NOISE MEASUREMENTS and REDUCTION
UPDATED 19 January 2005

A Typical Problem:
  Click the link below for a real world scenario of a noisy power line located approximately 2 miles from my hilltop site in Hillsboro Center, NH.  All measurements were made at 50 MHz using the degradation measurement described later and represent dB of degradation over a cold resistor (a baseline measurement that allows no external noise to enter the system) at the bearings shown.  This most recent problem of 20.5 dB degradation began just as the Fall 2000 F2 season became productive for world wide six meter DX.  The bearing of 110 degrees is particularly troublesome because SKEW PATH propagation, peaking in that general direction,  is the primary mode from here in early fall.
  The data obtained shows TWO important features.  Firstly,  the MINIMUM degradation realized over a cold resistor.  This turns out to be 8.0 dB and is the best that can be obtained at any bearing from this location.  This value shows approximately how much excess gain my low noise figure receiver system has in the real noise 50 MHz environment surrounding it.  The reduction of this excess gain is essential to insure the lowest level of receiver intermodulation products.  More about this later.  The second important feature shown is the level of degradation from the target source.  Knowing this value allows one to estimate the goal in terms of the amount of noise reduction required from this singular source.  That goal is approximately 15.5 dB reduction and is obtained as follows:
1.  If the single source contributor had the same noise power as the receiver NF then the degradation would be exactly 3 dB
     due to the summation of noise powers.  However,  the antenna will pickup other contributions so that 3 dB reserve is a
     conservative amount of extra noise reduction.
2.  If the best case degradation is 8 dB and assumed to be from non manmade sources,  then reducing the contributor noise
     much more that 3 dB below this value would offer little or no improvement of signal to noise ratio.
3.  Therefore,  the reduction goal for the 110 degree contributor  ~ 20.5 dB - 8 dB + 3 dB reserve = 15.5 dB

View a POLAR PLOT of the initial noise environment (NOISE BLANKER OFF)

  Note:  Assuming 6 dB of free space path loss every time distance is doubled and ~ 12 dB of noise loss required from this single contributor located 2 miles away to equal the quietest bearings means that this same noise source would cause approximately 3 dB of degradation at a distance of roughly 8 miles!!!!  Making matters worse is that the TOTAL degradation realized in the real world is the summation of all the noise powers that your antenna views.  Another,  less tasteful way of saying this is to consider a situation where you have TEN separate noise contributors,  each at the same level and you remove only one.  The net improvement realized will hardly be measurable!!!!!

  Here is another plot made about one month later.  This one was created when the weather conditions were very dry and cold (temperature 16F) and a wind out of the Northwest at 10-15 m.p.h.

This kind of MANMADE noise environment can ruin your chances of 6m DX!

  It took almost a year for the local electric company to resolve the above problem of the 110 degree "contributor" even after I pointed them right at the source of the problem.  They claimed that they found numerous other problems in that general area and that extensive and costly work was required but finally they fixed it as evidenced in the following plot.  Unfortunately, over that period of a year other loose hardware and failing insulators began to rear their ugly head and become problems from other directions.  Some of these may have shown subtle signs of becoming a problem in earlier plots. One of them, the Northeast noise source, was "fixed" by the electric company some years earlier.  The peak at 140 degrees was particularly troublesome during the PW0T DX-pedition- 19 dB of receiver degradation!  To me, the plot shows 1) a new Western contributor, 2) A Northeast contributor that showed up in earlier plots- now MUCH WORSE, 3) A SE (140 degree) contributor that showed up in earlier plots, 4) A SSW contributor that showed up earlier BUT- They fixed the 110 degree contributor!

Click here for a 25 February 2002 Plot taken with the 6M9KHW yagi mounted at 110' AGL

TO BE CONTINUED...



The Measurements:
    The best (good science) method of measuring and plotting receiver interference from an external source is the DEGRADATION TEST.  The minimum equipment required is an ISO-TEE, a good 50 ohm RF load and a signal generator.  If you utilize a preamp and wish to optimize gain for best sensitivity consistent with best receiver intermodulation rejection, then a step attenuator is temporarily required to determine the minimum FIXED attenuation value that should be placed between the output of the preamp and the input to the receiving system.  Each case is different and I strongly recommend that this also be accomplished!!
  When accomplishing a degradation test,  you will compare the measurements of receive sensitivity reduction in dB of:  A resistive load which does NOT introduce external noise versus the reduction of sensitivity with an antenna connected which does allow external noise to be introduced.  This requires the use of an ISO-TEE.

CLICK HERE to see the TEST SETUP and how to build an ISO-TEE

    As can be seen from the diagram,  the iso-tee is nothing more than a three port device with two of the ports directly connected (to maintain a 50 ohm Z-match) and a third port that is isolated from the other two.  The isolated port is typically capacitively coupled (you build a capacitor by adding the insulator) and will have a frequency rolloff characteristic of 6 dB per octave.  The isolated port is ALWAYS CONNECTED TO THE SIGNAL GENERATOR and will have a good deal of loss at 50 MHz.  40 dB loss is not uncommon so the signal generator must be capable of an output level significantly greater than the system sensitivity + worst case external degradation + ISO-Tee loss.  Therefore,  a receive system that is degraded by 20 dB and tested on the bench to have a sensitivity of -120 dBm with a 40 dB loss ISO-TEE will require more than -60 dBm of output from the signal generator.  Due to the higher than normal level on the cable leading from the signal generator to the ISO-TEE,  it is essential to use good quality LOW LEAKAGE coaxial cable for this connection!

Initial Setup:
1. Connect equipment as shown in the diagram with a 50 ohm dummy load in place of the antenna
2. If you utilize a preamp and there is an attenuator after it- replace attenuation (or switch attenuation) so the value is 0 dB
3. Turn OFF the receiver noise blanker
4. Stay with the SAME measurement technique throughout whether it be M.D.S. or receiver voltmeter S/N ratio
5. The tests simply compare the external noise to that from a "cold resistor"

Step A- Establish the baseline relative sensitivity:
    With a dummy load in place of the antenna,  inject sufficient signal generator level into the ISO-TEE isolated port to obtain M.D.S. or a given S/N ratio measured on a voltmeter from the receiver under test.  Record the signal generator level in -dBm or -dBw.  This is the best zero degradation relative receiver sensitivity that can ever be attained as there is no external noise contribution.  As an example- let's assume the value is -80 dBm.

Step B- Now connect the antenna system:
    Remove the dummy load and install the antenna.  Next set the signal generator for the same M.D.S. or given S/N ratio measured on a voltmeter as the above test and record the signal generator level in -dBm or -dBw (same as above step).  This test introduces the noise from the outside world. Let's assume the value is -60 dBm.

Step C- Calculate the degradation in dB:
    Simply subtract the two readings.  In the example above,  we see -80 dBm with no external noise contribution and -60 dBm (20 dB stronger) with external noise contribution (antenna connected) so that the measured degradation is the difference or 20 dB.  In other words,  an external signal (simulated by the signal generator injection via the isolated port) must be 20 dB stronger with real world noise present than with real world noise removed.  For low noise figure systems,  don't be surprised to see 6 dB of degradation or more from the quietest directions.  50 MHz is a very noisy band and this noise is from sources that are NOT generated by man!
  It is suggested that a plot be created at 5-10 degree intervals for a full 360 degree sweep.  This will allow you to determine all your external noise sources and the quietest bearings.

Step D- Eliminate external noise sources until satisfied:
    Record your progress using the methods above.  Once satisfied that the noise environment has been optimized and if you utilize an external preamp- then it's time to optimize it's gain.  This is a pretty basic rule:

    For each one dB of excess gain into your receiver that you can eliminate,  the receiver third order intermodulation products will fall by 3 dB!  There are not too many things in life that offer a "pay back" of three to one but this is one of them!

Step E- Optimizing Preamp Gain:
    This procedure will insure that you will realize best possible system sensitivity in the noise environment that surrounds you.  Each case is different and several factors are involved.  Many may suddenly realize that the brand new preamp they just installed is doing nothing but causing an S1 signal with a 6 dB S/N ratio to be an S9 signal with a 6 dB S/N ratio and a receiver that is severely overloaded in the presence of moderate signals.  This procedure will "tell all"!

1. Set a step attenuator to 0 dB and insert it between the output of the preamp and the receiver system.
2. Turn the noise blanker ON.
3. Duplicate the Steps B and C above and point the antenna to the quietest (least degradation) bearing.
4. Using M.D.S. method,  increase the step attenuator attenuation value carefully until you just begin to notice the weak
    M.D.S. begin to degrade (with a bias towards maximum attenuation). Record the amount of dB on the step attenuator.
5. Replace the step attenuator with a fixed value of the result in Step 4 minus 1 dB.  This is the MINIMUM optimized value.



NOISE and INTERFERENCE REDUCTION USING THE ANC-4 NOISE CANCELLER

    This device,  presently marketed by Timewave,  seems to offer great potential.  It contains a combiner and a phase adjustment network that allows the user to take signals from two antennas (a noise sense antenna and a main antenna) and adjust the phase into the combiner so that:

1. If the two signals are combined in phase,  the output will be the sum of the two.  Thus, space diversity (gain) is possible.
2. If the two signals are near out of phase,  the output will be the subtraction of the two.

  The secret to success appears to be that if one antenna receives a very weak signal AND a stable noise but the other antenna receives only the same stable noise AND NOT the desired signal- then the phase can be adjusted so that the noise can be canceled out typically 40 dB or greater.  The same reduction can be realized with QRM!  If I am pointing to Europe and bothered by weak qrm off the side of the main antenna and point the sense antenna directly at the QRM source so that it cannot "hear" the weak Europeans-  I may be able to null out the qrm entirely!  Needless to say,  it was an easy decision to purchase an ANC-4 after realizing it's potential.  I would much rather turn off my noise blanker to prevent receiver blanker  overload and have a "box" at the RF level take care of this task.

CLICK HERE to see the INITIAL SYSTEM DESIGN planned- using the ANC-4

    The concept is to utilize TWO very large 50 MHz yagis with very clean patterns.  The existing 11 Element yagi (two wavelength boom) is at 57' AGL mounted on the North Tower.  On the South Tower, 100' away is a 6M9KHW (9 El 50' boom) yagi mounted at 110' AGL (installed 6 July 2001).  This design will NOT have these yagis stacked.  They are each under the control of a separate rotor.  Antenna selection will be simplified (but more expensive to implement) by utilizing a single selector switch (S1) to control the routing of the TRANSMIT signal.  S3 forces the transmit antenna to always be the UPPER and allows reception on the lower dring those times when this configuration may be advantageous (local noise pickup on the UPPER antenna can be higher if the source is close and shadowed by terrain from the lower antenna). Receiving on the lower antenna in this mountainous area can sometimes offer a 15 dB signal to noise improvement!  RF relay K3 insures that whichever antenna has been selected as MAIN (also transmit if S3 is in the auto position)- the OTHER antenna is always selected as the sense antenna.  RF splitters are utilized for several reasons.  Most importantly,  they allow for non critical cable lengths to be used between their output and K3.  The splitter also allows tapoff of either the main or the sense antenna (switch selectable via RF switch S2) to drive a second receiver.  Separate preamps are utilized to obtain a low system noise figure.  The combination of the 6 dB specified loss of the ANC-4 added to the theoretical loss of a 4-Way splitter plus cabling and relay losses will near 14 dB so preamps are absolutely required.  Attenuation is placed after the preamps to swamp out excess gain.  This will be an expensive and time consuming project but it's just part of being a weak signal VHF DX chaser!

CLICK HERE to see a comparison of the HORIZONTAL patterns for the two antennas

CLICK HERE to see a comparison of the VERTICAL patterns for the two antennas
 

PROJECT PROGRESS and OBSERVATIONS

    Unpacked and installed ANC-4.  First observation was a rather noticeable insertion loss.  I didn't bother to measure this but I did remove the 6 dB attenuator after my preamp and determine if there was any measurable loss of receiver sensitivity using the degradation tests described earlier with the external (main) antenna connected.  There was NONE.  My conclusion is that the initial system design remains valid and that the dual  preamps and gain adjustment will be required and adequate to overcome the insertion loss of the ANC-4.

    Played around with various antenna configurations- unable to obtain a null on my most bothersome 50 MHz noise source- 2 miles away.  However,  I presently have only one 6m antenna and a very good off the air TV antenna system so maybe I could perform another test just to see if the ANC-4 really works:  Can it reduce or eliminate TV carriers/hash with a more efficient antenna system?
    With the off the air (mast mounted preamp with > 54 MHz low frequency cutoff) TV antenna pointed to the Boston area and peaked at a local TV4 station there on 67.240,  I plugged it's multicoupled output into the ANC-4 "sense antenna" port.  Using the 6m yagi as the main antenna- I tuned to the next closest offset (67.250) for a weak distant TV video carrier located at a bearing 90 degrees away from the Boston interference source.  Without the ANC-4 this very weak signal was buried in buzz from the locally generated source.  With the ANC-4 inline I was able to obtain a very decent null on the interfering signal and easily hear this weak signal.  However,  this was short lived.  Both large arrays were blowing around in high winds during this test and that was probably the reason for the instability of the null.  Conclusions:  The manufacturer spec. of 70 MHz is probably true and at least I know the ANC-4 works and WILL WORK at six meters.  Will long boom yagis with narrow beamwidths cause unstable nulls when the wind blows?

  LOSSLESS COMBINING- NEVER!  Don't forget that even though this design has preamps on both the MAIN and SENSE antenna ports and these preamps are designed to compensate for the 6dB insertion loss of the ANC-4 (to maintain a low system NF)- this concept oversimplifies what will really occur!  In the IN-PHASE (diversity combining mode) condition- consider each preamp to be terminated in a 50 ohm load with NO EXTERNAL NOISE CONTRIBUTION.  Also assume that the thermal noise (NF) from each preamp is identical.  Then it is practical to assume that if we vector sum only the white (incoherent) thermal noise contributions of both preamps- fed in-phase to a common node- then the theoretical & relative LOSS of system NF would be equal to 3 dB using this design because the preamp contributed noise powers will sum.  All references seem to disregard this physical reality!  However,  with the dynamic changes in arrival angle on 6m prop modes (if the antennas are significantly separated vertically) AND the diversity gain considerations (if the antennas are separated significantly- horizontally & vertically) it may still be practical to assume reasonable "smoothing" over a given period of time and apparent receive gain when using such a design at six meters in the in-phase (diversity combining) mode.  Time will tell!   At best, this design should be considered psuedo lossless combining ONLY BECAUSE the user can manually switch between Antenna A, Antenna B or BOTH to manually determine best receive - FOR THAT GIVEN INSTANT IN TIME.

    06 July 2001 UPDATE:  The second high gain yagi, an M2 6M9KHW (50' boom) is now installed at 110' AGL 100' South and 53' higher than the 6M1136 (57' AGL)  yagi.  Each yagi seems to have a markedly different noise environment.  My 1,000' distant neighbor's in-band consumer device "garbage" is essentially eliminated when the higher antenna is chosen.  However,  a 1.6 kM distant noise source is 11 dB stronger on the upper antenna.  Initial "isolation" tests at the apex of a triangle 83' away from each antenna and 40' AGL yield 12 dB less signal with the upper antenna.  Longer haul signals show improvement on BOTH antennas depending upon the propagation mode and conditions at the instant of selection.  Rarely are longer haul signals identical on both antennas.

  11 July 2001 UPDATE: I live in a mountainous area on the summit of a hill.  Some of the power lines are well below me in deep valleys (shadowed from the lower high performance antenna).   Today it was raining hard and a nearby insulator was arcing creating an additional 15 dB of noise on the upper antenna with NO NOTICEABLE NOISE DEGRADATION on the lower antenna (shadow loss).  To allow user selection of transmission on the upper antenna while receiving on the lower antenna  I have added switch S3 to the original design.  This modification will benefit ME but probably not those surrounded by smooth rolling terrain.  Also upgrading the high power relays to L versions with better isolation to handle 1.5-2.0 kW may be worth considering.

30 January 2002 UPDATE: The entire 6M station has changed. Thus, I must look over the original ANC-4 interface and make the appropriate design changes.  The new setup of a Mark V and FTV1000 no longer splits the receive and transmit paths so that dual relays will be required.  When I have time, I'll do the design and modify the schematic.  Priorities change- especially when the 30 year old exciter dies in the middle of a Solar Maximum!

    More to come- this will take some time...


6M NOISE & QRN REPORTS
    It is my hope that this section will someday be filled with good information and links to what others have found to cause degradation in the 6M band.  There are MANY sources out there besides noisey power lines.  Please share your experiences so we can make this a quieter place for all!

DSS Satellite Receive Systems:
  Several have reported strong 6m in-band hash generated by these receive systems.  I'm looking for more good information.  If the LNB is oscillating then it's a design issue.  If the 6m hash is generated by the receiving box,  then an in-line preamp should have enough reverse isolation to resolve the problem.  Sure- It's still a design issue but if I can fix it by giving my neighbor a $20 in-line preamp then I consider this a good (practical) solution.

FM Audio near 50.750 when in close proximity to a Channel 2 TV station:
  Consider this:  TV2 has a video carrier frequency of 55.250 (+/- 10 khz) MHz and in the USA/Canada- The Aural carrier is always 4.5 MHz higher so that it would fall on 59.750 MHz if the video carrier assignment were 55.250 (59.740 for - offset or 59.760 for + offset video carriers).  The composite signal is AM and therefore contains both an upper and lower sideband but the lower sideband is partially suppressed close in to the video carrier frequency while the full bandwidth of the upper sideband is utilized.  AKA- Vestigal Sideband.  If the lower sideband of the composite AM video signal were calculated,  it would place the aural FM exactly 4.5 MHz below the 55.250 video carrier or at 50.750 +/- offset.  If you are close enough to the source transmitter- then you will hear the signal.  At -60 dBc, this signal can be audible for many, many miles.
  Solution:  Can only be resolved by better filtering within the source transmitter.  Good Luck on this one!  An ANC-4 or similar setup may be able to resolve this problem.

MORE FM Audio near 50.750 when in close proximity to a Channel 2 TV station:
  It just so happens that 50.750 (+/- 10 khz) MHz is a perfect mathematical THIRD ORDER INTERMOD product (2 * 55.25 - 59.75) between the Aural and Video carriers of a composite channel 2 TV signal.  Unfortunately it may not be readily apparent that this could be intermodulation because as per the formula, the FM Aural signal deviation will NOT be multiplied by 2 (the AM video carrier will).  If the mixing is within the affected receiver (receiver intermod), attenuation or filtering between the receiver and antenna should help to eliminate the problem.  However, transmit intermod is also a possibility. In this scenario, non-linear joints (aka rusty) in close proximity to the broadcast station are illuminated and re-radiate the mixing products.
  Solution: Determine if receiver intermod by placing a step attenuator between the receiver and antenna.  If the degradation drops at a greater (typically 3 dB for 1 dB of attenuation applied) rate than the applied attenuation then consider yourself lucky and filter to prevent the 55.250 and 59.750 signals from reaching your front-end.

50.113 Carriers
  Mass produced 3.58 Mhz TV Color Burst crystals are commonly used in many consumer devices other than TVs.  The 14th harmonic falls near 50.113 and this can radiate a great distance.

BEEPING AT REGULAR TIMED INTERVALS:
  This can sometimes be caused by the local oscillator of a scanning receiver, leaking out the antenna port and being radiated for a great distance.  Although the reverse isolation of an external preamp will almost always solve this problem,  it is likely the owner of the scanner desires to listen to numerous bands and a multiband preamp with adequate preselection may not be easy to find.  Assuming a low side mixer injection and an IF of 10.7 MHz, a receiver tuned to 154.900 MHz may radiate a 144.200 MHz signal.

NOISE PULSES- 3 seconds on, 3 seconds off (HF and VHF):
  IONIC BREEZE QUADRA air purifier units emit a nasty pulsing noise across many frequencies. I first discovered this after installing an HF wire antenna system over 100 feet away from two of these units and receiving S8-9 noise bursts from them.  I contacted The Sharper Image and was shipped two special design LOW RFI replacement models.  The result was the sudden appearance of nasty noise bursts on 50 MHz and a reduction of HF noise bursts to S3-4 from S8-9.  After explaining this continuing problem to The Sharper Image, this was their last response: "We are more than happy to assist you. Please be advised that our Ionic Breeze Quadra Silent Air Purifiers (SI637) have been tested and found to comply with the limits for a Class B digital device, pursuant to Part 15 of the FCC rules. We apoligize for any inconvenience this matter may have caused you."  End of story I guess?


INFORMATIONAL LINKS ABOUT NOISE CANCELLATION

TIMEWAVE ANC-4 information and data sheets

Experiences of Tony (I0JX) with the ANC-4

Experiences of Alan (3C5I) with the ANC-4

N6CA's 50 MHz Noise Cancelling Receiver



INFORMATIONAL LINKS ABOUT NOISE

VE3HLS RF NOISE IDENTIFICATION





 

Best 73 & Good 6m DX!
Bob, K1SIX (ex-WA1OUB)