[time-nuts] EMI from florescent lights
GandalfG8 at aol.com
GandalfG8 at aol.com
Fri Aug 10 03:44:12 EDT 2007
In a message dated 10/08/2007 06:13:01 GMT Daylight Time,
hmurray at megapathdsl.net writes:
Does anybody know how much garbage comes from the standard CFL "60 watt"
replacements for incandescent lamps? I'm interested in both WWVB and GPS.
WWVB is 60 KHz which is the 500th harmonic of 120 Hz. 500 is a big number,
so maybe that's not a serious problem. I don't have any WWVB gear right now
so this part is just academic curiosity.
On the other hand, that fixture is only 6-8 ft from a couple of GPS antennas.
This seems to be one of those areas when yer pays yer money and takes yer
"Normal" fluorescents can wipe out LF and well into MF, say 3 or 4 MHz at
least, that harmonic ratio is no security.
In a previous house I had several fittings in a garage 6 or 7 metres from
the house, admittedly old and buzzing well, and if I forgot to switch these off
the RFI wiped out LF/MF totally....at least it was easy to spot and fix:-)
We had fluorescents and CFLs in our recent urban apartment but backround RFI
was so bad to start with it was impossible to detect any difference with
lights on or off.
The more recent house we have now also has a couple of fluorescents in the
garage, but in much better condition than those mentioned above. These also
generate interference but nowhere near as bad as before.
I've also had experience of nearby fluorescents where there didn't seem to
be any noticeable RFI at all.
I much prefer the light from fluorescents, over incandescent bulbs, for a
lab or workshop environment but do try to avoid them if possible.
CFLs though are not just "normal" fluorescents wound up small. In place of
the iron cored choke they use an electronic ballast.
I'm no expert on these but assume they can be considered to have similar
characteristics to a switch mode power supply, and they operate at a much higher
frequency than 50 or 60 Hz, I've just seen one example quoting 34 KHz.
Again, some appear to be better behaved than others but for anyone wanting
to receive on LF/MF, and hence WWVB, I would recommend avoiding them like the
The RSGB mag RadCom has carried reports on some of these for the EMC page. I
don't have these back copies to hand but there may be copies of the articles
online somewhere. I'm sure a Google check would provide more info.
Another item to avoid, or at least be wary of, is the dimmer switch for
These chop the incoming sinewave by varying the switching angle and can
again generate severe LF/MF interference.
I've heard bad reports of these but have also used LF loops just a few feet
above them with no effect, so again mileage varies.
Many domestic appliances, TVs etc, now carry switch mode PSUs which are
permanently on unless disconnected from the mains. These are becoming increasing
sources of LF/MF/HF interference and there's been many reports of severe
interference traced to such items.
Then there's good ole "Broadband Over Powerline" (BPL), another great source
of interference below 30 MHz.
So that's the sort of fun you might expect with WWVB:-)
Having said all that, I would not expect there to be any problems from
either type of fluorescent with GPS equipment operating in L band, providing good
RF practice is followed.
Fluorescent tubes interfering with low level 10GHz microwave signals is a
well perceived phenomena but this is due to the the ionised gasses being in the
It might be an interesting area for observation but would suggest changing
the antenna or checking wiring if a problem were observed.
If direct RF interference is unlikely, but there's always the possibility of
noise on power feeds with any equipment. Proper screening and grounding
should avoid any issues.
With GPS timing, the only low level signal is the GPS signal itself, signals
out of the receiver module are much higher amplitude so much less liable to
the sort of interference being considered.
Power line conditioning is worth considering for any lab environment though,
just because the amount of mains borne crud is increasing all the time.
Perhaps a Faraday screen or three might help too:-)
More information about the time-nuts