[CliffordK]

Oops. How do I make the GIF bigger?

After posting the message, go back to "Edit" and choose "Insert into Message".  At least it normally works with other images.

[CliffordK]

Two issues.  Initial pause is probably related to a capacitor that must charge before starting the lamp, like the ballast in the old Fluorescent lamps.
Once started, the lamp still takes a while (a minute or two) to get bright.  I think this may be a heat issue.

I have a couple of "Cold Cathode Compact Fluorescent Lamps".  I've been meaning to experiment with more of them.  However, I believe they may actually go from off to bright quicker than the normal CFL lamps.  They are supposed to also use less power than the normal CFL lamps.  Look for light bulbs with thin coils, and ones that say "dimmable".

[syhprum]

The current leaking thru the capacitance of the cable will feed into diode circuit within the lamp slowly charging a capacitor, when the voltage on the capacitor is sufficiently high which could take a minute or so a flash will be triggered.

I note that several other correspondents have put forward this rather obvious explanation.

Knowing the frequency of the power supply And the capaccitance of the cable to the switch (measure or check from tables) it is easy to calculate the current flowing into the capacitor on the output of the brige rectifier, if one knows how large this is one can calculate the rate of increase of voltage and if one knows the voltage required for a flash the freqency of flashes can be calculated.

 [ Invalid Attachment ]

[CliffordK]

Oh, as far as the flash...  Hmmm...
Whether or not the neutral line has power, the positive line should be shut off, and thus there should be no flow of power.

I have a touch dimmer switch in one room.  You should be able to tap it and the power goes out.  However, I noticed that when I replaced the incandescent bulbs with cold cathode compact flash bulbs, they would be on very dim when I had thought the switch was off.  So, now I use the manual switch below it.

Any chance you would be using a dimmer that doesn't truly cut off 100% of the power?

[Geezer]

What about Sean's point about the capacitance in the wiring? If you think about how the light switch is wired, there is usually a long out and back loop with the switch at the end of the loop. The out and back loop conductors are frequently in the same cable (in the US anyway). The capacitance between the conductors will appear in parallel with the switch.
 
Would that be enough to do it? 

[graham.d (OP)]

Geezer, Syphrum, Sean - this probably is the explanation (in fact I can't think of any other) although I did not think the wiring was done in this way - I confess to not bothering to go into the loft to find out, or even looking to see how many cables are behind the switch which may be easier. Had there been two switches operating the light (as alternatives) then this is always the case and there could be coupling from a Live wire to the switched out part of the Live side of the switch but I had not thought this likely with a single switch. However I see that it makes sense as it means only one cable is needed to be in the wall.

Clifford, yes, I have seen problems with dimmers and these lamps before, but there is no dimmer involved here.

I will look behind the switch when I'm next home to see, but I reckon you have the right explanation. Thanks chaps. I can now go back to counting sheep rather than light flashes.

[Geezer]

If that's what's causing it there is not too much you can do about it, although, I suppose you could always wire a small incandescent bulb in parallel with the CFL!

[syhprum]

There are two simple ways to stop it one is to put a resitor (100K 1watt) across the lamp or replace the switch with a change over type that grounds the line feed to the lamp in the off position (if a ground or neutral line is handy)

[graham.d (OP)]

Actually it's no more disturbing than the flashing LED on a smoke detector. I was really just curious as to the cause.

[syhprum]

There is a slightly more technical solution for those in the USA that have a bi-phase 110/220 volt main.
The reverse phase can be fed to the lamp via a capacitor of equal capacitance to that of the switch cable neutrolising the effect.
Make sure the capacitor can withstand 220v or there may be a bang

[Geezer]

There is a slightly more technical solution for those in the USA that have a bi-phase 110/220 volt main.
The reverse phase can be fed to the lamp via a capacitor of equal capacitance to that of the switch cable neutrolising the effect.
Make sure the capacitor can withstand 220v or there may be a bang

I think I prefer the changeover switch!

[Geezer]

Actually it's no more disturbing than the flashing LED on a smoke detector. I was really just curious as to the cause.

Rats!
 
I was hoping to make a killing on Skymall with "Geezer's Patent CFL Anti-flicker Resistor Technology".
 
Simply attach this remarkable invention to your CFL before you screw it into the socket.  Only $19.95 for a pack of five.