On Mon, 25 Jun 2007 23:47:00 GMT, Eeyore
put finger to keyboard and
composed:

Franc Zabkar wrote:

Eeyore put finger to keyboard and composed:
Sjouke Burry wrote:
Eeyore wrote:
Sjouke Burry wrote:

Capacitive leakage from the HOT wire to the switch
wire?

So why doesn't this happen all the time ? Answer, the capacitance is very low,
as is the frequency.

A leak cap has very high impedance, and via
the input rectifier can charge the input cap(slowly).
Then when a threshold is passed, the circuit
produces a flash.
And it only happens, if the input capacity of the
rectifier circuit is low compared to the leak cap.

It flashes every 2 seconds or so.

Clearly a very low leakage current won't do that.


This fellow experimented with a ~1mA leakage current at 240VAC:
http://groups.google.com/group/aus.e...5398c8c?hl=en&

"Changed to 270 kohms and now have a 13W CFL flashing at exactly 1
Hz".

I make that ~ 11nF !

You won't get that easily.

Graham

Yes, I would think that even 100pF is way too much. The hypothesis
regarding stray wiring capacitance does seem very far fetched.

Maybe this explanation is a better one, a least for the other thread:
http://groups.google.com/group/aus.e...9fbd911?hl=en&

AFAICS, you can *very* roughly calculate the amount of leakage current
required to account for the OP's observations by assuming that all the
current is used in charging the main filter cap, say 4.7uF. Assuming
also that the trigger voltage for the DIAC is around 50V, then ...

i = C . dV/dt
= 4.7 E-06 x 50V / 2s
= 0.12 mA

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.