[Lowfer] Government Mandated Noise ???

[Ed Phillips]

Peter Barick wrote:

>Andy, 
>The CFBs are expected to be short lived in consumer application. They will be superseded by LED types.
>
>There are several downsides to CFB with cost being one. That should be surmounted with LED bulbs.
>
>But when? Don't know.
>Peter
>
    At least some CFL's are horrid noise generators.  For example, here 
in the "radio room" I have 4 20 watt CFL's in a ceiling fixture.  They 
generate about an S9 noise level at 50 kHz on my NRD 535 [AM, max 
bandwidth] when I receive on my 50 foot vertical whose base is about 10 
feet from the fixture.  Remember that they use switching transistor 
oscillators and that the voltage across the bulbs is around 140 V p-p.  
I've been using them since they first appeard [in the few sockets where 
the enlarged base will fit and where the reduced light output is 
tolerable] and have had typical life estimated at maybe 2000 hours.  
Lots of premature failures and I've had two fail more or less 
catastrophically - something shorted inside and the fuse resistor in 
series with the line blew resulting in typical "burnt smell" and 
evidence of fire coming out of the case.  I think a clue to their 
reliability is that I always mark the 'in service' date and purchase 
date on the base and keep the receipts.  When I haul a dead one in to 
the store where I bought them there never seems to be any interest in 
looking at the paper work at "customer service", just the words 'go back 
and get another one off the shelf'.  The worst reliability we have had 
is with Lights of America lamps which also seem to have quite a bit less 
output than claimed.

    By the way, if you have failed ones you don't return pry the shell 
open and you'll find quite a few usable components including a full-wave 
bridge rectifier [usually 4 1N4007's], one or two 20 ufd 200 volt 
electrolytics, a pair of 400 volt switching transistors, a number of 
film capacitors, a neat ferrite-cored ballast inductor, one diac, and 
lots of resistors.  They usually also have a small toroidal oscillator 
coil.  No wonder the bloody things cost several bucks - there's a lot in 
them!

    As for light output when cold, that is real enough just as it is in 
conventional fluorescents.  Starting isn't a problem and they usually 
warm up within a minute or so.  This is in Socal where temperatures as 
low as freezing are pretty uncommon.  I don't know what would happen in 
really cold places like Minnesota or Alaska.

   The best price I've seen quoted for an industrial-grade screw in LED 
replacement for a 20 watt incandescent is about $60 in quantity - 
"there's the rub".  Prices will come down eventually [presumably if 
there's enough usage and manufacturing process improve enough] but 
initial cost will be always be high and the payback time based on energy 
usage will be many years.  Another problem is color temperature.  The 
"warm tone" LED's have only about 2/3 the efficiency of the harsh 
blue-white ones we usually see.  A web search will reveal lots of 
sources of information.  I have a friend who spends works for ON 
semiconductor and spends a good part of his time designing power 
supplies for industrial LED applications - they need current regulators 
at the very least.  According to him the place they are really gaining 
acceptability, in spite of the cost, is in places like display windows 
of stores where the cost and inconvenience of replacing incandescents 
more than justifies the higher first cost of the LED's.  Remarkable 
little gadgets to be true, great for pocket flashlights, but don't 
expect to find them in the consumer stores as replacement for 
incandescents for a number of years.

Ed