Do dimmers still blow their triac with crap 60W bulbs (x3)?
The triac used to be quicker at terminally disconnecting faults before
the cpd did (somewhat less dramatically :-)

If so is it worth locally protecting the dimmer with a local fuse?
eg, grid K4501 + grid fuseholder with 1A fuse.

CPD on the final circuit is 6A Type B so 5*In or 30A will disconnect a
fault in 0.1s, conversely a local 1A fuse with 7*In would get that
down to 7A with luck. Last time I bothered with a dimmer was 25yrs ago
and that one buzzed loudly, made the brightness vary up & down
randomly and despite being almost never used just decided to stop
working.


If modern dimmers are ok, a plain K1501 dimmer is cheaper.

In article ,
"js.b1" writes:
Do dimmers still blow their triac with crap 60W bulbs (x3)?
The triac used to be quicker at terminally disconnecting faults before
the cpd did (somewhat less dramatically :-)

If so is it worth locally protecting the dimmer with a local fuse?
eg, grid K4501 + grid fuseholder with 1A fuse.

CPD on the final circuit is 6A Type B so 5*In or 30A will disconnect a
fault in 0.1s, conversely a local 1A fuse with 7*In would get that
down to 7A with luck. Last time I bothered with a dimmer was 25yrs ago
and that one buzzed loudly, made the brightness vary up & down
randomly and despite being almost never used just decided to stop
working.


If modern dimmers are ok, a plain K1501 dimmer is cheaper.

There's a trade-off in cheap dimmers, and it's between having a
low minimum load (which requires a small triac die with a small
holding current, but can then drive a single 40W lamp) and having
the capability to handle the high fault current resulting from a
flashover as a filament breaks (which requires a large triac die
to absorb the heat pulse without melting, but can't drive a load
as small as a single 40W lamp).

The other option is a more expensive dimmer circuit which uses a
hard firing technique for the triac (which requires an integrated
circuit rather than just the simple resistor, capacitor, diac
circuit), in which case the minimum load problem goes away, and a
large die triac can be used which will both survive the high fault
current for long enough to blow a fuse/mcb, and can also drive low
power loads.

I have an expensive remote controlled DIN rail mounted dimmer
which I use in one location. It probably would survive a flashover,
but as belt-and-braces protection, I added a £5 B3 MCB on the rail
next to it -- that will trip faster than most 1A fuses on a fault
current -- it can quench the arc mid-half-cycle, which many fuses
don't do, so you get faster disconnection time.

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

On Oct 9, 9:19*am, (Andrew Gabriel) wrote:
There's a trade-off in cheap dimmers, and it's between having a
low minimum load (which requires a small triac die with a small
holding current, but can then drive a single 40W lamp) and having
the capability to handle the high fault current resulting from a
flashover as a filament breaks (which requires a large triac die
to absorb the heat pulse without melting, but can't drive a load
as small as a single 40W lamp).

That is what I recall.

The other option is a more expensive dimmer circuit which uses a
hard firing technique for the triac (which requires an integrated
circuit rather than just the simple resistor, capacitor, diac
circuit), in which case the minimum load problem goes away, and a
large die triac can be used which will both survive the high fault
current for long enough to blow a fuse/mcb, and can also drive low
power loads.

PWM for lights?

I have an expensive remote controlled DIN rail mounted dimmer
which I use in one location. It probably would survive a flashover,
but as belt-and-braces protection, I added a £5 B3 MCB on the rail
next to it -- that will trip faster than most 1A fuses on a fault
current -- it can quench the arc mid-half-cycle, which many fuses
don't do, so you get faster disconnection time.

Fuses are slow.

So I gain nothing by a 1A fuse next to a K1501 (iQ) dimmer?

If so it might be better to do grid dimmer+switch with a picture
light, that way there is still functional lighting in the room if
after cpd reset the dimmer is sat dead in the water. It is for an
elderly relative who irritatingly wants a low light to draw curtains
or refuses to put it on - no doubt followed by falling over in the
room.

In article ,
"js.b1" writes:
On Oct 9, 9:19*am, (Andrew Gabriel) wrote:
There's a trade-off in cheap dimmers, and it's between having a
low minimum load (which requires a small triac die with a small
holding current, but can then drive a single 40W lamp) and having
the capability to handle the high fault current resulting from a
flashover as a filament breaks (which requires a large triac die
to absorb the heat pulse without melting, but can't drive a load
as small as a single 40W lamp).
That is what I recall.
The other option is a more expensive dimmer circuit which uses a
hard firing technique for the triac (which requires an integrated
circuit rather than just the simple resistor, capacitor, diac
circuit), in which case the minimum load problem goes away, and a
large die triac can be used which will both survive the high fault
current for long enough to blow a fuse/mcb, and can also drive low
power loads.
PWM for lights?

It's PWM identically in both cases. Hard firing the triac just
means that a gate current is supplied continuously when the triac
should be conducting, rather than just a pulse to start the
conduction and relying on the load current being than the triac's
holding current to maintain conduction to the zero crossing point.
With pulse firing and a small load, the current drops below the
triac's holding current well before the zero crossing point,
which causes premature cut-off before the zero-crossing point.
Furthermore, triacs are not manufactured to have well matched
holding current in each direction, as that's not a property of
a triac which is generally expolited. This means you get 50Hz
flicker at low light levels because of the uneven cutoff in
+ve and -ve going mains cycles, and it screws up the RC network
timing for the next half cycle by prematurely starting to charge
the capacitor before the zero crossing point, and that can make
the light output go very unstable, flickering badly.
Hard firing avoids the vaguries of the holding current and uses
much better timing circuits. This is easily integrated into a
small monolithic IC together with things like soft-on/soft-off
control and/or touch control. Mullard/Signetics/Philips were
producing such ICs over 30 years ago.

I have an expensive remote controlled DIN rail mounted dimmer
which I use in one location. It probably would survive a flashover,
but as belt-and-braces protection, I added a £5 B3 MCB on the rail
next to it -- that will trip faster than most 1A fuses on a fault
current -- it can quench the arc mid-half-cycle, which many fuses
don't do, so you get faster disconnection time.
Fuses are slow.
So I gain nothing by a 1A fuse next to a K1501 (iQ) dimmer?

Depends so much on the fuse, and on the prospective short circuit
current at the lampholder. It won't do any harm, except it might
needlessly blow every time the lamp blows.

If so it might be better to do grid dimmer+switch with a picture
light, that way there is still functional lighting in the room if
after cpd reset the dimmer is sat dead in the water. It is for an
elderly relative who irritatingly wants a low light to draw curtains
or refuses to put it on - no doubt followed by falling over in the
room.

Some assistive technology schemes do away with light switches
and use occupancy sensors instead, and this is because an elderly
person might otherwise get up in the middle of the night and not
switch on the lights, before moving around and falling over
something.

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

On Oct 9, 2:12*pm, (Andrew Gabriel) wrote:
Furthermore, triacs are not manufactured to have well matched
holding current in each direction, as that's not a property of
a triac which is generally expolited. This means you get 50Hz
flicker at low light levels because of the uneven cutoff in
+ve and -ve going mains cycles, and it screws up the RC network
timing for the next half cycle by prematurely starting to charge
the capacitor before the zero crossing point, and that can make
the light output go very unstable, flickering badly.

I had noticed bad flickering on the old dimmer when set low.

Depends so much on the fuse, and on the prospective short circuit
current at the lampholder. It won't do any harm, except it might
needlessly blow every time the lamp blows.

Indeed catch-22.

Some assistive technology schemes do away with light switches
and use occupancy sensors instead, and this is because an elderly
person might otherwise get up in the middle of the night and not
switch on the lights, before moving around and falling over
something.

I'll retain the wall light on a dimmer.
However I will add a ceiling occupancy sensor & picture light, easy
enough.