Topband: The WD8DSB mini-flag antenna (LONG!)

Thu Feb 25 16:19:59 EST 2021

(Note:  what follows is a long, technical discussion about noise and “small” antennas.  I invoke some math and physics here, so if you are not comfortable with it, feel free to disregard or delete this e-mail.  I went through this exercise to help teach myself the limits here and maybe others might find it helpful or educational as well.)

My early tests with the WD8DSB mini-flag and some statements on the DX Engineering Web page led me to think that this antenna might be usable as a general-purpose receiving antenna with good sensitivity.  However, some of the comments posted on this reflector led me to reconsider my thinking.  To address this issue I decided to do some calculations and measurements, using my background in engineering.  Based on what I’ve now learned, I have a better understanding of the limits of using very small receiving antennas as replacements for larger antennas.  I thought I would share some of what I’ve done.  I should emphasize that this does not in any way make the WD8DSB mini-flag less useful as a tool for direction finding.  I still consider this antenna a terrific addition to my arsenal.

First, I’ll repeat what I said in an earlier post about the preamp I’ve been using with the mini-flag.  It’s a homebrew unit using a cascade of the UTO 511 and UTO 533 amplifier modules.  The datasheets give the noise figures of these amplifiers as 2.3 dB and 3.5 dB respectively, with gains of 16 dB and 17 dB  respectively.  In cascade they should yield 33 dB gain. 

Because the datasheets of these devices don’t specify operation below 5 MHz, I made my own measurements to see what they really do on 1.8 MHz.  I used my Elecraft P3 panadapter as a spectrum analyzer.  I’ve calibrated the scale of the P3 with the Elecraft XG3 calibrated signal source.  The XG3 provides selectable signal output levels of -107, -73, -33, or 0 dBm.  The specified accuracy is +/- 1 dBm at the 3 lower levels, so I’m pretty confident about the absolute calibration of the P3.

To assess the actual noise figure of my preamp on 1.8 MHz, I first did some calculations.  Johnson noise, which represents a thermal noise limit of a system, is -174 dBm in a 1 Hz bandwith (or -174 dBm/Hz) at a temperature of 300K, which is roughly room temperature.  For an explanation and derivation of this, see:  https://en.wikipedia.org/wiki/Johnson%E2%80%93Nyquist_noise.  This is a fundamental limit of physics (although when you get to optical wavelengths, quantum mechanical effects can become important).

Next I connected my preamp to my radio with a 50 ohm termination on the input of the preamp in place of the antenna and measured the noise floor in the P3.  The P3 gave me a value of about -118 dBm in a bandwidth of about 100 Hz.  The noise spectrum is flat so the noise in a 1 Hz bandwidth will be 20 dB (100 times) less, or -138 dBm/Hz.  To refer the output noise to the input of the preamp, where the antenna connects, I subtracted out the 33 dB preamp gain to yield an equivalent input noise of -171 dBm/Hz.  (I separately verified the 33 dB preamp gain on 1.8 MHz with the XG3 and P3).  This is only 3 dB greater than the theoretical thermal noise limit of -174 dBm/Hz.  That’s pretty darn close to what I would expect to see for a preamp with a stated 2.3 dB noise figure.  All of this so far is independent of any antenna.  It is all about the internal noise limits of my receiving electronics.

Next I set out to settle the question of whether the mini-flag (or any other low-gain antenna) could pick up enough external atmospheric noise to overwhelm the thermal noise limit that I just calculated/measured.   To do this requires knowledge of actual atmospheric noise levels.  You can look up numbers from various sources but those are only rough estimates and can vary greatly at different locations.  For my purposes it was best to try to measure it directly.

To measure atmospheric noise, I measured the daytime noise floor level in my P3 using my 160m transmit vertical antenna as the receiving antenna.  (On an average night, the noise comes up 5 to 10 dB from daytime levels at my QTH).   In reality this noise consists of a combination of some local man-made noise as well as atmospheric noise.  The P3 measured the daytime noise level as about -120 dBm/Hz.  I live in a suburban area and my numbers may or may not be representative of other suburban locations.  

To proceed further, I need to know the gain of my antennas.  I estimate the actual *average* gain of my transmit vertical is about -7 dBi, based on EZNEC modeling and some estimates of my ground loss.  Average gain is computed in EZNEC by averaging an antenna’s gain response over all of 3-dimensional hemispherical space.  It is necessary to do this type of analysis when doing atmospheric noise calculations because we have to account for noise contributions from all directions, not just the peak of the antenna’s main lobe.  This calculation assumes the atmospheric noise has a uniform distribution over all of space, which is probably not exactly true but we’ll make that assumption to get a ballpark result.  Note that an ideal lossless antenna would have an average gain of exactly 0 dBi, which, by definition, is the gain of an isoptropic radiator.  That results falls out from the concept of antenna gain averaged over all of space.

One caveat: the assumption of uniformly distributed noise is definitely not true for local man-made noise, which comes from specific directions and propagates via groundwave at essentially zero elevation angle.  The EZNEC average gain calculation uses a far-field antenna gain pattern and does not take groundwave into account.  

Next I created an EZNEC model of the DXE mini-flag and found it has an average gain of about -72 dBi on 1.8 MHz.  Because this is 65 dB lower than the average gain of my transmit vertical, the mini-flag should theoretically deliver 65 dB less atmospheric noise, or -185 dBm/Hz, to the receiver.  I make no claim these are super-accurate numbers but it’s the best I can come up with, at least at my location.  It does give an idea of the magnitude of the problem we’re dealing with when it comes to noise limits.

Because I already calculated/measured the thermal noise floor of my system as -171 dBm/Hz, the atmospheric noise floor level from the mini-flag at my QTH will be about 14 dB below thermal noise.  That means the atmospheric noise will be swamped by thermal noise in my electronics and not vice versa.  The higher atmospheric noise at night will be somewhat closer to thermal noise but probably not enough to change the conclusion here.  Thermal noise will limit one’s ability to hear very weak signals with this antenna.  This might be improved marginally with a preamp with a lower noise figure but 2.3 dB is already getting close to the limits of what can be achieved practically.  More preamp gain, without an improvement in noise figure, will not improve weak signal reception.  

This finding confirms what others have stated about very low gain receiving antennas being limited by thermal noise.  If you have an antenna with a different gain than the mini-flag, you can substitute its average gain into the calculation I just did.  

EZNEC says the gain of the mini-flag increases about 10 dB for each doubling in frequency above 1.8 MHz.  This means you start to have a chance of beating the thermal noise limit with this antenna at higher frequencies, maybe 7 MHz or above, depending on the ambient external noise environment.

For direction-finding purposes, you generally don’t need the ultimate in low-noise reception.  You only need the DF antenna to be good enough to hear the signal or interfering noise source well above whatever noise is in your receiver.  For this purpose, the WD8DSB mini-flag should work great in most situations.

I welcome any comments about this analysis and these results.

73, John W1FV

From: Don Kirk [mailto:wd8dsb at gmail.com] 
Sent: Wednesday, February 24, 2021 3:33 PM
To: john.kaufmann at verizon.net
Cc: TopBand List
Subject: Re: Topband: The WD8DSB mini-flag antenna

Hi John,

You started this mess (or maybe I did), and finally here is my response to a few of the questions you had and thanks so much for waiting in line for my response.

1)  As I have mentioned in a few other responses I suspect having the short 10 foot feedline helps to negate common mode noise and direct signal ingress into the feedline that often haunts us.  I do have one recommendation about pattern distortion as follows.  Check to make sure the peak and the null of the antenna are in agreement on the signal you are direction finding.  If you notice a slight skew (where they don't agree with each other), then move away from existing objects and this will correct that problem.  I sometimes notice a slight skew when in my backyard near my house (looks like the null shifts slightly from where it should be), and when I get out away from my house (I move to the sidewalk in front of my house) the slight pattern skew goes away.

2)  I don't know what the DX Engineering preamp noise figure is, and there definitely is no expectation from the designers standpoint that it's spectacular.  Even though I did most of the field testing of this preamp I have no idea what the amplifier part of the circuit is as DX Engineering tightly controls their designs.

3)  My antenna does not have a long mast, and it easily fits in the back seat of my very small 1996 Saturn Station Wagon.  It fits widthwise in my backseat, and I have to imagine it will do the same in almost any car since my car is about as small as they come.  Looks like the DX Engineering version of my antenna has a longer mast, and that can easily be shortened if it helps you fit the antenna into other areas of your vehicle.  I only use the extension on my antenna if I'm doing some stationary tests in my backyard, otherwise I use it as shown on the cover of QST.

I Really appreciate you jumping in so quickly on the antenna build and reporting some of your results.  Sometimes when I have a few minutes I will go outside and just play with the antenna and marvel on how simple it is, but how well it works. 

I hope I have now covered the majority of everyones questions and comments.  If there are other comments or questions we probably should move this discussion over to the RFI Reflector or e-mail me direct.

73,

Don (wd8sb)

On Tue, Feb 23, 2021 at 1:26 PM John Kaufmann via Topband <topband at contesting.com> wrote:

Some of you may have seen the article by WD8DSB in the latest issue of QST.
I believe WD8DSB is on this reflector.  His article describes a mini-flag
antenna that can be used for direction-finding.  The neat thing about this
antenna, besides its compact size, is that it is unidirectional and is very
broadband.  It works from the AM BCB through 10m.  It produces a sharp null
off the back which allows you to determine signal direction without the
direction ambiguity you get with a conventional unterminated loop.

DX Engineering is producing this antenna as a kit, along with a companion
preamp.  (Disclaimer:  I have no affiliation or commercial interest in DX
Engineering).  See:  https://www.dxengineering.com/parts/dxe-noiseloop.  I
just bought the flag kit last week and finished assembling it this past
weekend.  I see today that the kit is now back-ordered until April so it was
good that I ordered it as soon as I saw the QST article.  

It took me about 3 hours to assemble the mini-flag even though the DXE Web
site says it can be done in 1-2 hours.  There is a bit of fussy mechanical
assembly involved in getting the symmetry and dimensions just right,
although it's not hard work.  The flag is 42 inches wide and 21 inches tall.
The DXE version of the antenna has slightly smaller dimensions than those
given in the QST article, which results in a small reduction in gain, which
doesn't really matter, but the pattern is the same.

I did some testing of the mini-flag in the AM BCB.  The gain is very
low--about -65 dBi on 160m--so it needs a good preamp.  I used a homebrew
preamp made up of a couple of MMIC's that produce about 35 dB of gain.  The
DXE preamp for this antenna won't be available until April.  On the higher
frequencies, less preamp gain is needed because the gain of the mini-flag
increases with frequency.

My initial tests indicate this antenna clearly works.  By rotating the flag
for the deepest null, I could nail the heading an AM BCB station to a few
degrees.  

This antenna could also be used as directional receiving antenna on its own
Although it is not hugely directive, it can be rotated easily to peak or
null signals or noise, and it is better than a conventional unterminated
loop.  It has essentially the same RDF as other larger flag or pennant
antennas but is obviously far more compact.

This is a nice contribution by WD8DSB.  Now I have to go off with the
mini-flag and chase some local noise sources that have been plaguing me this
winter on the low bands.

73, John W1FV

_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector