MCB
>
> and
>
> http://wiki.diyfaq.org.uk/index.php?title=Calculating_A_Cable_Size
>
> gives more detail.
>
> It needs 160A to trip a 32A MCB quickly enough to meet fault protection
> requirements.
>
> But 160A is NOT the maximum fault current ir is a mimimum current. The fault
> current will be much higher than 160A in most cases. A house close to the
> substation could have a very low ELI and well over 1000A could pass through
> the cable in the event of a fault (even on a 6A lighting circuit).

I see.

> If the manufacturer of the oven specifies a 13A fuse needs to be fitted then
> you must fit one.
>
> Can you suggest a way that a 3kW oven could fail and allow greater than a
> 27A current (the maximum 2.5 T&E can carry) for any length of time?

I don't have any experience in trying to sabotage ovens. ;-)

Thanks for the explanation.

Rod Speed

unread,
Apr 21, 2012, 1:19:32 AM4/21/12
to
Mike Tomlinson <mi...@jasper.org.uk>
> tin...@isbd.co.uk wrote

>> However there are lots of dire warnings hereabouts that the
>>solder creeps and the wire fails at the solder/non-solder interface.

> Yes, it's a bad idea to tin wire ends where they will be
> secured by a screw. The solder, being soft, "creeps"

No it does not in that situation because the copper stops that happening.

> and a loose connection eventually develops.

Fantasy.

Rod Speed

unread,
Apr 21, 2012, 1:22:11 AM4/21/12
to
Mike Tomlinson <mi...@jasper.org.uk> wrote
> sm_jamieson <sm_ja...@hotmail.com> wrote

>> I shall not buy a Beko oven.

> They all come out of the same factory,

Completely off with the fucking fairys, as always.

> so you could have the same problem as the OP with any make.

Very unlikely since few have commented on having seen that.

brass monkey

unread,
Apr 21, 2012, 1:23:25 AM4/21/12
to

"Rod Speed" <rod.sp...@gmail.com> wrote in message
news:9veg99...@mid.individual.net...
> Mike Tomlinson <mi...@jasper.org.uk>
>> tin...@isbd.co.uk wrote
>
>>> However there are lots of dire warnings hereabouts that the
>>>solder creeps and the wire fails at the solder/non-solder interface.
>
>> Yes, it's a bad idea to tin wire ends where they will be secured by a
>> screw. The solder, being soft, "creeps"
>
> No it does not in that situation because the copper stops that happening.

That's bollocks, our rodders. You obviously have no experience of the
problem. Loosen they do.


brass monkey

unread,
Apr 21, 2012, 1:24:51 AM4/21/12
to

"Rod Speed" <rod.sp...@gmail.com> wrote in message
news:9vege8...@mid.individual.net...

Rod Speed

unread,
Apr 21, 2012, 1:33:33 AM4/21/12
to
js.b1 <js...@ntlworld.com> wrote

> If you want a high temperature 2.5mm solid core (Flat Twin & Earth).
> - Item 220592367957 on www.ebay.co.uk
> - 2.5mm LSF BS7211 6242BH, 90oC XLPE insulation.
> - 99p a metre.

> They have miss-listed it BS6004 (PVC).
> However white sheath 6242 in Harmonised colours can only be BS7211
> LSOH LSF LSZH 6242BH.

> You do not need it though, think someone higher up just did the
> adiabatic calculation on the minimum required CPC.

> Absolutely do not "solder flex ends" to make them solid.
> - It is against BS7671 17th.
> - Solder cold flow reduces the screw clamping force

Not in that situation it doesn't, because
the copper bundle doesn't change shape.

> - Thermal cycling exacerbates solder cold flow

There wont be any thermal cycling.

> - Joint becomes slightly loose, solder melts, loose connection

Fantasy.

> Now, that loose connection will heat the supply
> conductors back along the cable orange.

Even sillier.

> Hopefully creating a short circuit where the
> cable/flex bends and disconnecting the fault.

Even sillier.

> Otherwise it can create a fire.

Even sillier.

> The cooker's internal wiring will be high temperature
> glass fibre sheathed so will not usually be the the
> "fuse" in this situation. Obviously the cable connector
> block and internal wiring may still be a write-off.

> This is exactly how a lecturer in 1998 stuffed
> a Miele Pyrolytic oven that was over £1200.

Have fun explaining how come my oven
has worked fine for 40 years that way.

I did that because I chose to put a normal 10A plug on it
so I could use the oven quickly when building the house
and never got around to wiring it in permanently.

> He tinned the copper fine strands "because that
> is how they used to be supplied by manufacturers
> of appliances and you fitted a plug yourself".

Have fun explaining why the manufacturers did it that way.

> The supply cable was charcoal, the units
> scorched, the oven internal wiring stuffed.

Have fun explaining why mine has lasted 40 years fine.

> Big ovens

We arent talking about big ovens, it's a puny oven.

Adam Funk

unread,
Apr 21, 2012, 3:19:43 PM4/21/12
to
On 2012-04-20, Mike Tomlinson wrote:

> En el artículo <sph769-...@chris.zbmc.eu>, tin...@isbd.co.uk
> escribió:
>
>> However there are lots of dire warnings hereabouts that the
>>solder creeps and the wire fails at the solder/non-solder interface.
>
> Yes, it's a bad idea to tin wire ends where they will be secured by a
> screw. The solder, being soft, "creeps" and a loose connection
> eventually develops.

I wonder if this idea comes from the custom of tinning the ends of
stranded speaker wire --- which is of course a completely a completely
different kettle of fish, since the current is much lower & the wires
are held in place by spring clips.

Adam Funk

unread,
Apr 22, 2012, 7:03:30 AM4/22/12
to
Interesting, thanks.

Jeff

unread,
May 4, 2015, 1:44:03 PM5/4/15
to
replying to tinnews , Jeff wrote:
> wrote:
>
> Thanks for this detailed clarification, just what I needed to feel
> that the way I have wired it is satisfactory.


Good grief don't know what all the fuss is about on this, there are two
options to wire a medium powered fitted hob and a low powered fitted oven.
The medium powered hob is hard wired via 6mm T&E to a 45A cooker control
switch that has a 13A socket outlet, then the 6mm cable runs to the hob,
the 2kW fitted oven is wired with 1.5mm 3 core flex (heat resistance can
be used but not mandatory).
Fit a 13A-fused plug the other end & plug it into the cooker control
switch 13A socket.

The second option is the same for the hob but the oven can be hard wired
to a 13A switch fused spur, the spur unit can be fed with 4mm/6mm cable
from either the cooker control switch or a cooker connection unit but make
sure you have a 45mm mounting box for room, then wire a 1.5mm/2.5mm to the
2kW oven.
You must have easy access to fuses & switches so do not place anything
switched or fused behind fitted ovens, duel cooker connections are also
now available that can take up to 10mm cables.
The golden rule is you must check the power in watts of each appliance &
consider the total wattage for the circuit.
It is paramount that you adhere to the appliance electrical wiring
instructions.
If it states an appliance must have a 13A fuse, then it generally means
there is no internal fuse in the oven to protect the fan & mains input
wiring & components, a 32A cooker circuit fuse would provide unsuitable
protection for the appliance.


--


ARW

unread,
May 4, 2015, 1:57:12 PM5/4/15
to
"Jeff" <caedfaa9ed1216d60e...@example.com> wrote in message
news:16020$554777a2$cf3aab60$20...@news.flashnewsgroups.com...
That's nice advice dear. Now fuck off back to year 2012.



--
Adam

John Rumm

unread,
May 4, 2015, 2:25:09 PM5/4/15
to
On 04/05/2015 14:44, Jeff wrote:
> replying to tinnews , Jeff wrote:
>> wrote:
>>
>> Thanks for this detailed clarification, just what I needed to feel
>> that the way I have wired it is satisfactory.
>
>
> Good grief don't know what all the fuss is about on this, there are two
> options to wire a medium powered fitted hob and a low powered fitted oven.

Bzzzzrt - sorry you did not read the question. Please try again.

J

unread,
May 5, 2015, 4:44:03 AM5/5/15
to
replying to ARW , J wrote:
> adamwadsworth wrote:
>
> "Jeff" <caedfaa9ed1216d60e...@example.com> wrote in
message
> news:16020$554777a2$cf3aab60$20...@news.flashnewsgroups.com...
> That's nice advice dear. Now fuck off back to year 2012.
> --
> Adam


No need to swear Adam your blood pressure is going high, back to 2012 you
need to go back further & debate it with Eve, she may get a cooker instead
of lighting fires to cook food, oh well Rome wasn't built in a day but it
was built a dam site quicker than you making your mind up about installing
a cooker.

--


0 new messages