# A Beginner's Guide to Antennas for Blitzortung

**Draft v0.3 — for community review**  
**Built and documented by Alex Frankes, station 3304, Groningen, the Netherlands**

## The antenna routes covered in this guide

The guide begins with the three easiest routes and then describes the other antenna constructions found in the official project documentation.

| Design | Active conductor | Geometry | Why choose it? |
|---|---|---|---|
| **A. Station 3304 crossed-diamond antenna** | Two separated `0.75 mm²` copper speaker-wire conductors | Two three-turn diamonds at 90° on one three-axis wooden frame | Cheapest and simplest; built, tested and working |
| **B. Documented coax-loop antenna** | Coax inner conductor, with the shield used only as electrostatic shielding | Two three-turn circles of 1.00 m diameter at 90° | Shielded alternative described in the project documentation |
| **C. Ready-made FRA-200/12 ferrite antennas** | Factory-wound ferrite rods made for Blitzortung | Two compact ferrite antennas mounted at 90° | Easiest route when you prefer not to build an antenna |
| **D. Classic direct wire loops** | Insulated copper wire of at least `0.75 mm²` | Two large circular or square loops at 90° | Conventional construction with an easy-to-understand wire path |
| **E. Compact multi-turn loops** | Approximately 25 m of `1 mm²` insulated copper wire per axis | Two 38 cm circles with 20 turns each | Smaller frame, but much more winding |
| **F. Home-wound ferrite rods** | Enamelled copper wire wound on ferrite rods | Two horizontal rods at 90° | Compact experiment when a proven coil specification is available |
| **G. Möbius coax loops** | Cross-connected coax inner conductor and shield | Two advanced coax loops at 90° | More active turns from fewer physical rounds |
| **H. Transformer-coupled tube loops** | Copper or brake tube plus a toroidal current transformer | Two compact rigid loops at 90° | Robust advanced construction, potentially suitable outdoors |
| **I. E-field probe** | One short metal probe | One omnidirectional electric-field input | Very simple sensor, but more sensitive to local electrical noise |

These are the three easiest beginner routes. Later sections also cover classic square or circular wire loops, smaller loops with more turns, home-wound ferrite rods, Möbius loops, transformer-coupled copper-tube loops and an E-field probe.

The many valid alternatives are exactly why this guide does not claim to present a “perfect antenna”. Some later designs can be reproduced from fixed dimensions; others depend on a transformer core or ferrite winding specification and therefore cannot honestly be reduced to one universal shopping list.

For H-field reception, choose one construction type and build or buy two matching antenna axes. Mount those axes at 90 degrees. Do not connect several unrelated H-field designs to the same H1 or H2 input.

### Read this before starting

- Designs A, B and C are the recommended beginner choices. Designs D onward are alternatives for readers who deliberately want a different construction.
- Design A is the easiest do-it-yourself build and is the one actually tested at station 3304. Design C is the easiest overall route because the antennas arrive ready-made.
- The fixed measurements for designs A, B, D and E already include spare cable. **You do not need to calculate those lengths yourself.**
- `H1` and `H2` are the two receiving channels on PCB 23.1. Each channel gets one complete H-field antenna axis.
- A *turn* means one complete trip around the frame. Three turns means going around the same path three times before stopping.
- Keep the copper conductors of the two loops electrically separate.
- Follow the jumper instruction belonging to the chosen design and current preamplifier revision. Do not assume that one jumper setting applies to every loop, ferrite rod or transformer arrangement.

### If building an antenna still feels too difficult

There is absolutely no requirement to build an antenna yourself. Purpose-made [FRA-200/12 ferrite-rod antennas for Blitzortung](https://www.ferrite-rod-antennas.de/page-4) can be ordered ready-made. They are compact, neatly wound and described by the supplier as the standard ferrite antenna for the Blitzortung system.

Order **two antennas** and state that they are for one Blitzortung station. The supplier will try to select a pair with closely matched performance. The current listed price is **€35 per antenna**, so the pair costs **€70 plus shipping**.

The ferrite antennas and the System Blue station come from **different suppliers** and therefore require two separate orders. Place both orders at approximately the same time. That gives them a reasonable chance of arriving during the same period, although delivery on the same day cannot be guaranteed.

This is the easiest route: order the System Blue station, order the two ferrite antennas immediately afterwards, and wait for both parcels before completing the installation.

## Design A: the station 3304 speaker-wire antenna

### The result first

This is a simple dual H-field antenna made from three wooden battens and ordinary copper speaker wire. It uses no coaxial cable, ferrite rods, transformers, tuning capacitors or shield gaps.

![Alex Frankes' completed crossed-loop antenna](antenne.png)

The finished antenna contains two electrically separate, three-turn wire loops mounted at 90 degrees to each other. The actual station-3304 antenna is installed **indoors on a dry attic**, so it does not require outdoor weatherproofing. Both channels work and initially showed almost identical noise floors of approximately **13–14 mVpp** on automatic controller settings.

It may not be the most elegant antenna ever built, but it is cheap, simple and it works.

### The short build description

For other newcomers, this is exactly what was done:

1. Buy **11 m of 2 × 0.75 mm² copper speaker wire**.
2. Separate it into two individual 11 m wires.
3. Make a frame from **three 80 cm wooden battens**, crossing at the centre at 90 degrees—like a Czech hedgehog or anti-tank cross.
4. Wind each wire three times around the frame, creating two loops at 90 degrees to each other.
5. Connect one loop to `H1+` and `H1−`, and the other to `H2+` and `H2−` on preamplifier PCB 23.1.
6. Leave both solder jumpers open.
7. Connect the preamplifier to the System Blue Basic/Mini controller and begin with the normal automatic settings.

That is the complete electrical recipe. The remainder of this guide explains why it works, how to reproduce it reliably and how to diagnose problems.

![Crossed-loop construction and wiring principle](blitzortung-speaker-wire-crossed-loop-wiring.svg)

> The drawing shows the electrical and mechanical principle. Always follow the labels printed on your own preamplifier revision.

### What this antenna actually is

The wooden frame has three axes:

- one vertical batten;
- one horizontal batten forming the first loop plane;
- a second horizontal batten at 90 degrees, forming the second loop plane.

Each loop follows four endpoints of the frame and therefore forms a diamond. The two diamonds share the same physical top and bottom support points, but their copper wires remain electrically separate.

This arrangement gives two perpendicular H-field channels in a single compact structure. A signal weak in the null of one loop can still be received by the other.

The exact compass orientation is not critical. The important relationship is the approximately 90-degree angle between the two loop planes.

### Why ordinary speaker wire works

A lightning discharge produces a very short burst containing many frequencies at once. The antenna therefore does not need to be adjusted to one exact frequency. The wire loops receive the magnetic part of that burst; this is what “H-field” means here.

The official Blitzortung magnetic-antenna documentation states that loop shape is not critical. Total enclosed area across all turns determines the signal strength. A circle encloses the most area for a given conductor length, but a square or diamond can be easier to construct and still be effective.

The `0.75 mm²` conductors used here also meet the documentation's suggested minimum wire cross-section for a direct wire loop.

There is no coaxial shield in this design. Consequently:

- there is no shield to interrupt;
- no midpoint shield gap is needed;
- there is no shield connection to a third terminal;
- each loop simply has two conductor ends.

### Why 11 metres of speaker cable is enough

You do not need this section to build the antenna; the shopping measurement is simply **11 m of twin speaker cable**.

After pulling the two insulated halves apart, you have **two separate wires, each 11 m long**. One wire is used for H1 and the other for H2.

On the 80 cm frame:

- one complete diamond-shaped round uses approximately **2.26 m** of wire;
- three rounds use approximately **6.8 m**;
- an 11 m wire therefore leaves more than **4 m spare** for the two connection ends, routing and mistakes.

The diamond formed by the 80 cm battens encloses approximately **0.32 m²**. That figure only describes the size of the loop. It is not a setting, a target or something that must be measured during construction.

Small differences in batten length or winding position will not make the antenna suddenly stop working.

## Materials for design A

### Required for the antenna

- 11 m of `2 × 0.75 mm²` copper speaker wire;
- three wooden battens, each approximately 80 cm long;
- screws, a small centre bracket or another rigid non-conductive/mechanically quiet centre joint;
- cable ties, tape or small plastic fasteners to hold the turns in position;
- strain relief near the preamplifier;
- PCB 23.1 H-field preamplifier and its normal controller connection cable.

### Useful tools

- screwdriver and drill;
- wire cutter and stripper;
- multimeter with continuity mode;
- ruler or tape measure;
- soldering iron only if the chosen connection method requires it.

Avoid a metal outer frame or closed metal ring around either loop. Small screws or the centre bracket are mechanically convenient, but unnecessary conductive structures following the loop perimeter should be avoided.

## Expected cost

Prices vary by country and supplier. These figures are realistic European estimates with a reference date of **27 August 2026**.

### The antenna itself

| Item | Expected cost |
|---|---:|
| 11 m of 2 × 0.75 mm² copper speaker wire | approximately **€10–€20** |
| Three 80 cm battens, or one longer batten cut into three | approximately **€4–€10** |
| Centre fixing, screws, cable ties and strain relief | approximately **€3–€10** |

**Expected antenna cost: approximately €17–€40.** It can be cheaper when wood and fixings are already available.

### Complete receiving station from scratch

| Item | Expected cost | Notes |
|---|---:|---|
| System Blue Basic PCB 22.2b package | **€318** | Current official package includes PCB 22.2b, PCB 23.1 preamplifier, piggyback PCB, aluminium housing, panels and controls; all parts are currently supplied soldered |
| Tracked DHL shipping from Germany | approximately **€15–€25 within nearby EU countries** | Exact amount is confirmed before payment |
| External active GNSS antenna, SMA, suitable for 3.2 V bias | approximately **€10–€25** | Required separately; the GNSS receiver module itself is already fitted to PCB 22.2b |
| Reliable 5 V USB-C power supply | approximately **€10–€20** | Avoid an electrically noisy no-name supply |
| Ethernet and connection cabling | approximately **€10–€30** | Depends on required lengths and existing materials |
| Crossed-loop antenna described here | approximately **€17–€40** | Speaker wire, battens and fixings |

**Realistic complete total: approximately €380–€460.** A safer round budget is **about €400–€500**, allowing for separate shipping, better cabling and unexpected small materials.

For design B, budget approximately **€30–€75** for 21 m of coax, two one-metre loop frames, fixings, insulation and strain relief. The resulting complete-station budget remains approximately **€400–€500**.

There is no Blitzortung subscription or membership fee. Continuous electricity use and a small amount of network traffic are the normal recurring costs.

As the ready-made alternative, order two [FRA-200/12 ferrite antennas made for Blitzortung](https://www.ferrite-rod-antennas.de/page-4) together with your station order. They currently cost €35 each, or €70 plus shipping. This is a separate order and parcel from a different supplier; ordering both at the same time merely increases the chance that they arrive during roughly the same period.

### Rough material budgets for the additional designs

These wider estimates reflect the large price differences between reclaimed material, local hardware shops and specialist electronic suppliers. They exclude the detector and PCB 23.1 already listed above.

| Design | Rough material budget | Main cost driver |
|---|---:|---|
| D. Classic direct wire loops | **€25–€80** | Copper wire and two large frames |
| E. Compact 20-turn loops | **€60–€170** | Approximately 50 m of `1 mm²` copper wire |
| F. Home-wound ferrite rods | **€20–€80** | Two identical rods and a proven coil specification |
| G. Möbius coax loops | **€30–€90** | Coax, two frames and reliable insulated joints |
| H. Transformer-coupled tube loops | **€70–€200** | Copper tube, suitable toroids, winding wire and good coax |
| I. E-field probe | **€5–€40** | Probe enclosure, F connector and coax; only when a compatible input already exists |

Do not choose design H merely because a cheap toroid is available. The core material must suit the transformer design; a random ring core with the wrong material is not a bargain.

## Building the frame

### 1. Mark the centres

Mark the centre of all three 80 cm battens. Accuracy helps the structure remain balanced, but millimetre precision is unnecessary.

### 2. Form the three-axis cross

Join the battens at their centre points so that each axis is approximately perpendicular to the other two. Viewed in three dimensions, the result resembles a Czech hedgehog or anti-tank obstacle.

The six endpoints define:

- top and bottom;
- left and right for loop 1;
- front and back for loop 2.

Make the joint rigid enough that the loop planes cannot slowly collapse or change angle.

### 3. Check the loop paths

Before adding wire, visualize the two diamonds:

- loop 1: top → left → bottom → right → top;
- loop 2: top → front → bottom → back → top.

The planes should cross at approximately 90 degrees.

## Preparing and winding the wire

### 1. Separate the twin speaker cable

Carefully pull the two insulated conductors apart over their full 11 m length. Do not damage the insulation while separating them.

You now have:

- one 11 m conductor for H-field channel 1;
- one 11 m conductor for H-field channel 2.

Mark the conductors or use their original colour/stripe identification so they cannot be confused at the preamplifier.

### 2. Wind loop 1

Measure **1 m from the first end** of the H1 wire and mark it. Leave this first metre hanging loose as the connection end. Start winding at the bottom of the frame.

Follow this exact route: **bottom → left → top → right → bottom**. Returning to the bottom is one complete turn. Follow the same route two more times, always in the same direction, until you have made exactly three turns.

Keep the three turns together and secure them at the four frame endpoints. Do not pull so tightly that the wooden structure twists.

After three turns, stop at the bottom. Leave at least **1 m** of the second end loose. Do not cut off the remaining surplus until the preamplifier has been positioned and the antenna has been tested. The two loose ends of this one wire will become `H1+` and `H1−`.

### 3. Wind loop 2

Repeat the same procedure with the H2 wire around the perpendicular diamond. Leave 1 m loose, then follow **bottom → front → top → back → bottom** three complete times. Leave at least 1 m loose at the other end. These two loose ends will become `H2+` and `H2−`.

The two loop wires may physically pass near each other at the shared top and bottom points, but their copper conductors must never make electrical contact.

### 4. Add strain relief

Secure the connection tails to the frame before they reach the PCB. Movement of the antenna or cable should not pull directly on the preamplifier terminals.

## Testing before connection

Disconnect the antenna from all electronics and use a multimeter:

Set the meter to continuity mode. On most meters, a beep means that the two probe points are electrically connected. No beep means they are separate.

1. Measure between both ends of loop 1: continuity is expected.
2. Measure between both ends of loop 2: continuity is expected.
3. Measure between any loop-1 end and any loop-2 end: open circuit is expected.
4. Flex the wires near frame corners and connection tails while testing for intermittent breaks.

If the meter displays resistance, both loops should show a low and reasonably similar value. Exact equality is unnecessary. A major difference suggests damaged wire or a bad connection.

## Connecting PCB 23.1

Connect the first loop's two wire ends directly to:

- `H1+`
- `H1−`

Connect the second loop's ends directly to:

- `H2+`
- `H2−`

It does not matter which end of loop 1 is connected to `H1+` and which is connected to `H1−`; the antenna will receive signals either way. The same applies to H2. Never connect an H1 wire to an H2 terminal.

For this direct speaker-wire construction, **leave both solder jumpers open**, as in the working station-3304 build. Do not close a bridge merely because the wire was sold as “speaker cable”; the jumpers relate to the preamplifier input configuration, not the product name printed on the cable.

Do not connect either loop to a local earth stake or protective earth at the antenna.

## Design B: two documented coax loops

The project documentation also describes a coaxial H-field loop. In this design the coax **inner conductor** is the active three-turn winding. The surrounding metal shield helps reject electrical interference. The shield must have one complete break halfway along the cable so that it cannot act as a shorted loop.

Build two identical loops and mount their planes at approximately 90 degrees.

![Three-turn coax-loop wiring principle](blitzortung-coax-loop-wiring.svg)

### What to buy

- **21 m of coaxial cable**, which will be cut into two 10.5 m pieces;
- two non-conductive circular frames, each measuring 1 m straight across its centre;
- plastic cable ties or clamps;
- insulation and strain relief for both shield gaps and all cable ends;
- PCB 23.1 H-field preamplifier and its normal controller cable;
- multimeter with continuity mode.

For this direct-loop build, the official documentation permits inexpensive satellite coax with a copper-plated-steel centre conductor.

### Cut the two cables

1. Measure and cut the 21 m cable into **two equal 10.5 m pieces**.
2. Label one piece `H1` and the other `H2`.
3. Do not remove any cable yet.

Each 10.5 m piece provides:

| Purpose | Cable length |
|---|---:|
| Three rounds around one 1 m frame | approximately **9.42 m** |
| Two loose connection ends | approximately **1.08 m total** |
| Complete piece for one antenna | **10.5 m** |

In practice, leave approximately **54 cm loose at each end** and wind the cable section between them around the frame three times. Do not shorten the ends until everything works.

These measurements apply only to circular frames measuring 1 m straight across the centre.

### Make one break in the cable shield

Repeat these steps for both 10.5 m cables:

1. Lay the complete cable out straight.
2. Measure **5.25 m from either end** and mark the spot. That is exactly halfway.
3. Carefully remove a short ring of the plastic outer jacket at the mark.
4. Cut away a complete narrow ring of the metal foil and/or braided metal shield underneath.
5. **Do not cut or damage the insulated centre wire.** It must remain unbroken from one end to the other.
6. Check with a multimeter that the metal shield on one side of the gap is no longer connected to the shield on the other side.
7. Insulate and support the gap so that loose metal strands cannot reconnect it later.

The required result is simple: the centre wire remains continuous, while the surrounding metal shield has one complete break halfway along the cable.

### Wind and position the coax loops

1. Leave approximately 54 cm cable loose at the first end.
2. Wind the cable three complete times around its own 1 m frame.
3. Keep the three rounds together and secure them with plastic fasteners.
4. Leave the second end loose for connection to PCB 23.1.
5. Repeat for the second cable and frame.
6. Position the two completed loop planes at approximately 90 degrees to each other.

Each coax cable still has one unbroken centre conductor with two free ends. Never connect that centre conductor directly to its own shield.

### Connect the coax loops to PCB 23.1

PCB 23.1 has six antenna-input pins:

| PCB pin | Connect this |
|---:|---|
| 1 | Metal shields from **both ends** of coax loop H1 |
| 2 | Centre conductor from the first end of coax loop H1 |
| 3 | Centre conductor from the other end of coax loop H1 |
| 4 | Centre conductor from the first end of coax loop H2 |
| 5 | Centre conductor from the other end of coax loop H2 |
| 6 | Metal shields from **both ends** of coax loop H2 |

The two shields belonging to H1 both go to pin 1. The two shields belonging to H2 both go to pin 6. Do not connect the shield to an earth stake, protective earth or the antenna frame.

For this direct coax-loop construction, **leave both solder bridges open**. The fact that the cable may be sold as “75-ohm coax” is not a reason to close them.

### Test each coax loop before connecting the PCB

Set the multimeter to continuity mode and test:

| Touch the probes here | Correct result |
|---|---|
| Centre conductor at one end and centre conductor at the other end of the same cable | Beep/continuity |
| Shield at one end and shield at the other end of the same cable | No beep/open circuit |
| Any centre conductor and any shield | No beep/open circuit |
| Any part of H1 and any part of H2 | No beep/open circuit |

While testing, gently move the cable at the shield gap and both ends. The result must not change.

## Design C: ready-made FRA-200/12 ferrite antennas

This is the simplest route because the sensitive coils arrive ready-wound.

1. Order **two FRA-200/12 antennas** from the [Blitzortung ferrite-antenna supplier](https://www.ferrite-rod-antennas.de/page-4).
2. State that both are for one Blitzortung system so the supplier can try to select a closely matched pair.
3. Order them at approximately the same time as the System Blue station. They come from a separate supplier and parcel, so simultaneous delivery is not guaranteed.
4. Mount both rods **horizontally** at 90 degrees to each other. They may cross or lie side by side.
5. If the fine connecting wires are enamelled, carefully scrape off the varnish and solder on short, thicker insulated leads that the screw terminals can grip reliably.
6. Connect one complete ferrite winding across `H1+`/`H1−` and the other across `H2+`/`H2−`. The marked or knotted wire end is irrelevant for PCB 23.1's balanced inputs.
7. Use the high-impedance ferrite input arrangement; on PCB 23.1 this means leaving the loop-input solder bridges open.

Ferrite rods are compact, work indoors and need no exact compass alignment. Two rods at 90 degrees are sufficient.

## Other documented antenna constructions

The following designs are valid alternatives described in the [official magnetic-antenna documentation](https://docs.lightningmaps.org/hardware/antennas/hfield/). Designs D and E can be built from fixed measurements. Designs F, G and H require a proven winding specification, the official connection drawing or a correctly selected transformer core. That distinction matters: a beginner guide should not invent missing electrical details.

### Design D: classic direct wire loops

This is electrically similar to design A, but uses two separate flat frames instead of one three-axis diamond frame. Use insulated copper wire of at least `0.75 mm²`.

Choose **one** of these shapes and build two identical copies:

| Shape | Frame | Turns | Wire to prepare per antenna |
|---|---|---:|---:|
| Circle | 1 m diameter | 3 | **10.5 m** |
| Square | 1 m × 1 m | 3 | **13 m** |

For the circular version:

1. Make two non-conductive circular frames measuring 1 m straight across the centre.
2. Prepare two 10.5 m lengths of insulated copper wire.
3. Leave approximately 54 cm loose at the first end.
4. Wind the wire three complete times around the frame.
5. Leave the remaining end loose and do not trim it until installation is complete.

For the square version:

1. Make two non-conductive square frames measuring 1 m along each side.
2. Prepare two 13 m lengths of insulated copper wire.
3. Leave approximately 50 cm loose at the first end.
4. Follow all four sides and return to the starting corner. That is one turn and uses 4 m of wire.
5. Repeat the same route twice more, making three turns in total.
6. Leave approximately 50 cm loose at the other end.

Mount the two finished loop planes at 90 degrees. Connect the two ends of one loop to `H1+` and `H1−`; connect the other loop to `H2+` and `H2−`. The loops must not be connected to each other or to earth. Test them exactly like the direct speaker-wire loops in design A.

The official documentation also mentions a historical eight-turn, one-metre square used with System GREEN. System BLUE normally needs only three or four turns, so copying the old eight-turn version is unnecessary.

### Design E: compact 20-turn wire loops

This version trades frame size for more winding. The documented example uses a 38 cm diameter frame, 20 turns and `1 mm²` insulated copper wire.

For two antenna axes, obtain:

- two non-conductive circular frames, each **38 cm in diameter**;
- two **25 m lengths** of `1 mm²` insulated copper wire—**50 m total**;
- plastic fasteners and strain relief.

Build each axis as follows:

1. Leave approximately **55 cm** loose at the first end.
2. Wind the wire around the 38 cm frame exactly **20 complete times**, always in the same direction.
3. Count every return to the starting point as one completed turn. Mark every fifth turn while working so that you do not lose count.
4. Keep the turns together and secure them at several points.
5. Leave at least 55 cm loose at the other end; trim surplus only after installation.
6. Repeat for the second antenna and mount the two loop planes at 90 degrees.
7. Connect one coil to `H1+`/`H1−` and the other to `H2+`/`H2−`.

The expected inductance is roughly 500 µH, but a beginner does not need an inductance meter merely to reproduce the documented dimensions. Both coils should be built identically and should pass the same continuity and isolation tests as design A.

Historical System BLUE examples used the loop-input solder bridges with this design. Because preamplifier revisions and matching arrangements differ, confirm the required bridge setting for your exact PCB before soldering anything permanently.

### Design F: home-wound ferrite-rod antennas

The official documentation gives the mechanical limits but does **not** give one universal number of turns or wire diameter for a homemade ferrite coil. A precise recipe therefore requires a proven coil specification from the rod supplier or an already tested Blitzortung design.

The documented construction principles are:

1. Obtain two identical ferrite rods. Each rod should be between **12 cm and 30 cm long**; longer rods offer no useful advantage for this application.
2. Obtain a tested winding specification that states the enamelled-wire diameter, number of turns, winding length and preferably the expected inductance.
3. Wind both rods identically. Keep the winding neat and secure it with insulating tape, shrink tape or varnish.
4. Scrape the enamel completely from both fine wire ends.
5. Solder a short, thicker insulated connection wire to each delicate coil end so it can be clamped reliably in PCB 23.1.
6. Mount both rods horizontally at 90 degrees to each other.
7. Connect one complete winding across `H1+`/`H1−` and the other across `H2+`/`H2−`. The marked or knotted winding end does not determine polarity on the balanced PCB 23.1 input.

The 2 kΩ input arrangement is intended for ferrite antennas. On historical System BLUE preamplifiers this meant leaving the loop-input bridges open; verify the labels and documentation for the actual PCB revision in front of you.

Ferrite shielding is optional. If a conductive foil or metal tube is used, it must have a continuous lengthwise slot so it cannot form a closed conductive turn around the rod. For most beginners, the ready-made matched antennas in design C are the safer and easier ferrite route.

### Design G: Möbius coax loops

A Möbius loop cross-connects the coaxial centre conductor and shield so that two physical turns act as four active turns. It is compact and effective, but its connection is less forgiving than the ordinary coax loop in design B.

The official one-metre example is made as follows in principle:

1. Wind coax cable for **two physical turns** around a 1 m diameter non-conductive frame.
2. At the midpoint specified by the official drawing, cut and prepare both the centre conductor and shield.
3. Cross-connect the shield from one section to the centre conductor of the other section, and do the opposite for the remaining pair.
4. Insulate every joint separately and verify with a multimeter that no unintended short remains.
5. Build a second identical Möbius loop and mount its plane at 90 degrees to the first.

After the cross-connections are complete, each Möbius antenna must present exactly two intended output connections. Connect the first antenna output across `H1+`/`H1−` and the second across `H2+`/`H2−`, following the official diagram. If your construction leaves an unexplained third output, a joined centre conductor and shield, or continuity where the drawing shows isolation, stop and correct it before connecting the PCB.

Do not treat this paragraph as a substitute for the [official Möbius connection pictures](https://docs.lightningmaps.org/hardware/antennas/hfield/#mobius-antenna). A normal midpoint shield gap, as used in design B, does **not** turn an antenna into a Möbius loop. A mistaken cross-connection can instead short the received signal. Use this design only when you can follow and verify the official drawing conductor by conductor.

### Design H: transformer-coupled copper-tube loops

This construction uses a low-resistance metal loop as the one-turn primary winding of a current transformer. The documented compact example uses a loop approximately 38–40 cm in diameter made from 5–6 mm copper, brake or air-conditioning tube, together with a toroidal transformer carrying approximately 50 secondary turns.

The mechanical outline is:

1. Form the tube into a 38–40 cm circle while leaving a controlled connection gap.
2. Arrange the tube through or at the toroidal core exactly as shown in the official current-transformer drawing; the metal loop forms the primary side.
3. Wind the specified number of insulated secondary turns evenly around the toroid.
4. Connect the transformer secondary to the preamplifier through good-quality, pure-copper, heavily shielded 75 Ω coax.
5. Build a second identical assembly and mount the two loop planes at 90 degrees.

Each transformer secondary supplies one balanced H-field channel: the first goes to H1 and the second to H2. The precise coax shield and input-matching connections must follow the proven transformer drawing used for the chosen core; do not infer them from the ordinary direct coax loop in design B.

This is **not a universal parts recipe**. Toroid material, core size, winding wire, insulation and exact interconnection determine the result. The official documentation demonstrates the principle but does not specify one currently stocked core that every reader should buy. Select a proven transformer design or ask an experienced participant to verify the core and winding before construction.

### Design I: the E-field probe

An E-field antenna is not a loop. It is one high-impedance probe that detects changes in the electric field between the atmosphere and ground. It is omnidirectional, so only one probe is used.

If the controller revision provides the appropriate E-field input, the official basic construction is:

1. Use the controller's built-in probe, or cut one **10 cm** piece of `1–2.5 mm²` wire.
2. Connect the probe to the designated E-field input—not to `H1` or `H2`.
3. For a remote probe, use 75 Ω television coax and the specified F connector between the probe assembly and controller.
4. Position the probe as high as practical and several metres away from mains wiring, power supplies, LED lighting, motors and other electrical equipment.
5. Start with the 10 cm length. A longer probe produces more signal but usually also more local noise; longer is not automatically better.

An exposed outdoor E-field probe is a conductive object in an electrically active environment. A beginner should keep it indoors or use a properly engineered protected installation, never improvise an exposed mast near overhead wiring.

### Choosing between designs D to I

| If you want… | Choose… |
|---|---|
| A familiar large loop with straightforward wiring | Design D |
| A smaller loop frame and do not mind winding 20 turns | Design E |
| A compact homemade ferrite experiment with a known coil specification | Design F |
| An advanced coax experiment and can verify every cross-connection | Design G |
| A rugged advanced loop and understand current transformers | Design H |
| One simple omnidirectional probe and have a suitably quiet location | Design I |

For a first station, designs A or C remain the least ambiguous choices.

## Controller setup for H-field designs A to H

1. Connect the PCB 23.1 preamplifier to the correct amplifier port on the System Blue Basic/Mini controller.
2. Connect the separately purchased external active GNSS antenna to the on-board GNSS receiver, then connect Ethernet and the specified 5 V USB-C supply.
3. Wait for stable GNSS timing and network status.
4. Open the controller's **Signals** page.
5. Confirm that channels `1A` and `1B` are present and responsive.
6. Begin with the normal automatic settings, as used for this build.

On the original installation both channels worked immediately and showed almost equal noise floors around **13–14 mVpp**. `mVpp` is simply the controller's way of expressing the height of the background-noise trace. This result comes from one location; it is not a value every station must reproduce.

The official web-interface reference gives approximately 20–25 mV as a useful noise-level target when manually adjusting gain. This is an adjustment guideline, not a requirement to make a stable automatic configuration at 13–14 mV noisier.

Leave normal sampling at **500 kSps**, **512 samples** and **256 samples before trigger**. These are controller recording settings; a beginner does not need to calculate or change them unless an experienced project member gives a specific reason.

## Placement

An H-field loop does not need a view of the sky and does not need to be mounted high. Separation from electrical interference is usually more important.

Keep the antenna several metres away where possible from:

- switch-mode power supplies and USB chargers;
- computers and monitors;
- LED lighting, dimmers, WLED controllers and ESP devices;
- Ethernet switches, PoE equipment and mains wiring;
- motors, inverters and appliance electronics.

Move the antenna or switch suspected devices off one at a time before redesigning a working loop.

The station-3304 antenna described here stands indoors on a dry attic. That is a perfectly valid installation: an H-field antenna does not need to see the sky.

Outdoor weatherproofing is therefore **not part of the tested build**. Only readers who deliberately choose an outdoor location need to protect the wood, wire, connections and preamplifier from rain, condensation and UV. Unnecessary outdoor height also increases weather and lightning exposure without improving magnetic-field reception by itself.

## Reading the important indicators

| Indicator | Meaning |
|---|---|
| Mode, green | Controller voltages and basic system state are okay |
| Network, green | Connection to Blitzortung.org is established |
| GPS, blue, once per second | GNSS timing is operating |
| Signal, yellow | A signal triggered the detector |
| Fault, yellow | Too many unwanted signals are being received |

A busy Fault LED normally points to interference, excessive triggering or unsuitable gain—not automatically to a defective antenna or controller.

On the Signals page:

- a green timestamp means the signal was released for transmission;
- a red timestamp means it was invalid or filtered;
- `L` beside a channel means that channel stayed below its threshold for that event;
- one loop can trigger while the perpendicular loop remains below threshold;
- the browser may not display every event at high signal rates.

## Troubleshooting

### Only one channel works

- Repeat continuity measurements on both loops.
- Check that loop 1 is on `H1+`/`H1−` and loop 2 on `H2+`/`H2−`.
- Inspect stripped wire ends and terminal clamping.
- Verify that the two individual speaker-wire conductors were not accidentally joined or confused.
- Swap the complete loops between H1 and H2. If the problem follows the loop, inspect the antenna; if it stays on the channel, inspect configuration and preamplifier wiring.

### Noise floors differ greatly

- Confirm equal turn count and similar connection-tail length.
- Check resistance and connections.
- Rotate or move the structure: a local noise source may couple much more strongly into one loop plane.
- Keep both loops equally far from mains and data cables where possible.

### Fault LED is very active

- Check the live trigger rate and waveforms.
- Temporarily switch off nearby electronic equipment one device or circuit at a time.
- Increase distance from the suspected source.
- If gain is being adjusted manually, reduce excessive gain gradually and observe both channels.
- Do not enable every filter immediately; that can hide the source while discarding useful signals.

## What speaker-wire design A does not require

- Coaxial cable.
- A coax shield interruption.
- A third shield terminal.
- Ferrite rods.
- A transformer.
- A tuning capacitor.
- Exact north/south alignment.
- A tall mast.
- Copying another station's gain and threshold values.

## What success looks like

A useful station does not need the highest trigger rate or largest waveform. Look for:

- stable GNSS timing and network connection;
- both H-field channels visible and responsive, or one responsive E-field channel when design I is used by itself;
- similar, manageable background levels;
- occasional clean broadband pulses;
- no continuous interference mode;
- useful station participation statistics over days and weeks.

Station 3304 demonstrated that this inexpensive speaker-wire construction can provide two balanced, working channels. That validates it as a practical beginner design—not as the only correct antenna.

## Items for community review

Experienced participants are invited to review:

1. the PCB 23.1 jumper explanation across current board revisions;
2. recommended automatic/manual commissioning practice for current firmware;
3. optional weatherproofing methods for readers who want to adapt this indoor attic design for outdoor use;
4. proven current component specifications for the home-wound ferrite and transformer-coupled alternatives;
5. useful waveform examples distinguishing lightning from common household interference;
6. any terminology that may still confuse a first-time participant.

## Primary references

- [Magnetic antennas](https://docs.lightningmaps.org/hardware/antennas/hfield/)
- [Electric antennas](https://docs.lightningmaps.org/hardware/antennas/efield/)
- [General antenna information](https://docs.lightningmaps.org/hardware/antennas/general/)
- [Current detector and PCB 23.1 hardware information](https://www.limaps.org/hardware.html)
- [Initial setup](https://docs.lightningmaps.org/station-operation/setup/)
- [Controller overview and LEDs](https://docs.lightningmaps.org/station-operation/overview/)
- [Signal examination](https://docs.lightningmaps.org/station-operation/signal_examination/)
- [Web interface reference](https://docs.lightningmaps.org/station-operation/web_interface_reference/)
- [Current System Blue Basic order list](https://www.limaps.org/Order-List.pdf)
- [FRA-200/12 ferrite antenna alternative](https://www.ferrite-rod-antennas.de/page-4)
- [Antennas for System BLUE — Mini Guide](https://forum.blitzortung.org/mybb/showthread.php?tid=1708)

---

**Draft status:** v0.3 documents the actual indoor attic installation of the station-3304 speaker-wire antenna, a complete coax and ready-made ferrite route, and practical construction outlines for the other antenna types in the official documentation. Component-dependent advanced designs remain explicitly marked for community verification.