EMC Compliance Dipole, attn:radio8z

Robert,

It is a shame that all the information which was contributed to that site is now unavailable to us. How much would it cost them to keep the forum up as a read only library? But... their site, their rules.

Anyway, I don't keep copies of posts so all I can do is try to reconstruct this. As I recall I posted a few times on this subject so if questions remain please ask.

I have had good results by using a TV antenna balun core which was removed from a snap together balun available for a couple of $. These are usually binocular cores with the windings placed around the center part. They are usually wound for a 300 to 75 Ohm reduction in impedance but I have encountered 1:1 units. If you use a half wave folded dipole the balun can be used as is to match to 75 Ohms but if you want to match to 50 Ohms it will need to be rewound.

The rule of thumb for this type of transformer is that the self inductance should be at least five times the operating impedance at the frequency of interest. What this means is that the number of turns on each winding should be about 5 to 10 turns but this is not critical. What is critical is the turns ratio.

The impedance ratio is proportional to the square of the turns ratio. If you want to transform from a 72 ohm dipole (not folded) to 50 ohms then the turns ratio is SQRT(72/50) with more turns on the antenna side. It becomes a matter of finding integer number of turns in the range of 5 to 10 turns or so which gives this ratio.

One such ratio is 5T receiver side and 6T antenna side. Another would be 10T to 12T.

The impedance seen looking into the receiver side would then be (5/6)^2 X 72 = 50 ohms.

There will be loss due to core loss and leakage inductance but this would be nearly constant and can be accounted in the antenna calibration.

If the purpose is to balance the antenna currents without impedance transformation then a suggestion posted by John (WCDX) which was to place a choke core around the coax at the antenna may work.

Hope this helps.

Neil

 Accurate measurement of field intensity requires knowledge of the voltage or power produced at the output terminals of the receive antenna, relative to the arriving field.  This is commonly termed "Antenna Factor," or AF.

That answer will depend on the frequency, the configuration of the receive antenna, its orientation in space with respect to the arriving field, and for VHF and above, its elevation above the earth -- and other factors.  This all takes takes a bit of study to determine.

The first link below develops this for a particular line of spectrum analyzers.  The comments about the antenna given on pages 9 & 10 there are useful.

http://www.home.agilent.com/upload/cmc_upload/All/AD2003_module3.pdf

Here is a link to a "standard antenna" based on those used at the NIST:

http://w4neq.com/htm/nist.htm

Commercial VHF field intensity meters such as the Potomac Instruments FIM-71 included the antenna, and the documentation provided with the meter accounted for the calibration of those meters with those antennas, in a standard field.

i have a spec analyzer and a known part 15 certified legal signal source to calibrate it against. just a matter of building the antenna and calibrating it to my scope using the known signal source

"just a matter of building the antenna and calibrating it to my scope using the known signal source"

That is one approach, but in that case it wouldn't matter what the receive antenna was, or whether or not it used the right balun -- as long as the receive system used the same hardware configuration and test procedure during both the reference "calibration," and when checking the system under test.

Another possible concern is that the reference source used in this process could produce less than 250 µV/m at a distance of three meters, which would be compliant with FCC 15.239, but not at the upper limit.

Unfortunately, the only way to measure fields ~accurately is to use equipment designed and calibrated to do that.

Unless a quantity is defined by the measurement there is always an error involved in measurements.

Measurements can be very useful even if they contain errors as long as the maximum error is known. For example, a 10 volt meter with a maximum error specification of 3% full scale reads 5.0 volts. If the error spec. is true then the actual voltage is between 5.0V +- (0.03 X 10V) which gives that the true voltage is between 4.7 and 5.3 volts.

Now suppose Robert has a calibration source which is known within +- 5% and he calibrates his meter to this. The error is now +-5% at this signal strength so if he measures a signal which is nearly this value the error range is essentially +-5%. Unfortunately, the error in measurement for signals not near the calibration value may not be known and may be greater than 5% due to non-linearities in the meter.

So, here's a thought for which I hope Rich will give some input. Suppose Robert knows of an FM station a certain distance from him and he knows their ERP. He takes several measurements at his location in different spots and averages the readings in an attempt to cancel the effects of multipath and other position dependent factors. He then uses this number as the calibration point for his meter so he can measure his part 15 FM field strength.

Rich, is it possible to reasonably estimate the error range in this type of situation? Though Robert's measurements may not be precise it may be useful if the error range is known.

For example, if the error range is +-50% at 150 uV/m and he measures the field strength of his transmitter to be 150 uV/m at 3m then the actual reading would be between 75 and 225 uV/m which would be useful in determining compliance providing the error limits of the calibration are known with confidence.

Is this realistic in this scenario?

Neil

"...Rich, is it possible to reasonably estimate the error range in this type of situation? Though Robert's measurements may not be precise it may be useful if the error range is known. ..."

This method would be subject to a considerable number of unknowns, unfortunately.  The VHF wavelengths used for FM broadcasting are much more subject to obstruction losses and reflections than those used in AM broadcasting.  Such effects at the receive antenna on the signal using a line-of-sight path to the transmit antenna cannot easily be predicted.

Below are some Powerpoint clips I wrote before I retired, showing some of these conditions.  The clips were from a microwave talk I gave, but the same principles apply at VHF.

Here is a page from a microwave system planning handbook, showing how the received signal varies with Fresnel clearance:

Another factor is the azimuth and elevation plane radiation from the transmit antenna used by the FM station, which can vary as much as 20 dB depending on the radial path to the receive point.  This is a function of the transmit antenna configuration, and re-radiation by the tower supporting that antenna.  Almost no FM broadcast station radiates a truly omnidirectional signal.

The FCC uses a statistical approach when predicting the fields of an FM broadcast station, giving a result for 50% of locations, 50% of the time.

Short paths are much less vulnerable to obstuctions and reflections, which probably is why the FCC uses 3 meters as the reference distance in §15.239.

Thanks, Rich, for answering as you did. I knew this would be complicated but was hoping there was something worth pursuing.

Back in the day when FM receivers were being considered for car applications I was working for General Motors and saw a study they did using the then new main frame computer to plot the field strength effects from metal street poles along a boulevard. From the spatial variations it appeared that it wouldn't work but perhaps due to very good AGC it does (or maybe the model was wrong).

Neil

http://www.hamradio.me/antennas/coax-velocity-factor.html