Here's a letter I plan to send to Danlers. I just wondered if anyone
here could cast any light on the problem before I send the letter.

We have a problem with the TWSW twilight switch which has us completely
baffled. We had two of these items for a while. We got a replacement
from CPC for the first one thinking that it was faulty, but didnt
immediately send it back.

Initially the first unit was installed to replace an existing item
(bought from RS; expensive) that has been in place for many years but
had failed. The location is on a north facing wall about 15ft above the
ground. The unit installed normally and responded to the change of
light. We were able to set up the Lux threshold correctly. However, the
next morning the lighting was on in daylight. On examination the TWSWs
LED was lit, meaning that it should be holding the lighting off. We
tried various things over the next few days but the problem persisted:
the unit would still power the lighting after dawn, and would continue
to power it all day.

The replacement unit was initially set up in a test rig on the bench,
where it worked perfectly. We installed it in place of the first unit
and the same fault occurred. However the fault was intermittent, and
some mornings the lights would be switched off correctly. Wondering if
the lighting load was to blame we tried the unit with a temporary load
of a 100W tungsten lamp but with the unit still on the wall. The fault
persisted. Incidentally the normal load is three 70W sodium lamps, six
10W LEDs, a 1A 12V psu, and four 17W CFLs.

We took the second unit off the wall in case the fixings had distorted
the case and caused the fault but it made no difference. Concerned that
the load might have damaged the units we tested them both on the bench
and they both then worked perfectly, although we only took them through
a few cycles, using the (very bright) room lights to simulate daylight.
Without the lights the room is pitch black.

I then had the second unit on test indoors, looking out through an
east-facing window so it could €˜see daylight. It worked fine for four
days and nights.

Everything seems to point to a wiring fault, except that when the fault
condition is present the units LEDs are lit yet the switched line is
still live. We are aware of the 30 second delay function. What could
cause the LED to be lit yet leave the switched live output still
permanently live? After the unit has failed to operate in this way we
have found that brief power-cycling doesnt make it work normally. When
the power is restored it does the 30 sec thing but then the LED is on
and so is the lighting. To get the unit to turn the lights off in
daylight it seems to be necessary to leave it unpowered all day, and
then power it before the light starts to fade. It will then turn the
lighting off after 30 secs, and will turn it on again at dusk, but will
sometimes malfunction as described at dawn. We did have a good run of
three mornings with the second unit, and I thought the problem had
somehow fixed itself, but then the fault started to occur again.

At first we thought the problem might be caused by the very low outdoor
temperatures weve had here, but it persisted even when the weather
improved.

We just cant think of anything else to try. The units seem to work
anywhere except on the north-facing wall, where they work sometimes.
Have you had any reports of problems caused by strong RF fields? Its a
long shot but we have a low powered local radio transmitter 2km away to
the north.

Bill

On Friday, 20 April 2018 02:19:23 UTC+1, Bill Wright wrote:
Here's a letter I plan to send to Danlers. I just wondered if anyone
here could cast any light on the problem before I send the letter.

We have a problem with the TWSW twilight switch which has us completely
baffled. We had two of these items for a while. We got a replacement
from CPC for the first one thinking that it was faulty, but didnt
immediately send it back.

Initially the first unit was installed to replace an existing item
(bought from RS; expensive) that has been in place for many years but
had failed. The location is on a north facing wall about 15ft above the
ground. The unit installed normally and responded to the change of
light. We were able to set up the Lux threshold correctly. However, the
next morning the lighting was on in daylight. On examination the TWSWs
LED was lit, meaning that it should be holding the lighting off. We
tried various things over the next few days but the problem persisted:
the unit would still power the lighting after dawn, and would continue
to power it all day.

The replacement unit was initially set up in a test rig on the bench,
where it worked perfectly. We installed it in place of the first unit
and the same fault occurred. However the fault was intermittent, and
some mornings the lights would be switched off correctly. Wondering if
the lighting load was to blame we tried the unit with a temporary load
of a 100W tungsten lamp but with the unit still on the wall. The fault
persisted. Incidentally the normal load is three 70W sodium lamps, six
10W LEDs, a 1A 12V psu, and four 17W CFLs.

We took the second unit off the wall in case the fixings had distorted
the case and caused the fault but it made no difference. Concerned that
the load might have damaged the units we tested them both on the bench
and they both then worked perfectly, although we only took them through
a few cycles, using the (very bright) room lights to simulate daylight.
Without the lights the room is pitch black.

I then had the second unit on test indoors, looking out through an
east-facing window so it could €˜see daylight. It worked fine for four
days and nights.

Everything seems to point to a wiring fault, except that when the fault
condition is present the units LEDs are lit yet the switched line is
still live. We are aware of the 30 second delay function. What could
cause the LED to be lit yet leave the switched live output still
permanently live? After the unit has failed to operate in this way we
have found that brief power-cycling doesnt make it work normally.. When
the power is restored it does the 30 sec thing but then the LED is on
and so is the lighting. To get the unit to turn the lights off in
daylight it seems to be necessary to leave it unpowered all day, and
then power it before the light starts to fade. It will then turn the
lighting off after 30 secs, and will turn it on again at dusk, but will
sometimes malfunction as described at dawn. We did have a good run of
three mornings with the second unit, and I thought the problem had
somehow fixed itself, but then the fault started to occur again.

At first we thought the problem might be caused by the very low outdoor
temperatures weve had here, but it persisted even when the weather
improved.

We just cant think of anything else to try. The units seem to work
anywhere except on the north-facing wall, where they work sometimes.
Have you had any reports of problems caused by strong RF fields? Its a
long shot but we have a low powered local radio transmitter 2km away to
the north.

Bill

Do they switch using relay or triac?
Either way your 12v psu (if it's an smpsu) has a biggish capacitor on its input which will produce high peak currents at switch-on.


NT

Not fully understanding this, but I was indeed thinking that the input from
the actual sensor to the switching logic module could be effectively giving
the on signal expected from the actual sensor. Whether mains borne stuff is
getting to this line adding enough noise to make it trigger is hard to say.
If those circuits are physically separate, some decoupling could be tried,
but of course this would invalidate any warranty.
if the whole thing is one big chip, them the only option is to attempt to
power it from a really well filtered line making sure all paths in are using
ferrites or some other method of filtering.
I do know from experience some years ago, that a PIR battery light I
modified to make an audio sound when motion was detected to help me find my
back gate also malfunctioned when a guy nearby used his mobile phone. This
was unexpected and to my mind makes me suspect a high impedance input
somewhere in the tiny module. I never did fix it.
Brian

--
----- -
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...

Blind user, so no pictures please!
"Bill Wright" wrote in message
news
Here's a letter I plan to send to Danlers. I just wondered if anyone here
could cast any light on the problem before I send the letter.

We have a problem with the TWSW twilight switch which has us completely
baffled. We had two of these items for a while. We got a replacement from
CPC for the first one thinking that it was faulty, but didn't immediately
send it back.

Initially the first unit was installed to replace an existing item (bought
from RS; expensive) that has been in place for many years but had failed.
The location is on a north facing wall about 15ft above the ground. The
unit installed normally and responded to the change of light. We were able
to set up the Lux threshold correctly. However, the next morning the
lighting was on in daylight. On examination the TWSW's LED was lit,
meaning that it should be holding the lighting off. We tried various
things over the next few days but the problem persisted: the unit would
still power the lighting after dawn, and would continue to power it all
day.

The replacement unit was initially set up in a test rig on the bench,
where it worked perfectly. We installed it in place of the first unit and
the same fault occurred. However the fault was intermittent, and some
mornings the lights would be switched off correctly. Wondering if the
lighting load was to blame we tried the unit with a temporary load of a
100W tungsten lamp but with the unit still on the wall. The fault
persisted. Incidentally the normal load is three 70W sodium lamps, six 10W
LEDs, a 1A 12V psu, and four 17W CFLs.

We took the second unit off the wall in case the fixings had distorted the
case and caused the fault but it made no difference. Concerned that the
load might have damaged the units we tested them both on the bench and
they both then worked perfectly, although we only took them through a few
cycles, using the (very bright) room lights to simulate daylight. Without
the lights the room is pitch black.

I then had the second unit on test indoors, looking out through an
east-facing window so it could 'see' daylight. It worked fine for four
days and nights.

Everything seems to point to a wiring fault, except that when the fault
condition is present the units' LEDs are lit yet the switched line is
still live. We are aware of the 30 second delay function. What could cause
the LED to be lit yet leave the switched live output still permanently
live? After the unit has failed to operate in this way we have found that
brief power-cycling doesn't make it work normally. When the power is
restored it does the 30 sec thing but then the LED is on and so is the
lighting. To get the unit to turn the lights off in daylight it seems to
be necessary to leave it unpowered all day, and then power it before the
light starts to fade. It will then turn the lighting off after 30 secs,
and will turn it on again at dusk, but will sometimes malfunction as
described at dawn. We did have a good run of three mornings with the
second unit, and I thought the problem had somehow fixed itself, but then
the fault started to occur again.

At first we thought the problem might be caused by the very low outdoor
temperatures we've had here, but it persisted even when the weather
improved.

We just can't think of anything else to try. The units seem to work
anywhere except on the north-facing wall, where they work sometimes. Have
you had any reports of problems caused by strong RF fields? It's a long
shot but we have a low powered local radio transmitter 2km away to the
north.

Bill

On 20/04/2018 03:35, wrote:

Do they switch using relay or triac?

There's an audible click then the 30 sec pause, then the lights go on/off.

Either way your 12v psu (if it's an smpsu) has a biggish capacitor on its input which will produce high peak currents at switch-on.

Yes. There's a big inrush. A big spark!

Bill