To some extent, the project was inspired by KK5JY’s Loop on Ground (LoG).

This article presents measurements and the three terminal equivalent impedance model.

Above is the three terminal equivalent impedance model. Elements Z1, Z2 and Z3 are derived from measurements Za, Zb, and ZC as discussed at Find three terminal equivalent circuit for an antenna system.

A NanoVNA was calibrated for a fixture designed for such measurements, and a scan from 1-30MHz for each of Za, Zb, and ZC was captured and saved to the SD card.

Above is the measurement of ZC, the common mode impedance Zcm which is of interest as it informs strategies for minimising injection of common mode feed line current into the coax interior.

Whilst a small loop in air tends to have a very high common mode impedance, the LiG is quite the opposite.

Above is a plot of the Za and Zb measurements. Though the wire geometry is quite symmetric, the electrical symmetry is not perfect, probably due to less than uniform soil characteristic over the antenna site.

Above is a calculation of the various elements at 3.61MHz.

A work in progress…

]]>To some extent, the project was inspired by KK5JY’s Loop on Ground (LoG).

This article presents a comparison of Signal to Noise Degradation metric (see Signal to noise degradation (SND) concept) for both antennas, the common elements being:

- based on NEC-5.0 models (as detailed in earlier LiG articles);
- soil parameters used are σ=0.01, εr=20 (calibrated to measurements at the LiG test site);

The LoG models are for 4.6m sides, 2mm wire at 10mm height above ground, and an approximately optimal 450Ω:50Ω transformer.

The LiG models are for 3.0m sides, 2mm wire at 20mm below ground, and an approximately optimal 200Ω:50Ω transformer.

Above is a plot of SND for both antennas over the range 0.5-15MHz.

They are quite different responses.

It can be seen that from about 3-11MHz there is not a lot of difference between the antennas, but the LiG degrades slowly above 11MHz whereas the LoG degrades quickly below 3MHz.

A work in progress…

]]>The Loop in Ground project is about a receive only antenna for low HF, but usable from MF to HF. The objective is an antenna of that is small, low profile, and can be located outside the zone where evanescent modes dominate around noise current carrying conductors, like house wiring to minimise noise pickup.

The antenna comprises a square loop of 3m sides of 2mm bare copper wire, buried 20mm in the soil.

This article reports measurement of feed point impedance and a ‘calibrated’ NEC-5.0 model.

Australia is experiencing a La Nina weather pattern this spring / summer, and it has rained and rained and rained. The ground has varied from saturated to nearly saturated, so measurements are a little atypical. Several measurements have been made, and the ones reported here are at a less saturated time, but the same broad pattern has been observed with all measurements.

Whilst the measurements to not calibrate exactly with the NEC model, they are quite close, and the model used here is adjusted for better reconciliation in the range 1-5MHz. The soil parameters used are σ=0.01, εr=20, which are suggestive of very ‘good’ soil.

The calibrated NEC model is probably the best predictor of behavior that other constructors might experience.

Key to the performance is system gain which depends greatly on MismatchLoss, which is quite dependent on the load impedance presented to the LIG. A spreadsheet model was constructed from the NEC feed point impedance and expected ambient noise (per ITU P.372-14) and Signal to Noise Degradation (SND) calculated (Signal to noise degradation (SND) concept).

Provision made for a transformer as part of the system model. So parameters to the spreadsheet model included:

- NEC-5.0 model of feed point impedance and average power gain of a 3m a side square loop of 2mm HDC buried 20mm in soil σ=0.01, εr=20;
- ITU P.372-14 ambient noise precinct ( Rural used here);
- transformer ratio (2:1 turns, 4:1 impedance, with 1dB loss used here);
- receiver Noise Figure (6dB used here); and
- transmission line (10m of Belden 8215 RG-6/U used here).

The 75Ω feed line is chosen as a low cost feed line even though it is not 50Ω, in a real implementation with a buried feed line, flooded RG-6/U or RG-11/U might be very practical choices.

Above is the graph for the Rural precinct as mentioned, SND is in blue. It can be seen that is less than 3dB from 1.0-13MHz.

If you live in a very noisy neighborhood, the Residential precinct may be more appropriate to your environment.

Above is the graph for the Residential precinct as mentioned, SND is in blue. It can be seen that is less than 3dB from 1.0-28MHz.

Most designs of small Loop-on-Ground antennas use higher transformer ratios, and they may or may not be appropriate, but for this LiG in the specified soil, it is clear from the spreadsheet model that choosing other integer ratio transformers gives poorer SND response.

An IC-R20 was used for a listening test as with near zero length feed line, system gain is not significantly affected by common mode feed line contribution. Tests were conducted at 19:00 local time on the MW BC band, 80m, 40m and 20m bands. In all cases, external noise was audibly greater than internal noise (assessed with a 50Ω termination, some very minor contribution from the termination), and the receiver performed pretty much in line with prediction. There was no evidence to question the predictive models for Rural in this location. It was interesting that even an hour before sunset, many MW broadcast stations were heard at very good strength even though this location is not the the formal service area of any MW broadcast signals, stations in Canberra some 200km distant were at very good strength and excellent quality (steady signal, no buzz or other significant intermodulation distortion.

A work in progress…

]]>The antenna comprises a square loop of 3m sides of 2mm bare copper wire, buried 20mm in the soil.

Above is the site marked out for earthworks, but excavation of a narrow slot 25mm deep. On the far side of the loop is an already installed plastic irrigation valve box for the transformer.

Above, the excavation implement… an attachment for the 62cc weed wacker which is designed for cutting neat vertical edges in the grass along paths etc. It was not very expensive, so seemed worth a trial. It worked very well.

Above, wire laid and slot backfilled. This will be watered and rolled with the mower over coming weeks to grow grass roots and settle the soil.

Above, ABS tent pegs were used to secure the corners of the wire loop.

In a couple of weeks, the feed point impedance will be measured and compared to models.

After one week of settling, including lots of rain, a quick preview with a hand held receiver is promising.

A work in progress…

]]>This article explains a little of the detail behind the graph.

The graph is based on a series of NEC-5.0 models of the loop in ground antenna. Key model parameters are:

- 3m a side;
- ‘average’ soil (σ=0.005, εr=13);
- depth=0.02m; and
- frequency 0.5 to 10MHz in 0.1MHz increments.

The models were scripted by a PERL script, and the output parsed with a Python script to extract feed point Z, structure efficiency, and average power gain (corrected to 4πsr).

The summarised NEC data was imported into a spreadsheet and an approximate model of the system built, comprising:

- Receiver input impedance 50+j0Ω;
- a length of transmission line (10m of Belden 8215 RG6/U);
- an ideal transformer (4:1);
- source impedance derived from the NEC data.

Calculation includes:

- transmission line loss and impedance transformation;
- transformer assumed ideal plus an allowance for transformer loss (1dB);
- mismatch loss; and
- average antenna gain.

Above is an extract of the spreadsheet.

Mismatch loss is an important element of the system behavior. A convenient place at which to calculate mismatch loss is the feed point of the loop in ground.

Above is a plot of the loop feed point impedance, the source impedance in the receive scenario.

Above is a plot of the loop load impedance, the receiver impedance transformed by transmission line and transformer. The varying impedance is a result of using 75Ω line.

The combination of these allows us to calculate mismatch loss.

Above is a plot of the calculated mismatch loss which must be added in to the system gain model.

From the system model, and an estimate of ambient noise from ITU-R P.372-14, we can calculate SND.

Above is a plot of SND.

Note that P.372-14 is based on a survey with short vertical monopole antennas, so it is likely to overestimate noise received by a horizontally polarised antenna (and therefore the SND estimate will be low).

Antenna performance is sensitive to soil parameters, especially those close to the surface and subject to variation with recent rainfall etc.

This is after all a feasibility study, and within acceptable uncertainty, the antenna system would seem to be feasible for low HF and even 160m receive.

]]>Let’s take ambient noise as Rural precinct in ITU-P.372-14.

An NEC-5.0 model of the 3m a side LiG gives average gain -37.18dBi. An allowance of 2.7dB of feed loss covers actual feed line loss and mismatch loss.

Above, calculated SND is 0.9dB. For this scenario (ambient noise and antenna system), the receiver S/N is 0.6dB worse than the off-air or intrinsic S/N ration. For Residential precinct ambient noise, SND is less at 0.3dB.

The above graph shows the system behavior over 0.5-10MHz, it is a combination of the effects of noise distribution; antenna gain; mismatch; transformer and feedline losses; and receiver internal noise.

I have been asked by several correspondents why I used #43 when the consensus of online experts is that #75 is a clearly better choice for the application.

Let me say that almost all such articles and posts:

- are absent any quantitative measurement of their proposed design;
- they tend to use medium to large toroids; and
- the few that expose their design calcs treat permeability as a real number that is independent of frequency.

#75 mix is a high permeability MnZn ferrite and subject to dimensional resonance in the frequency range of interest for this application, a problem exacerbated by using larger cores.

Permeability is a complex quantity that is frequency dependent and any analysis that pretends otherwise is not soundly based.

Above is a plot of Fair-rite’s published permeability complex data for #75. Note the rapid descent of the real component µ’ to zero, a strong sign of dimensional resonance in the measured core. Dimensional resonance is a function of cross section dimensions and larger cores may have dimensional resonance evident well below 1MHz.

The chosen small binocular #43 cores do not exhibit dimensional resonance in the frequency range of interest, the windings are a practical number of turns, the performance is predicted by a simple model, and the measured performance documented in the article reconciles well with the predictive model.

- Snelling, E C. Soft ferrites properties and applications. Iliffe books 1969.

To test the prototype, I thought it an interesting exercise to use a low end rx only SDR for the instrumentation, providing a graphic quantitative measure of performance that is within the reach of most hams.

The first device trialled was a RTL-SDR v3 dongle with Sdrsharp (SDR#) software under windows, a very low cost option ($40). I was unable to find meaningful NF specifications or end user measurements for the thing in direct sampling mode.

The v3 incorporates hardware for direct sampling, so allowing input up to ~30MHz. Well, it might have some of the hardware, but it lacks preselector filters which are essential to alias free behavior and the option was quickly dismissed as not suitable for my intended use from 0.5-15MHz.

So, that kicked off a review of ‘mid price’ rx only SDR, starting with a documentation review. Again, finding meaningful NF specifications is very difficult, but some state sensitivity or MDS in ARRL terms (an appeal to the ham market no doubt)… but a worry in its own right as these figures are almost always based on nominal 500Hz bandwidth, not the actual quivalent noise bandwidth of the receiver and the error can be significant (ARRL Test Procedures Manual (Rev L) – Noise Figure calculation).

I settled on an SDRPLAY RSP2 and did a more detailed search for issues, and lo and behold, SDRPLAY advice on their website of loss of several consignments of the things in the supply chain, and to protect someone (themselves mainly) they have “blacklisted” the stolen serial numbers to “render them worthless” (SDRPLAY) in the hands of the user. Their strategy is to attack the easily identified soft target, each buyer, irrespective of whether the buyer knew they were buying stolen goods (shades of the FTDI debacle here).

The next candidate is very new technology, very interesting, but a greater risk than more mature technology. That can wait for a little more user experience (this sounds like the young man who waited for SDR technologies to plateau).

So with a little delay low end rx only SDR are dismissed as a low risk low cost instrument for the project’s measurements, and more conventional devices will be used.

Work continues…

]]>The characteristic of typical medium µ ferrite mixes, particularly NiZn, are well suited to this application.

This article continues with the design discussed at BN43-2402 balun example, but using a BN43-202 with 5t primary and 10t secondary for a nominal 1:4 50:200Ω transformer (though at high ratios, the transformation is only nominal).

Lets consider a couple of simple starting points for low end and high end rolloff.

A simple model for these devices with low flux leakage is an ideal transformer with primary shunted by the magnetising impedance. To obtain low InsertionVSWR, we want the magnetising impedance in shunt with 50+j0Ω to have a low equivalent VSWR.

Typically complex permeability changes in-band, and although it tends to decrease, increasing frequency means that the critical point for magnetising impedance is the low end.

At the high end, transformation departs from ideal usually when the length of wire in a winding exceeds about 15°.

A small core makes for short windings to obtain high frequency performance, and sufficient turns are needed for low end… but not too many as it restricts the high end.

There are lots of rules of thumb for minimum magnetising impedance, most treat the inductor as an ideal inductor and these ferrites are not that.

A quick analysis using the method in BN43-2402 balun example hints that a 5t primary on a BN43-202 core is probably good enough down to 0.5MHz, depending on one’s limit for InsertionVSWR. We are not being too fussy here… this is not an application that demands InsertionVSWR < 1.5.

Above is a plot of expected magnetising R and X for a 5t winding using my common mode choke design tool. Z at 0.5MHz is 3.6+j172Ω, or Y=0.0001216-j0.005811S. (If your design tools are not giving you similar values, you might consider validating them.) Adding the shunt 50Ω (Y=0.02), we get Yt=0.0201216-j0.005811S, and plugging that in to calculate VSWR…

…we have InsertionVSWR=1.33, that is fine for our application. This is only a rough indication of suitability, the final test is a VNA sweep of a prototype.

The next step is to make one up and measure it.

Above, the prototype with terminals for antenna field test.

Above is a VNA sweep of a prototype wound with 0.25mm ECW.

The load resistance is a resistor with DC resistance of 148.6Ω in series with the VNA 50Ω input port. InsertionLoss at less than 0.2dB is quite within expectation.

Above is a disaggregation of InsertionLoss into TransmissionLoss (or simply Loss) and MismatchLoss. It can be seen that MismatchLoss is worse at the lowest and highest frequencies, it is what limits the VSWR bandwidth.

The response is quite acceptable for the application (a very low gain, ie lossy, antenna). The compensated transformer is slightly better at the high end.

The transformer was revised to add a centre tap to the secondary winding for optional grounding for feed line common mode current mitigation.

The transformer is wound by forming a trefoil bundle of wires and winding five turns, but one wire is split out at the start end to tuck a further half turn in, and at the other end, the same wire is split out of the last half turn so that one five turn winding has its terminals at one end, and the other two windings have their ends at the other end so that they can be joined start to finish to produce a centre tapped secondary of 10 turns. This construction is to minimise leakage reactance, the enemy of broadband transformers.

Above is the equivalent series inductance of the transformer with short circuit termination.

Above is the measured InsertionVSWR of the uncompensated transformer.

Above is a calibrated Simsmith model of the rewound transformer with optimal compensation. A 33pF capacitor was added to the prototype.

Above, the revised transformer assembly with compensation capacitor installed.

Above is InsertionVSWR for the revised transformer with 200Ω load.

- Duffy, O. 2015. A method for estimating the impedance of a ferrite cored toroidal inductor at RF. https://owenduffy.net/files/EstimateZFerriteToroidInductor.pdf.
- Snelling, E C. Soft ferrites properties and applications. Iliffe books 1969.

]]>

This article documents the selection of the trial loop in ground configuration as a development from the loop on ground antenna (KK5JY).

The baseline is a minor variation of a design by KK5JY, a 15′ square loop 20mm above average ground, with 9:1 transformer and 50Ω load middle of one side.

Above is a plot of feed point impedance when the loop is driven. At 3.6MHz, the source impedance for a rx system is 43+j852Ω, and the mismatch loss to a 450Ω load is 11.0dB, a direct contribution to Antenna Factor (AF).

Note that these values are quite dependent on model parameters such as wire diameter, height above ground, soil type etc. NEC-2 may have issues with some aspects of the model, so it may not produce similar results to the NEC-4 models

For excitation being a plane wave at elevation 45° from direction of maximum response, AF is calculated at the transformer secondary 50Ω load to be 22.7dB.

Above is a plot of feed point impedance when the loop is driven. At 3.6MHz, the source impedance for a rx system is 240+j136Ω, and the mismatch loss to a 200Ω load is 0.4dB, a direct contribution of Antenna Factor (AF). Note that in practice source impedance depends on soil parameters, and can be expected to vary with moisture content and so the impedance matching solution is a very approximate solution for a untuned loop.

For excitation being a plane wave at elevation 45°, AF is calculated at the transformer secondary 50Ω load to be 13.9dB.

The models are sensitive to ground parameters.

The smaller shallow buried loop has AF 8.8dB better than the baseline configuration.

AF=13.9dB @ 3.6MHz implies Gain=-32.5dB (@45° elevation) which at first might seem unusable, but when the ambient noise figure (Fam) is more than 50dB, such a low gain antenna system still captures sufficient external noise to dominate receiver internal noise and there is very little degradation in achieved S/N.

]]>