Ian Stirling wrote in message .. .
N. Thornton wrote:

OK, the factors stated above do not ensure no 50Hz current is present.
snip
With such a setup, if one emitter has total failure, the tube can
conduct current one way only, giving intolerably strong 50Hz flicker.

No, it can't.

An electronic ballast has a rectifier at the front of it.
The waveform after the rectifier is NOT a 50Hz sine wave.

I'd have to draw the cct and label the waveforms, and I'm too busy
today. I could sit here and assure you it can flicker at 50Hz, but
that wont convince either of us. Maybe if I get the time later I'll
draw it and we'll see what comes up.

I'm not saying it's impossible.
snip
What can happen with a marginal bulb or broken driver is that it may
try to restrike the tube often, and it doesn't quite work, leading to
pronounced flicker.


I think I understand where our wires are crossing. There are 2 ways to
wire an electronic ballast.

One way is to rectify the mains, smooth it, generate hf ac, and apply
that to the fl tube. That method wont produce 50Hz flicker.


The other way is to put the ballast in series with the tube, like
this:

~L--------ballast---------tube---------+
|
~N-------------------------------------+

In this case 50Hz mains is going through the tube, not hf ac. The
ballast contains the usual rectifier and circuitry, but the tube is on
the ac side, not the dc side of this rectifier. Now if one tube
electrode loses emission, the light will function but at 50Hz instead
of 100Hz. It will light every other half cycle.

In this type of setup there is nothing to prevent 50Hz flicker. It
should not happen as long as the tube is capable of supplying
sufficient current each way, since the ballast limits the i flow
equally. But when one emitter is performing poorly, 50Hz flicker can
become strong. And when one emitter breaks the 50Hz flicker is
intense.

These ballasts work by turning the rectified ac into hf ac, passing
that through a choke. The 'output' of the choke is simply shorted, the
real load being the tube in series with the whole of the ballast
circuit.


Regards, NT