Broadband noise on 160 meters from new Apartment Complex

Oct 5, 2014

On 160 meters I detected a rise in daytime noise level by approximately 4 dB compared with a year ago when using my 300 degree pointed pennant RX antenna. Direction finding on 160 meters using a 14" diameter shielded loop over the last couple of days indicates the noise is originating from a new apartment complex that's located 0.4 miles at a heading of 307 degrees from my house. The noise is continuous across the 160 meter band and when received on an AM receiver using a wide filter the audio output sounds like "white noise".

Here is a link to a youtube video showing the increased noise level compared with my 40 and 160 degree pointed pennants when using my Kenwood TS-180s with a 500 Hz CW filter: http://www.youtube.com/watch?v=cKW1uKz2cA4&feature=youtu.be

Here is a link to a youtube video recoding of the noise using a DX440 receiver in AM mode using its wide filter: http://www.youtube.com/watch?v=WQ22Fsm3ggI&feature=youtu.be

Here is a link to a youtube video recording of the noise using a Kenwood TS-180s in my car while parked in the back lot of the apartment complex: http://www.youtube.com/watch?v=Z2OuhfegtpE&feature=youtu.be

Here is a link to a youtube video recording of the noise while located at the apartment complex where the signal is extremely strong and I had to remove my preamp that was providing approximately 46 db of gain, and I also had to back off the gain on the receiver in order to easily detect a null on the shielded loop antenna: http://www.youtube.com/watch?v=W2B87ExiLHo&feature=youtu.be

Below is the direction finding plot that led me to the apartment complex.

(click on image below to enlarge)

Oct 24, 2014 Update :

Noticed this morning that the signal had a repeating pattern in which it was on for 30 minutes and off for 5 minutes which I recorded at my home. I then jumped in the car and drove over to the apartment complex to confirm the signal originating from that location did in fact follow the 30 minutes on 5 minutes off pattern and was in time sync with the signal at my house, and it certainly was (which I was happy to see). This confirms with 100 % certainty that the interfering signal at my home is indeed the same signal that I tracked down to the apartment complex. I recorded this event at the apartment complex in which I predict when the signal will go off and come back on, and here is a link to the youtube video : http://www.youtube.com/watch?v=6AiBarOvMbs&feature=youtu.be

Nov 14th, 2014 Update :

Interference time study from approximately 11:15am until 6:15pm on November 13th, 2014 yielded the following plot.

Note : the pattern appears to change at night, but it appears to closely follow the pattern shown below between the hours of 6am and 6pm.

Nov 18, 2014 Update

At 4pm today the outdoor temperature is 18 deg F and the interference is still present with on time of approximately 29 minutes and off time approximately 8 minutes.

Dec 11th, 2014 Update

Worked with the maintenance manager of the apartment complex and quickly tracked the source of the noise down to the "Kitchen air makeup unit" which is part of their HVAC system. When we turn off the disconnect to this unit the interference goes away, and when we turn the disconnect back on the interference returns.

Jan 28th, 2015 Update

Located the variable speed drive for the above mentioned HVAC system based on information received from the regional maintenance manager. Examined the contents of the VSD enclosure (with local building maintenance manager in tow), and determined that shielded cable which is recommended by the drive manufacturer for EMC compliance was not used between the drive output and the 5 HP motor (estimate the motor is located approximately 45 to 50 feet from the drive, and it's remotely located on the roof which is 3 stories up). Metal conduit was used to house the drive output wires (a wire for each phase along with a ground wire), but not sure if there is really good electrical bonding between the conduit and the enclosure due to the painted surfaces on the enclosure, etc., and no idea if the metal conduit is properly terminated on the motor side. Also not sure if the metal conduit is in contact with other metal structures along its run which would be non desirable in regards to the transmission of common mode current. Now in contact with the drive manufacturer (Lenze) to determine what recommendations they have based on the existing installation, and to also get clarification on the cable shielding diagram in the Lenze VSD manual which appears to be in error.

Note : while examining the VSD enclosure we were fortunate to see both drives shut down at the same time and interference on my 1.8 Mhz receiver which was sitting next to us disappeared, and then a short time later the smaller 3 HP drive started back up and the interference returned, and then a short time later the 5 HP drive started back up and the inference on 1.8 Mhz increased a substantial amount. This pattern matched the pattern I heard on Oct 24th when I first heard a faint motor start up, followed by a much stronger motor start up. I believe the 3 HP motor runs an exhaust fan in the kitchen whereas the 5 HP motor which is located on the roof is the kitchen air make up system motor.

Picture of the 3 HP and 5 HP drives located in the Kitchen of the apartment complex

Picture of the conduit exiting the VSD enclosure (5 HP metal conduit center of picture)

March 30, 2015 Update

Bonding of the conduit between the drive and the motor has been improved and common mode choke (Danfoss High Frequency Common Mode Cores part # 130B3257) installed on the output side of the drive, and there has been no improvement (no reduction) in the signal strength of the interference. RF spectrum during start up of the drives was recorded on Saturday 28MR15, and the recorded spectrum plot has been uploaded to youtube at http://youtu.be/LIWd2Dd-Fa0

The spectrum plot is very interesting (see below), as you can see very distinct signals spaced every 4 khz upon start up, and then quickly the 4 khz spaced signals blend together and cause wide band interference across the entire spectrum (more like white noise across the entire spectrum).

April 22nd, 2015 Update

Input line filters made by the drive manufacturer were installed on both drives on April 22nd, 2015, and they did have a positive impact on reducing the level of interference. Can still detect start up of the drives 0.4 miles away, but the interference peaks during start up appear to have been reduced by approximately 10 dB (they are now about 5 dB above my noise floor), and the steady state broadband "white noise" is now hard to detect from 0.4 miles away. Interference level in the parking lot of the apartment complex is still extremely strong. (Note : I suspect the Danfoss common mode cores that were previously tried (unsuccessfully) were not installed correctly based on recent information I obtained.)

May 20, 2015 Update

Screen Capture on May 20th, 2015 shown below that shows the 4000 hertz spaced signals during VFD drive startup are now 5 db above my noise floor (they were 15 db above my noise floor prior to the installation of the input line filters. Here is a link to the actual screen capture video that has been uploaded on youtube : http://youtu.be/g7XJcwfa0kY

Website created and maintained by Don Kirk (wd8dsb) Oct 5, 2014