** FSL
Antenna Design Optimization**

** All-out
Experimentation to Determine Weak-Signal Performance Potential**

By Gary DeBock, Puyallup, WA, USA
March 2012

__Introduction__ The
publication of Graham Maynard’s ferrite sleeve antenna article in March of 2011
kicked off a torrent of experimentation in the Ultralight radio enthusiast
group, as hobbyists quickly discovered the DXing potential of the new design.
Massive orders were placed for Russian surplus ferrite and Chinese Litz wire,
while wholesale lots of swimming floatation aides were drafted into service as ferrite
padding material. Each individual experimenter had his own ideas about FSL
design, and backed them up with serious financial outlays. Multiple design
optimization questions went unanswered in the process, but since the new
antennas seemed to be working fairly well for everyone, attempts to bridge
opinion gaps were few and far between.

One school of thought held that coil diameter was the most
important gain factor, while another was that ferrite size and weight was
paramount. Ferrite rod versus ferrite bar opinions were also diverse, and even
Litz wire variables became part of the controversy. Any newcomer wishing to
build an FSL antenna had no shortage of advice—the only problem was which
advice to follow.

This experimental free-for-all continued for a full year, with
various (and diverse) FSL antenna designs published in the process. The
performance variables needed to be sorted out in detailed A/B experimentation,
conducted by an experimenter with a completely open mind, and with no axe to
grind. Although all of my previously-published FSL designs had been based on
ferrite rod components, I recalled that a large supply of Russian surplus 100mm
x 20mm x 3mm ferrite bars had been received here for some time, and had gone
unused. In a serious effort to bridge opinion gaps and provide complete clarity
to the design factors influencing any FSL’s weak-signal DXing performance, it
was time to wipe the slate clean and test out all possible variations of the
antenna-- letting the chips fall where they may.

__Experimental
Objectives__ The ambitious agenda of the testing included all experimentation
necessary to judge the weak-signal performance of ferrite rods compared to
ferrite bars, to judge the weak-signal performance of large-diameter FSL models
compared to smaller-diameter models with more ferrite weight, and if at all
possible, to establish a design formula which could accurately predict the
weak-signal DXing performance of any FSL antenna model before it was
constructed (as well as rate the effectiveness of any existing model). These
were challenging goals, but I was determined to continue the experimentation
(and build as many test models as necessary) until complete clarity of all the
design optimization factors was reached.

In order to carry out this agenda seven new FSL test models were built, including a diverse variety of ferrite bar models, short ferrite rod models and long ferrite rod models. Deliberate use was made of common material like Russian surplus ferrite (of 400 permeability), single coils composed of 660/46 Litz wire, a common 381 pf variable capacitor and standard PVC-pipe frames. Each test model was constructed with a coil diameter shared by other models using dissimilar ferrite material (bars, short rods or long rods) so that the performance differences between each ferrite type would become obvious. Finally, all test models would be judged in relative weak-signal performance outdoors in an open-air test range, competing with each other in the reception of at least four fringe daytime DX test stations on frequencies throughout the AM band. During this testing relative-reception MP3’s would be recorded for each antenna match up, with the antennas switched in the middle of the recording (within 7 seconds, so that the fringe station’s signal level would not change). Both FSL test models would be placed on a 5’ PVC base for equal elevation, and the antennas would be separated by at least 50 feet so that there would be no possibility of interaction. Additionally, in cases where a relative-strength deadlock was apparent between two FSL test models, a reference antenna (4’ sided PVC air core loop) would be used to check if either test model had better performance against the reference antenna. In summary, every possible preparation was made to provide equal testing parameters for accurate relative performance results.

__Three Diameter Classes__ The test
models were divided into 3”, 5” and 7” diameter classes. The 3” class included
a 200mm long-rod model, a 100mm bar model, and a 65mm short-rod model. The 5” class included a 140mm long-rod model,
a 100mm bar model, and a 65mm short-rod model. Finally, the 7” diameter class
included a 140mm long-rod model and a 100mm bar model.

Each diameter class would first have relative performance
tests against other antennas of the same class, then the top performers in each
class would have match ups against each other. With the diverse mix of rods,
bars and diameters in the test models, it was presumed that obvious patterns of
superior performance would emerge—and they quickly did.

__Ferrite
Rods or Bars—Which are Better?__ In each diameter class a clear pattern
emerged immediately, with every model having a longer ferrite material (rod or
bar) beating out every model with a shorter ferrite material (rod or bar). The
thickness or weight of the ferrite material was irrelevant—even though the 65mm
rod models had greater weight and ferrite thickness than the 100mm bar models,
they lost out to the bar models every time. But this had nothing to do with any
superiority of the bar models—to even the score, the 100mm bar models always
lost out to the 140mm and 200mm rod models. So the question of rods or bars had
finally been answered—whichever ferrite material was longer had the superior
DXing performance. Even a 35mm or 40mm advantage in ferrite length was enough
to provide a noticeable edge in weak-signal performance, for a given coil
diameter.

__Smaller
Diameter FSL Beats Out Two Larger Diameter Models__ The 65mm
short-rod FSL models proved to be experimental duds, losing out to every other
FSL model their respective diameter classes. But before final elimination from
the competition, the 5” short-rod (65mm) model was matched against the 3.5”
long rod (200mm) model—and was clearly inferior to it. The 3.5” long-rod
(200mm) model not only outpaced the 5” short-rod (65mm) model, but clearly
outperformed the 5” bar model (100mm) as well. This was the first experimental
proof that a smaller diameter FSL could outperform a larger diameter FSL—and
another indication that ferrite length extremely important to overall DXing
performance.

__The “Odd Couple” DXing
Deadlock__ With the overachieving 3.5” long-rod (200mm) model running
roughshod over two of the 5” diameter FSL models, it was given a shot at the 7”
diameter bar (100mm) model. In what was probably the most important FSL match
up in the entire project, the two completely different antennas deadlocked in
performance on all 5 of the test signals. This was the first weak-signal
deadlock in 10 different FSL match ups, so it was quite a surprise. Both antennas
were then tested against the 4’ air core reference loop—and both FSL’s
deadlocked with it, also! There was now no doubt that the DXing performance of
these two radically different models was identical, which certainly was a
puzzle considering their polar-opposite design.

But when I looked at both models closely, I finally noticed a
very peculiar relationship between them… the 3.5” long-rod model’s ferrite
length (200mm) was twice that of the 7” bar model (100mm), while the 7” bar
model’s coil diameter was twice that of the 3.5” long-rod model! Could this
mean that ferrite length was equally important to coil diameter in determining
an FSL’s overall performance? It certainly was in these two FSL models—the two
models each had their own design advantage, which was resulting in a DXing
deadlock!

After stumbling on this concept, I quickly made up a
provisional formula to predict an FSL’s overall DXing performance: In
comparison to other models having equal component parameters (Litz wire type, coil
orientation, and ferrite permeability), an FSL’s coil diameter times its
ferrite length would determine its overall weak-signal capability. Since every current
FSL experimenter used inches to measure coil diameter and millimeters to measure
the Russian surplus ferrite length, I decided to risk the wrath of purists and
simply multiply the two dissimilar measurements together in my formula, which
in the case of the 7” bar FSL worked out to an equation of 7 x 100 = 700
points. In the case of the 3.5” long-rod FSL the equation worked out to 3.5 x
200 = 700 points, an identical figure.

But wait—I knew that the 5” Mini FSL model (composed of the
140mm long ferrite rods) had also deadlocked with the 4’ air core reference
loop, which meant that its performance was also identical to these two
deadlocking FSL models. Here was now yet another FSL model which had a design
pretty much in the middle of the two radically different models, with equal
performance. Theoretically, it should then have the same performance “score” as
the other two FSL models in the new equation. It was with some excitement that
I quickly did the math… 5 x 140 = 700! All three performance scores agreed with
each other, and I was thrilled that the new formula had accurately predicted
the 5” Mini-FSL’s participation in a 3-way DXing deadlock (as shown in the
large group photo of the three FSL’s together, at the beginning of this
article).

__Further
Confirmation of the FSL Performance Formula’s Accuracy__ When
considering the results of all the FSL test model match ups conducted
previously, I finally realized that the new performance formula could have
accurately predicted all of the experimental results before any of the test
models had even been constructed! But there was one test model match up that
had not been tried yet, which was to provide the final confirmation of the
formula’s accuracy. The 5” short-rod (65mm) model was an under-performer, and
had never been matched against the 3” bar (100mm) model. However, the apparent performance
formula scores of these two models were extremely close (325 and 300,
respectively), and it was time to see how they shook out in an actual
weak-signal Shootout. Before running this final confirmation test I carefully
measured the actual diameters of the two models, though—and was very lucky to
find that the 3” bar model was actually a little bit larger than 3”, having a
coil diameter of 3.25”…giving it a dead-even performance score with the 5”
short-rod model! I didn’t even need to modify either model, to have another two
FSL antennas which (theoretically) should be equal in weak-signal performance,
according to the new design formula.

Once again, it was with some excitement that I set up both FSL
test models on their 5’ PVC stands in the back yard test range, eager to see if
the performance formula had once again clearly predicted the final result.
After receiving the very first fringe signal (540-Burien TIS) on both models I
knew that I could have saved my trouble… switching between the two antennas in
the MP3 recording produced no change in signal quality at all. It was another
rare performance deadlock of two diverse FSL test models--- accurately
predicted by the new performance formula.

__Experimental Wrap Up__ The
objectives of the testing had all been met, and accurate weak-signal design
factors for the FSL antenna had finally been determined. To summarize, here
were the new discoveries:

1) There is no weak-signal advantage to be gained by using either
ferrite bars or ferrite rods. The only performance factor related to ferrite is
the ** length**
of the ferrite material, whether it is in bar or rod form. For a given FSL coil
diameter, as long as the ferrite permeability is identical, longer ferrite
material will always outperform shorter ferrite material, and the weight and
thickness of the ferrite material is irrelevant to weak-signal performance
(this was proven in seven FSL match ups).

2) The ** length** of the ferrite material in an FSL antenna is equally
important to its coil diameter in providing weak-signal performance. An FSL
with twice the ferrite length of another FSL model can still match its
weak-signal performance even with a coil diameter half the size of the other
model (proven in the case of the 3.5” long-rod model and the 7” bar model).

3) Assuming that all component parameters (Litz wire type, coil
orientation and ferrite permeability) are identical between two FSL antenna
models, the weak signal performance of any FSL antenna in relation to another
can be accurately determined by multiplying the coil diameter times the length
of the ferrite material in the sleeve. This product gives a “Performance Score”
which can accurately predict the weak-signal reception capability of any FSL
antenna in comparison to other FSL models having the same component parameters (proven
in three FSL match ups here, with identical signals in switched MP3 recordings).

__Summary__ The ability
to accurately determine an FSL antenna’s DXing performance even before
construction opens up a fascinating new perspective on design improvements.
Aided by the new performance design formula, a hobbyist can concentrate on building
highly effective models at a minimum cost and weight. The FSL antenna’s great
advantage in relation to other antenna types is its ability to provide low
noise, high gain performance from an extremely compact size—and this new design
formula emphasizes compact effectiveness, proving that ferrite length is
equally important to coil diameter in DXing performance. Future FSL antenna
models that combine both ferrite length and coil size are the wave of the
future, and now that definite answers have been found for most of the design
controversies, we can proceed to design and build our new models with complete
assurance of success!

73 and Best Wishes,

Gary DeBock

**Relative Strength “Shootout” MP3 Links**

For determination of the design factors influencing an FSL
antenna’s weak-signal performance, four of the FSL test model match ups were
considered very important. These four match ups (with the design factors
clarified by the results) were as follows:

1) 3.25” Bar model (100mm ferrite length) outperforms 3.25”
Short-rod model (65 mm ferrite length)

*Longer ferrite sleeve material
is superior in performance to shorter ferrite sleeve material, even if the
shorter material is heavier and thicker than the longer material.*

2) 3.5” Long-rod model (200mm ferrite length) outperforms 5”
Short-rod model (65 mm ferrite)

*A smaller diameter FSL model can
outperform a larger diameter FSL model if the smaller diameter model has a
ferrite sleeve length much greater than that of the larger diameter FSL.*

* *

3) 3.5” Long-rod model (200mm ferrite length) deadlocks in weak-signal
performance with 7” Bar model (100mm ferrite length)

*The ferrite sleeve length of an FSL
model is of equal importance to its
coil diameter in determining weak-signal performance. An FSL model with twice
the ferrite length of another model can match its weak-signal performance even
with a coil half the diameter of the other FSL model.*

* *

*Assuming that component parameters (Litz
wire type, coil orientation and ferrite permeability) are identical, any FSL
antenna’s weak-signal performance in relation to another FSL antenna can be
accurately determined by multiplying its coil diameter times its ferrite sleeve
length, and comparing this “Performance Score” with that of the other FSL
model.*

* *

4) 3.25” Bar model (100mm ferrite length) deadlocks in weak-signal performance
with 5” Short-rod model (65mm ferrite length)

*The “Performance Score” formula can
accurately predict the identical weak-signal performance of two FSL antennas,
even when they have different coil diameters and ferrite sleeve lengths.*

* *

The detailed MP3 recording links for each of these important
FSL test model match ups follow.

1) 3.25” Bar model (100mm ferrite length)
outperforms 3.25” Short-rod model (65mm ferrite)

**(3.25” Short-rod FSL reception for first 18 seconds,
followed by that of the 3.25” Bar FSL)**

540-Burien TIS (15 watts at
25 miles)

http://www.mediafire.com/?jqja5qneh3f9chl

790-KGMI (5 kW at 140 miles)

http://www.mediafire.com/?fbobcmhv4d27c47

1230-KWYZ (1 kW at 80 miles)

http://www.mediafire.com/?76nglcufzz1c7pd

1430-KBRC (5 kW at 110 miles)
http://www.mediafire.com/?76nglcufzz1c7pd,lzzz0bamjdt3jgg

2) 3.5” Long-rod model (200mm ferrite length)
outperforms 5” Short-rod model (65mm ferrite)

**(5” Short-rod FSL reception for the first 18 seconds,
followed by that of the 3.5” Long-rod FSL)**

540-Burien TIS (15 watts at
25 miles)

http://www.mediafire.com/?y6j2q0p4e76po6c

1230-KWYZ (1 kW at 80 miles)

http://www.mediafire.com/?i6575pbpkimg4vw

1410-CFUN (50 kW at 160
miles)

http://www.mediafire.com/?1v8qa3g3j0dcth3

1430-KBRC (5 kw at 110 miles)

http://www.mediafire.com/?zl3ba6dg6pd65ab

3) 3.5” Long-rod model (200mm ferrite length)
deadlocks in weak-signal performance with 7” Bar model (100mm ferrite)

**(7” Bar model FSL reception for the first 18 seconds,
followed by that of the 3.5” Long-rod FSL)**

** **

540-Burien TIS (15 watts at
25 miles)

http://www.mediafire.com/?bbuzgs8c819wsrj

750-KXL (50 kW at 150 miles)

http://www.mediafire.com/?87d0ccexi81yhaw

1040-CKST (10 kW at 160
miles)

http://www.mediafire.com/?i8d4jjd4q2vctml

1070-CFAX (10 kW at 90 miles)
http://www.mediafire.com/?ghh0omngnn5o9le

1410-CFUN (50 kW at 160
miles) http://www.mediafire.com/?lgfq7r9763g59h8

4) 3.25” Bar model (100mm ferrite length)
deadlocks in weak-signal performance with 5” Short-rod model (65mm ferrite)

**(5” Short-rod model FSL reception for the first 18
seconds, followed by that of the 3.25” Bar FSL)**

** **

540-Burien TIS (15 watts at
25 miles)

http://www.mediafire.com/?5n7vm2upt1zxfn9

750-KXL (50 kW at 150 miles)

http://www.mediafire.com/?ok82g3ey77bxncr

1070-CFAX (10 kW at 90 miles)

http://www.mediafire.com/?1cy43cahub24ylh

1410-CFUN (50 kW at 160
miles)

http://www.mediafire.com/?1v8qa3g3j0dcth3

** ****FSL Test Model Specifications**

** **

__Model__ __Ferrite Size__ __Ferrite
Wt.__ __Total Wt.__ __Construction Cost*__ __Performance Score**__

3” Short rod 65 x 8mm 1.5
lbs. 2 lbs. $45 195

3.25” Bar 100 x 20 x 3mm 1.0 lbs. 1.5 lbs. $40 325

3.5” Long rod 200 x
10mm 4.0 lbs. 4.5 lbs. $95 700

5” Short rod 65 x 8mm 2.5 lbs.
3.0 lbs. $65 325

5” Bar 100 x 20 x 3mm 2.0 lbs. 2.5 lbs. $50 500

5” Long rod 140 x 8mm 4.0 lbs.
5.0 lbs. $90 700

7” Bar 100 x 20 x 3mm 3.0 lbs. 4.0 lbs. $65 700

7” Long rod 140 x 8mm 6.0 lbs.
7.0 lbs. $130 980

* Total cost of components
(excluding shipping charges) as of March 2012

** Coil diameter (in inches) times ferrite sleeve length (in
millimeters)

* *