<<

A new idea for a very small directional 145MHz antenna. pa0nhc. 
Page opdate 2021
072
7  New version with modified matching circuit.
==>> Now FIRST PRESS F5 <<==
COPYRIGHT. The use, copy and modification of all info on this site is only permitted for non-commercial purposes,
and thereby explicitly mentioning my radio amateur call sign "PA0NHC" as the original writer / designer / photographer / publisher.

pa0nhc PCB service

This is an experimental  design for a very small and sturdy, directional 2m band foxhunt (ARDF) antenna.
It is  in the shape of a 162x195mm PCB card.

New version 20210726-4  .


Test values :
Coil L= 2 pieces 7.5mm long RG316 screening.
Soldered in parallel at PCB. LxWxT = 7.5x5x3 mm.
C3 = 3p3, C4 = (56p//15p). Total = 3.15 pF.
(Tuning rate = 35 MHz / pF).

 

PCB design .


VSWR measurement. B=0.86MHz.
The antenna must be tuned to a favorite
band portion (144.5 MHz or 145.5 MHz).


Serial RLC measurement.

 

This is a 145MHz delta loop with delta matching. I called it "Double Delta Loop". 

Good match is now achieved by introducing an inductance L, and balanced output tapping in the center of the loop.
The transformer and its tuning Cs are omitted. The construction is now simpler, and with a good CMC, its losses are now minimal.

After some experiments with a modified 20210701 PCB, good match was achieved. However, the -3dB bandwidth is narrower (0.86MHz), so the loaded Q of the loop is much higher (169). REM : optimal transformer matching results in lower loaded Q than other matching methods.

This low loss high-Q "pre selection" adds much to RF selectivity. Resulting in less chance of hinder due to : spurious (mirror) reception of airplane communication, pager stations, and receiver overload.

But, this high-Q antenna version does not cover the whole 2m band. It must be tuned to a favorite band portion (center at 144.5 MHz or 145.5 MHz).

As C3 and C4 are small in size, and have low operating voltage, max transmitting power is estimated to 1W.

A new version 20210726-4 PCB is ordered. Once it is tested, I will publish the results.


PCB version 20210701 with transformer matching.


Pa0nhc prototype. 
Shown nearly to actual size on a 60cm diagonal
HD screen.


 

 

Schematic. pa0nhc 20210724.

This very small size 165mm (6.6") wide and 195mm (7.6") long, sturdy and light weight (40g) ARDF foxhunt antenna is less suspicious, and easier to handle, especially in difficult terrains (brushes, mountains). A good antenna for recreational RDF hunts and starters too.

When held in parallel to the horizon, it shows a very nice directional pattern for horizontally polarized signals, with its maximum signal strength when the searched object is in the direction of the arrow. It has a single, deep minimum in the direction of the operator. 
Color is white to prevent loosing it out of site, or damage due to stepping on it. Other antenna colors (green) can be ordered.
It can also be used as a nearly invisible horizontally or vertically polarized directive antenna, for a low power foxhunt transmitter, or for a two way portable radio, horizontally or vertically mounted beside a (fiber) mast. 

To minimize the weight of the antenna to about 40 gram, all PCB material was removed just outside the copper tracks, and small components are installed.
All corners were rounded to prevent problems with catching twigs and leaves.

A rectangle shaped extension, enables fixing the antenna to a boom, mast or directly to a receiver.
The antenna itself must stay free from other objects, to prevent de-tuning.

            Design.
It is an at 145MHz resonating delta loop with nearly a 1/4 wavelength circumference. The loop conductor width is 10mm. The element conducting surface equals therefore to that of a 6.4mm thick copper wire. It is effectively protected from humidity and oxidation by the PCB solder mask. Three SMD resistors prevent static charge building up and high noise levels. The loop, the transformer TR1, and the transmission line are fully balanced in design and construction. The output impedance is 50 Ohms with a reasonable low VSWR. Thanks to the very effective common mode choke "CMC" at the bottom end of the balanced transmission line, a coax can be used to connect the antenna to the receiver without disturbing the antenna pattern. The horizontal directive radiation pattern will not be influenced by (the always present) common mode currents on the screening of the coax feeder. 

The antenna will thus show a symmetrical, wide angled signal maximum, and (important !) only ONE sharp and deep minimum, at exactly the back of the antenna, which is needed for EXACT and fast determination of the direction to the transmitter-fox.

            Expected practical results :

of a horizontal 1/4 wave circumference transmitting DELTA loop shows, that the maximal signal strength made by such a loop is in the direction of the lowZ feed point. An up to -20dB signal minimum is in the direction of the tuning capacitor.

When the delta loop is compared with the circle shaped loop, the minimum in signal strength from a delta loop is deeper, than that of a circular shaped loop. This is an important property for a direction finding antenna.
After first testing a square loop I therefore ended designing a delta loop.

By turning such a loop from horizontal to vertical position, it will show 2 very sharp minima the horizontal plane. Just like a horizontal straight dipole. When used, both patterns allow very fast and accurate bearings of horizontally polarized signals.
Advise : Practice to know your antenna !

          Calculations with software from internet for this VHF loop failed :
"KI6GD loop calculator" will not accept higher resonance frequencies than 30 MHz.
"Pacific66 LoopCalc" gave very strange, surely unusable results.
"MManagal" directive pattern calculation attempts failed also, i could not insert a Ctune for resonance.
Efficiency : My impression is that the efficiency could be better than -1 dB.
That could be a few dB less than a good yagi antenna.

 is therefore recommended. 

 

CMC : 15t 2x0.2mm twisted 1t/cm. Wound @ FT50-61 over 270degr. 3 layers top coat applied.
REM : Due to painting, all properties of the coil changed, and were only stable until completely dry.
The same, but not painted CMC, will show different properties !
Paint first, let dry, then measure.


Common mode isolation, measured with
NanoVNA : 36dB, Z=3k14 @ 145 MHz.


Insertion loss : only 0.5dB @ 145 MHz.


Minimum VSWR should be near 145MHz.

          COILS.


The TR1 secondary connections.

The 0.7mm wire has passed the
core hole 3 times. The turns must be
divided over the whole core circumference,
and stay close to the core surface.
Use a tapered ball point pen to press
the windings inside the hole close to the
T50-17 core surface.

 

 


The TR1 primary connections.
See schematic.
At the secondary coil center turn, remove some insulation at both bottom corners. Then pre-tin them.
Solder these tinned coil points directly onto the PCB primary contacts.

REM : All capacitors and resistors should be SMD 1206 types.

The used ferrite cores are 12.5mm in diameter from Amidon, Fairrite or Micro Metals. Details.

Tuning the loop for the best VSWR without series tuning the transformer, will result in a 
frequency difference between loop resonance and lowest VSWR frequency
.
The loop antenna efficiency will be (much) lower than expected.

          Tr1 :
1. First wind the 3 turns secondary winding with 0.7 mm thick lacquered copper wire on a T50-17 iron powder core.
    The wire should be wound very close to the core, and the 3 turns MUST be divided as good as possible over the hole core circumference. See picture.
2. The single primary turn connections will be created, by removing some isolation lacquer at the PCB sides of the secondary single center turn, and tinning them.
    See photo and schematic.
3. The tinned coil contact points will be soldered directly onto the PCB, without connecting wires.

          CMC (update 20210724)
1. Fold 60cm 0.2mm lacquered copper wire to a 30cm long parallel running wire pair.
2. Firmly twist this pair 30 times. Kinks are not allowed. 
3. Then wind the twisted pair 15 times through the hole of a FT50-61 core.
    Be sure there is no free space between the turns an the core body.
    After painting, no water may come between wire and core.
   Turns may NOT overlap.
   Turns must be evenly divided over 270 degr. of the core circumference. 
4. Paint all turns and core at least 3 times with "Top Coat" (transparent nail lacquered). 
    Let dry it a bit in-between. 
    Spaces between wire and core should be filled with lacquer. Once finished, let dry at least 12 Hrs.
5. Check common mode and line impedances (see graphs). 
    The lowest VSWR value should occur around 145 MHz. See graphs.
6. Solder the CMC in place, and glue it to the PCB with Top Coat.


Measuring IL and SWR of the 15t
CMC with a NanoVNA.

Disconnect both ground wires to
measure Common Mode suppression.

          C1 and C2 :
    They tune the leakage inductance of TR1 to series resonance at 145 MHz. The leakage series inductance then vanishes, and best match and lowest losses can be achieved. 
1. Solder for C1 and C2 both SMD capacitors of 82 pF. 
2. Protect them with 2 layers transparent nail lacquer.

If, due to spreading in CMC core material or winding technique, the VSWR on 145MHz shows to be higher than 1.6, you could try to change the value of C1 and C2 both equally. 
1. First solder 22pF on top (in parallel) of them.
2. If the VSWR becomes worse, change C1/2 both for an equal smaller value and fine tune with small value parallel Cs.
3. After every change, adapt the value of C4 for resonance at 145 +/- 0.5 MHz and measure VSWR.

          Ctune :
In practice, it will be two SMD capacitors (they are connected in series by a track on the antenna PCB).
In my prototype I used an air trimmer of max. 7 pF. Its capacitance was about 4.1 pF when tuned to 145MHz, 
1. Solder for C3=4p7 and for C4=33p.
2. Check with a VNA that the loop resonance (minimal VSWR) is at 145 +/- 0.5 MHz. 
3. If needed, change the value of C4 a little to 27pF or 39 pF.
4. For fine tuning, you can solder a small value SMD capacitor on top of (that is in parallel to) C4, and check resonance again.
5. Once finished and tested, protect C3 and C4 with 2 layers transparent nail lacquer.
6. Note the VSWR at resonance. See "C1 and C2". 

REM: a 5pF trimmer capacitor can be used for C3/4 while testing and adjusting C1/2. As it will not be protected against moist, it causes failure while hunting in rainy circumstances.

          Design details.
When developing this small 2m loop, i used my experiences with designing and optimizing my 40-80m TX loop. and the research by G0CWT.

In the center of the top of the loop is the RF current maximum, which is also the low impedance feed point. It is connected to a   1turn : 3turns   iron powder matching transformer. To avoid coupling with the loop, a balanced 50 ohm PCB strip line is used as transmission line. It runs from the top to the bottom end of the antenna part. The transformer at the op of the loop matches the balanced 5.5 Ohms loop impedance, to the balanced 50 Ohms impedance of the transmission line. By series tuning the primary winding of this transformer to 145MHz, a low VSWR value and optimal low losses are achieved.

Without a CMC (balun) in the transmission line, distortion of the radiation pattern of the antenna (asymmetry or skewing) will be caused by the combination of signals received by the loop, and common mode signals picked up by the outside of the coax screening. With a good CMC, these signals cannot combine. At the bottom of the loop a good "balun" is therefore inserted into the transmission line.
REM : To avoid magnetic RF coupling between the connected coax and loop, the coax should run at an angle of 90 degrees away from the loop.

At the bottom end of the balanced PCB strip line, a 50 Ohm Common Mode Choke ("balun") is therefore inserted. It also functions to convert the balanced "two wire" 50 Ohms line, to an unbalanced  50 Ohms coax line. 

            Coil materials, tuning and matching.
I did not have optimal transformer core material, so I first tried FT50-61 for TR1. It was wound and experimentally tuned for best VSWR @ 145 MHz. S21 measurements on this 3:1 turns transformer, loaded with 50 Ohm (RX) and 5.5 Ohm (loop) by my NanoVNA, suggested that this material is unusable for magnetically coupled 145MHz transformers, as core losses are to high. FairRite specs for #61 material also sate clearly : The losses in #61 material rise sharply above 20MHz. 
#61 material is however excellent for low loss use below 20MHz.

A specially suited material for up to 180 MHz should be ferrite #68. As FiarRite produces # 68 VHF cores only to customer specifications, these are not available.

The best and only remaining core material for a 145 MHz transformers is therefore iron powder type 17. 

Tests with a T50-17 core succeeded. Tight coupling between pri. and sec. windings, and a low number of turns showed to be important. 
A transformer with separated pri. and sec. windings did not succeed. The coupling between pri. and sec. was insufficient. I ended therefore with a core wound as "auto transformer". 
In an "Auto transformer" is the secondary winding a part of the primary winding. On other words : the secondary is made by tapping the primary.
The balanced connections for the single primary turn must here be made around the single center turn of the balanced secondary (see schematic and photos).
Both contacts were created, by removing a small part of the lacquered insulation on the PCB side of the center winding.
These two primary contacts were then directly soldered onto the PCB. Do NOT use wire connections. 

As the wire inductances of capacitors C1 and C2 add to the harmful transformer leakage inductance, (wireless) SMD capacitors are used for C1 and C2. These capacitors series tune the transformer leakage inductance to about 145 MHz. After a lot of measurements, the optimal values were found, and VSWR was about 1: 1.5 at 145 MHz. From the measured bandwidth, a loaded loop-Q can be calculated as about 72. This seems to be OK, as my 1/4 wave circumference 40m loop has a Q value in that region to. 
REM : The very high Qs and small bandwidths reported by loop calc software are valid for very small loops, not for 1/4 wave circumference loops.

REM : When the loop is experimentally tuned to 165MHz, a lower VSWR is noted. The loop feed point impedance is then probably more near the 5.5 Ohm observed by G0CWT and me. Resulting in better match. The cause is, that this loop its circumference is deliberately chosen a bit to short for 145MHz, in order to get a bit handier value for tuning capacitor C3+C4. Its feed point impedance at 145 MHz differs a bit from 5.5 Ohms.

          Weather proofing.
Paint all components and connections with at least two coats of transparent nail lacquer ("Top Coat"). Use it also to glue the CMC to the PCB.
Be sure no water can intrude the connected coax nor the CMC windings. Water intrusion can cause a massive change in transmission line impedance and large losses.

          Measuring tips :
Mass contact resistances in BNC to SMA adapters can cause unreliable measuring results, especially when measuring VSWR, IL and resonance.

Use good quality, machined connectors of well known brands. Do not use cheap Chinese nickel plated die castled junk, which miss the slits in de mass contact bus, and will create diode contacts.

Apply to all connectors (do not forget USB) a little oxide and sulfide removing, contact improving fluid, like CRC 2.26.
After that, turn BNC bayonet rim round, and move BNC plugs in-out a few times, to loosen/remove oxides etc. before staring measurements.