LZ1AQ / pa0nhc wide band loop
Listen to this antenna at the websdr of PA0EBC : http://mwiek.nl:8073/
download, print and save these 6 downloadable PDF files, and keep
them as documentation :
To be able to read them, you should have installed "Acrobat reader ".
Update : 20190828
Update : 201900921
|Overvoltage protection detail||Amplifier housing||READ
Frequency characteristic of the whole system.
Measured with NanoVNA.
Loop replaced by a 1.1 uH choke.
C5/10 = 150 pF.
NanoVNA port "ch0" connected with 2x 470 Ohms in balance to "Loop".
NanoVNA port "ch1" connected to splitter RFout.
Blue = Fchar.
In practice is 16 kHz still well readable.
Reception using Airspy HF+ and SDRconsole in a VERY noisy city location.
30dB "Visual Gain", resulting in signal indications equivalent to a full sized dipole antenna in equal field strength.
This antenna acts almost as a pure magnetic transducer. The input impedance of the amplifier is so low that any currents induced by electric field become very small compared to the currents induced by magnetic filed. This antenna does not need shield or any type of grounding. For vertically polarized low elevation angle signals the antenna has very sharp null. The directivity for the sky wave signals is not determined since their polarization is stochastic. The influence of nearby non-resonant conductive object is negligible. The differential circuit also reduces the influence of common mode currents. It works from height almost zero above the ground (there is almost no change in signal levels when the loop side is placed several centimeters above the ground in field environment). The wideband properties are excellent - from LW to upper HF even 50 MHz band can be included. The dynamic range obtained from on the air tests on the bands is good and no apparent non-linear distortions are found. The circuit is very simple, stable and cheap and there is nothing critical for adjustment. The antenna can be mounted outdoor and connected with FTP cable to RX and PS parts. The FTP cable is widely available and the associated connectors are very reliable and cheap. This is my favorite antenna for my city office where nothing else can survive the EMC pollution. The only drawback of this active antenna is its relatively higher noise floor specially for frequencies above 10 MHz which is several dB above the atmospheric noise levels for quiet rural locations at some frequencies and times of the day (for single loop 1m diam.) . The antenna noise floor is acceptable and suitable for all locations where the man made noise is moderate and above. The noise floor limit of these types of WSM loops is essential - see the Appendix section for more details. The noise floor can be reduced by using “fat” , parallel or parallel crossed loops especially for places where the electromagnetic noise is very low.
General properties :
- Pure H-field (magnetic) antenna.
- Antenna and transmission line are fully balanced.
- Power supply and signal transport by means of one standard "straight 1:1" UTP network cable.
REM: STP/FTP is NOT recommended.
- Insensitive to E-fields.
- No loop screening needed
=> Simple construction.
=> Simple installation.
- Loop and splitter are not grounded
- Only the connected receiver should be connected to a safety ground.
- Very effective internal common mode chokes and balanced transformers.
=> In practice NO transmission line de-noising measures are needed.
- Useable at very low height above ground with little signal loss.
system should withstand :
1. Induction from lighting at 100m distance,
2. Up to 1,5 kW radiated power at 10m distance.
3. Static charges and common mode voltages.
4. Supply over-voltage and wrong polarization.
External noise free power supply : 12.5 Vdc to
17Vdc / 135 mA.
The LZ1AQ / pa0nhc amplifier unit.
- Fully balanced.
- Two stages for ample amplification (48 dB) and large bandwidth.
- This amplifier is stable, and can cope with very strong signals.
- No signs of oscillations were noted.
- No IMD occurred even when receiving VERY strong signals on
LW, MW and SW at the same time.
- Input impedance abt. 3 Ohms balanced.
- The loop runs therefore in "short circuit mode".
- Output impedance 100 Ohm balanced to match a UTP twisted wire pair.
- Amplifier bandwidth (measured with nanoVNA) :
50 kHz -20 dB. REM : reception of 22.5 kHz still very good.
500 kHz to 23 MHz +0/-3 dB => WIDEband.
30 MHz -5 dB. Amplifier noise floor is noticeable.
70 MHz -45 dB. Good suppression of FM local broadcast.
- Loop bandwidth (circumference 2 m) 15 MHz. Above 15 MHz the loop sensitivity changes (more sensitive ?).
- The splitter RF output impedance is 50 Ohms asymmetrical.
prevention against strong electromagnetic fields.
Fast switching diodes D1-8 limit to high RF signal voltages on both amplifier inputs.
To high common mode and static voltages are limited to +-15V by R25, D9, D10, Z2 and Z3, and are blooded to receiver ground via one free UTP wire.
I deliberately chose for easy installable wired components and wide PCB component holes as :
- IMD free film and NP0 capacitors are mostly available as wired.
They are needed for low IMD, and temperature stability (outside temperatures can vary between -20C and +50C).
- Common mode chokes wound on optimal #31 ferrite cores for better cable noise rejection above 1 MHz.
- Added Tr2,, a 100 ohms alance to 50 Ohms unbalance transmission line transformer with separated windings.
- Transformers Tr1 and TR2 are both wound on #77 ferrite cores for wide bandwidth down to 25 kHz.
- Fine tuned the values of C5 and C10 for optimal bandwidth, together with good FM broadcast suppression.
- Designed two double sided PCB's with :
RJ45 busses and
a large number if ground plane via's.
- Made a component ordering list with ordering details.
Output, bandwidth and noise floor :
Antenna output of a dipole antenna :
In a field strength : 0,02 V/m, a correctly loaded 40m long 3.65 MHz dipole at 20m height, will give 0dBm output over a 50 Ohms load.
The output of this loop antenna calculated with LTspice :
A 60cm diameter loop antenna output will then be 0,014 mA RF current.
The loop amplifier will amplify thus to -43 dBv or -30 dBm (S9+43 dB) over a 50 Ohms load.
The output difference between this loop and an 40m long dipole is therefore -30 dB.
To get practical. signal strength
indications, i suggest to set your software's gain to + 30 dB.
This will compensate for the differences between dipole en small loop outputs, and signal strength indiactions will be more comparable with these of other stations using a wire antenna.
Measured with the loop replaced by a 2.2 uH choke, B=10 kHz, 50 Ohms output, and +30 dB software gain.
On 28 MHz the systems noise floor is at least 4 dB lower than my locations man mad noise.
MHz : 1.8 3.65 7.1 14.175
-dBm : 112 115 115 117
Loop construction :
I used easy available 2m long 15x2mm aluminum strip for the construction of the loop. This loop is connected to a special, by LZ1AQ designed two stage loop amplifier. It is fully balanced, and has a very low input impedance of abt. 2x2 Ohm. The loop therefore behaves like a current source with the
loop self inductance of 1.6 uH in series with it. If such a lowZ loaded loop is located in a magnetic H-field of constant strength, it generates a constant RF current for all wavelengths larger than 10 loop circumferences.
This loop is in its near field
(radius = lambda / 6) only sensitive to
E-field noises sourced from within its near-field (radius Lambda / 6), will be capacitive coupled. They are by nature of high impedance. And will cause weak noise currents into the loop. These weak currents will cause across both lowZ amplifier inputs very weak voltages (U=IxR). Due to the small loop dimensions, these noise voltages will have at both loop connections (nearly) equal amplitude and phase. As the amplifier has a balanced input and output, these noises will further be attenuated, resulting in a nearly pure H-field antenna.
The influence of the loop self
inductance to the amplifier noise floor.
Up to 14 MHz, this 60cm loop self noise is weaker than the atmospheric noise.
In city environments, having a higher noise level than in the country site, the antenna self noise is not important.
According to LZ1AQ, a loop
diameter of abt. 1m is optimal.
- Low noise input transistors will NOT lower the noise floor.
- A better conducting loop (copper and/or silver plated) will NOT lower the noise floor, as it does not lower the loop inductance.
- A larger loop circumference makes the S/N WORSE, as it results in a higher loop inductance.
- A loop with more than one winding connected in series will WORSEN THE S/N, as it has a far higher self inductance.
- A thicker loop (f.i..36mm) WILL IMPROVE the S/N, as a thicker loop has a lower loop inductance.
- A loop with more than one winding connected in parallel WILL IMPROVE the S/N, as it has a lower loop self inductance.
Transmission line and splitter.
Common knowledge : a balanced line is the most insentive to noise fields. An UTP network cable contains four balanced lines. In this design, a (weather resist) "straight / standard" UTP network cable is used. STP nor FTP are recommended.
As the used wire pairs are twisted, they are balanced, and insensitive to surrounding noise fields. A vast plus over coax use. It makes the system "plug and play", easily deployable for temporary use, and the transmission line very insensitive to noise fields. One wire pair is used for signal transport. Another wire pair is used for amplifier power. Both remaining wire pairs are only connected to the ground surface of the splitter. One wire is used for bleeding static voltages connected to the amplifier "ground" through a 100 k series resistor.
Common mode noises at the UTP network cable in the amplifier blocked by the secondary of Tr1, and Common Mode Choke L7. In the splitter Common Mode Currents they are blocked by CMC's L10, L12.
The specially by me developed transmission line output transformer Tr2 ensures very good separation between the RF transmission line and the 50 Ohms output. The RF transmission line remains effectively floating and balanced, as both transmission line transformers Tr1 and Tr2 have no galvanic connection between primary and secondary windings. Ensuring very good wideband decoupling of common mode signals on the UTP cable.
Two extra #43 + #31 CMCs onto the UTP cable showed to make no difference on noise levels in my very noisy antenna location.
Power to the antenna switched off, resulted in signal or noise levels to drop by 53 dB (MW) to 32 dB (SW).
Z1 and F1 in the splitter prevent damage from to high or wrong
Fuse : 5 x 20 mm, 200mA T (slow).
The needed external power supply.
- MUST be a LINEAR LOW NOISE type, using a 50/60Hz mains
- Must have a fully closed METAL cabinet.
- The 12Vdc output must be floating (insulated from the outside of the cabinet).
- The mains safety ground MUST be connected to the OUTSIDE of the cabinet (NOT to the inside of the power supply).
See mod. below.
- Check if the rectifier diodes in it are each paralleled by a 0.1 uF capacitor.
If not, install these capacitors.
- This power supply may ONLY be connected to the splitter, NOT to other equipment.
TIP : On the Internet some small cheap
noise HiFi audio power supply" are available.
For instance : "Breeze Audio", 12V / 15 W : only $56 at Aliexpress.
Their output voltage is internally adjustable between 11V and 14 V.
Minimum needed supply voltage.
The LM2940SX-10 low drop stabilizer on the amplifier PCB needs at least 11 Vdc to work well.
With the supply current of 135 mA, every 10m UTP cable causes abt. 0.5V voltage supply loss.
The minimum to the splitter connected external supply voltage therefore depends on the length of the UTP line.
With the power supply adjusted for 12.5 Vdc, the max. length of the network cable is 30 m.
Even with a short network cable connected, it is safe to supply up to 16 Vdc.
Grounding, connecting and de-noising.
The antenna, nor the splitter nor the power supply 12Vdc output may be grounded.
A connected receiver, or a connected PC MUST be grounded.
If you experience a high noise level, this noise
can be injected into the system in various ways :
a. Through the mains power supply grounding wire.
you use a power cord with on the power supply end a three-pole equipment
You MUST correct this FIRST.
Fist open the power supply
cabinet, and check if the mains power bus ground pin is connected to the
PCB or the inside of the baninet.
If it is connected, modify the power cord as follows :
the power cord from the mains supply.
1. At abt. 10 cm away from the power supply plug, cut the power cord.
2. The ground wire end which is still connected to the mains supply end of the plug, has to be lengthened by a piece of (yellow/green) wire.
3. This piece of ground wire must be connected to the OUTSIDE of the power supply cabinet.
4. Reconnect and insulate both live and neutral wire end.
Use shrink hose for insulation.
From now on noises on the mains wire will stay on the outside of a fully closed metal cabinet (cage of Faraday). And cannot enter the inside of the power supply, nor the 12Vdc output
5. The 12Vdc output bus MUST be insulated from the cabinet (be a plastic type connection).
Through remaining cabling.
The balanced lines inside the UTP cable in this design have proven to be VERY insensitive to noise fields.
With correct wound CMC's and transformers, it is unlikely that noises induced into these balanced lines will be of influence to the system performance.
In practice cable noises are much weaker than noises received by the loop itself.
you could try the following :y the following :
1. Change the routing of the network cable.
Installing a common mode choke :
2. If apply able, on the USB cable, as close as possible near the SDR unit.
3. On the 12Vdc power cable, near the splitter.
4. On the 240V~ power cord, near the power supply.
5. On the 50 Ohm coax, close to the (SDR) receiver antenna input.
Reception through the loop antenna itself, try a combination of the following :
- Rotate the loop.
- Change the location of the loop.
- Change the loop height above noisy surroundings.
For vertically polarized low incident DX or near field signals,
try to minimize the most annoying noise by turning the loop antenna.
For NVIS signals (1.8 MHz - 7.2 MHz from 30 km to 300 km) a vertical loop shows no directivity. Signals coming from above (NVIS), or from below the antenna (mains and network cabling), cannot be attenuated by turning the loop.
to install CommonModeChokes :
Wind as much turns as possible THIN (1/10 ") cable through the hole of one FairRite mix #31 core.
Use a - Ring core 2631801202
- "Snap-It" core 0431173551
IMPORTANT for good performance :
The following is also valid for winding the transformers and the common mode chokes on both PCBs.
- Always wind in ONE direction around the core circumference. Never wind back.
- Every turn MUST lay beside another. Never crossing another.
- When the core circumference is fully filled, it is allowed to wind further in the same direction
(see the coil on right).
One incorrect wound turn will cancel one other turn, resulting in TWO inactive turns.
Using a split core :
When both halves of a split ferrite core do not close in perfect mechanical contact with each other, the CMC does not work at all,
- To be sure both core halves close perfectly, the cable turns
should fit loosely into the hole of the split-core. Check it.
- Apply extra pressure to the core halves by installing a black ty-wrap around the core housing. Pull it tight.