Hello John, G4GVC
please find enclosed some more comments on your
EMI problem. Sorry for the "long message".
G4GVC> comments are very much appreciated and useful in
G4GVC> getting to the bottom of this
Collecting "ideas" is one way to work your problem
out. With such a check list you could confirm or
disclose possible noise sources.
G4GVC> The latest situation hasn't changed much as I
G4GVC> have been very busy again this week, so have
G4GVC> had little time to go out DFing.
Sorry for the situation hasn't changed (into positive
direction). Before I settled in my present QTH, I
experienced hard man made noise problems with little
left on my side to combat it.
We settled in the former QTH, a multiple party
house, when only families were living in there.
No apparent problems with receiving signals on
SW, MF, LF. The third party settled in the house,
a few months later receiption on MF and
SW up to 80m got interferred continuously.
The fourth party settled in the house and another
source of interference showed up after few weeks.
The latter interference was active daily for some
period of time (hours) but very intense.
Initial DF with portable MF/LF BC receiver showed
that the possible sources must be inside the house.
Being away from the house for more than 6-10m did
totally calm down the interference to a insignificant
and even undetectable level.
Using the portable BC receiver (utilizing the built-in
magnetic antenna i.e. ferrite antenna) I scanned all
electric, TV and telephone wiring. The interfering
signals peaked up when being close to any conductive
matter, even when close to water and heating pipes.
This way I could not trace down the culprit source(s).
Then I attempted a more systematic strategy.
(1) I logged the occurrence of interference and classified
it in strength. From that I knew ther must be at least
two sources, one being active 24h/7d with moderate level,
others being active daily for hours having high levels
which were also different from case to case.
(2) I tried to characterize the interfering spectrum
in frequency and amplitude, using the same antenna and
receiver, bandwidth etc. All parameters on the receive
side were keept constant.
Again I could discern two different spectra. Both were
built up from 50Hz/100Hz spectral lines starting at
the lowest frequency of the receiver (25kHz) and
being still detectable in the 10m band (I do not receive
anything above 30MHz, TV being the exception).
The 24h/7d signal lost significance (but still being)
detectable) above about 8MHz. The sporadic signal
lost it's significance above about 4MHz.
Both spectra showed an envelope function, densely
filled with mentioned 50/100 Hz spectral lines.
The envelopes had maxima at frequencies which
were multiples of around 86kHz and minima at
the same distances but being 43kHz away from the
maxima.
I list up the amplitudes for both spectra
(the envelope frequencies were slightly
different) and got a frequency and amplitude
"fingerprint" for the signals.
(3) Next I checked wether the 50/100Hz lines were
phase synchroneous with mains voltage - they were.
No phase synchronism detectable for both
envelope frequencies.
(4) Using frequencies where only one of the
interfering spectra were of significance, again
I scanned the house wiring.
Now I could make a difference on several mains
wire bundles as well as on telephone lines.
The portable BC reveiver did not have sensitive
or precise signal strength indicator so I had to
rely on the aural impression of the demodulated
audio.
The continuous spectrum I could trace to the
mains wiring going to the neighbour at the next
floor, the sporadic spectrum I could trace to both,
the mains wiring as well as the telephone wires
going to the neighbour at the lower floor.
The neighbour in the next floor was cooperative,
so I could find the source of interference.
It was a fax-TAM operated from 24 Volts DC,
which was produced by a SMPS external to it.
Using a combination of mains filters and
ferrite cores, this interference source could
be cured.
The neighbour in the lower floor wasn't that
cooperative, I had to call the authority to
finally find the interference sources. He was
operating halogen incandescent lamps (12Volts)
supplied by SMPS like electronic "transformers".
The authority checked the compliance of the
installation as such as well as the individual
electronic units with the EMI specs. Unfortunately
they were in spec but really close to the
allowable limits. The specs call for measurments
on the mains side (power in side) of the electronic
units, not on the low voltage side (power out side).
But latter side and the associated wiring was
causing the interference.
So I could not use the EMI regulations to argue
with the neighbour in order to modify the installation.
If he had used normal 230 to 12 Volts transformers,
he would have obtained the same efficiency and the
same functionality of his illumination installation
with ZERO interference at the same cost.
To cure the remaining sporadic interference I would
have to buy filters for the 12Volt power lines
or buy mentioned 230 to 12 Volts transformers.
Finally I decided to change the QTH for that and
for the reason that I could not erect TX
antennas at the former QTH.
Why I describe my "EMI-case" in such detail?
It may give you an impress in which steps a
systematic approach could lead to find the
culprit sources and find possible solutions
to cure the problem. Also the measurements
and fingerprinting obtained in my private
search process were valuable to argue with
the authority to send a search team.
I did not pay anything to the authority, because
I could prove a interference situation with
my antenna/receiver installation which made
"normal" receiption of BC signals impossible
or at least greatly impaired. The expenses for
the authority were equalized by the broadcasters
which pay for utilizing BC frequencies and for
curing interference.
G4GVC> The noise is still running continuously at S9 on
G4GVC> 136kHz,
It would be of benefit (in case you need to call
your authorities) to know the field strength of
the interfering signal.
For the same reason I recommend to confirm that
the spectrum interfering with receiption on 137kHz
is the same as interference observed on MF BC band.
This could be done by the "fingerprint" method
outlined above.
G4GVC> Yes, the poles where the noise peaked are definately
G4GVC> telephone poles and NOT power-line poles.
As in my case, the interfering SMPS coupled into the
mains wiring as well as into the telephone lines, as it
was supplying a telecommunication equipment, attached
to both. So don't be puzzled if you can detect the
interference on both kind of wiring.
G4GVC> The local transformer happens to be right outside
G4GVC> the telephone exchange, but as mentioned last time,
G4GVC> the noise didn't show a peak anywhere in that immediate
G4GVC> area, and I'm not climbing over the fence to get in
G4GVC> really close!
From that we can preliminarily conclude, that the
mains wiring from house to house and the telephone
wiring from house to house is just distributing
the interfering signals but not necessarily the
cause of the interference.
G4GVC> I happen to have a bat-detector, so if it is an arcing problem I
G4GVC> am already equipped to 'hear' it
Arcing problems are frequently found on but not
limited to lines carrying "high" voltages i.e.
_more_ than 1kVolts. If arcing occurs, then the emitted
spectrum is indeed broad band. Mains supply companies
utilize a portable VHF DF finding equipment to localize
defective isolators causing arcing. The emitted signal
appears as a continous spectrum (envelope function)
with small differences in between maxima and minima
and being filled with 50/100 Hz spectral lines.
There is a similarity with Aurora signals appearance.
The ultrasonic detecting equipment is well for
the close in search process, to find the individual
isolator on a pole, where the VHF DF equipment is
good to sort out the pole while driving along a
HV line or while flying in a helicopter above it.
G4GVC> I used the Walkman to check for noise at our house
G4GVC> consumer box.
Does "Walkman" mean that you used the built in BC receiver?
Or did you utilize the magnetic head of its tape recorder
to detect magnetic fields?
G4GVC> Yes, it is louder there, but so are all the BC signals.
G4GVC> The noise is louder still where the central-heating
G4GVC> pipes run up the wall and into the loft (attic), so
G4GVC> I think this is all just a re-radiation effect.
Yes, apparently. At first I made the same observation
with my EMI case. I had to made a spectrum fingerprint
first (selecting a frequency component associated with
only one EMI source) befor I could make a more direction
sensitive search. Observing the amplitude and/or the
antenna's position while scanning along the house
wiring bundles did the thing to discern which wiring
actually carrying the interfering signal.
G4GVC> the medium-wave noise appears to NOT be local to me.
G4GVC> On foot, I observed it getting weaker, but in the car
G4GVC> driving to work I discovered that it comes up again as
G4GVC> you get further away and there appears to be a fixed
G4GVC> 'standing-wave' effect of peaks and troughs in the
G4GVC> noise. I could still hear it well on the car radio
G4GVC> at work - 5 miles away in the centre of the city!
G4GVC> Now I have to investigate whether I am confusing the
G4GVC> issue by listening on medium wave - maybe that is a
G4GVC> different source from my 136kHz problem? If not,
G4GVC> it's one heck of a noise source.....
Please check that the interference on 137kHz and
the interference on MF BC band is in fact the same
(fingerprinting method). You (as happened to me
initially) get easily fooled if you cannot
discern the source(s).
OK John, I hope this info is of some benefit for you.
Please do not hesitate to get in contact with me for
any assistance.
Best 73
Gamal Soegiono (SWL)
50N01 / 08E27 / JO 40 fa
[email protected]
| 