There is very little information available about this frequency band on the Chinese internet. I have gathered some translated materials from various sources to provide a starting point for ham radio enthusiasts who are unfamiliar with this band, and I hope that more people will become active in this band. Due to time constraints, I am still working on improving the content, and any corrections or suggestions are welcome.
History and Origins
136 kHz was likely the first frequency band used by amateur radio enthusiasts. In the early days of radio development, before any regulations were in place regarding electromagnetic waves, early radio enthusiasts were experimenting with coils, Leyden jars, and transformers similar to those found in modern crystal radios, creating sparks within their homes and directing a small amount of energy into the antennas, which then radiated out into the vast sky.
However, in reality, 136 kHz is a relatively young amateur band.The legal allowance for 136 kHz in the amateur band began on November 9, 2007.On that day, at the International Telecommunication Union's (ITU) 2007 World Radiocommunication Conference, which was held in Geneva. （WRC-07） I agree to allocate the frequency range of 135.7–137.8 kHz for amateur use as a secondary service. Although some countries had already based their allocations on the ITU allocation before it was formally established.CEPT/ERC Recommendation 62-01 E ("Use of the 135.7–137.8 kHz band by the Amateur Service", Mainz, 1997)Allocate this frequency band to amateur use. This proposal suggests allocating the 135.7-137.8 kHz range to amateur use as a secondary service, with a maximum power of E.I.R.P. 1W. Prior to this, the 130–148.5 kHz band is primarily used for maritime mobile and fixed services. The main users are single-direction transmissions from submarines and radio location systems.
Furthermore, some countries in ITU Region 2 have also authorized the LowFER band, such as the United States and Canada, which allow anyone (without requiring a radio operator's license, only needing to submit an application) to conduct radio communication experiments between 160-190 kHz. However, they require:
- The total input power at the final RF stage (excluding the filament or heater power) shall not exceed 1 watt.
- The total length of the transmission line, antenna, and ground wire (if used) shall not exceed 15 meters.
- All spurious emissions below 160 kHz or above 190 kHz must be at least -20 dB below the unmodulated carrier. The determination of whether the requirement of -20 dB attenuation is met can be based on measurements at the radio's antenna output end. If the radio uses a permanent additional antenna, compliance should then be verified by measuring the radiated electromagnetic field.
The current state of regulations in China
The amateur radio frequency band of 135.7-137.8 kHz in China originated from the implementation of the "Regulations on Radio Frequency Allocation in the People's Republic of China" (Decree No. 16 of the Ministry of Industry and Information Technology), which came into effect on December 1, 2010.
To adapt to the development of radio business and radio technology, and to meet the needs of various industries and departments in China for radio frequency resources, while ensuring consistency with the "Regulations on Radio Management of the People's Republic of China," the Ministry of Industry and Information Technology decided in 2008 to revise the "Regulations on Radio Frequency Allocation of the People's Republic of China," which were promulgated and implemented in 2006.
The current status of amateur radio frequency allocation in this band is as follows:Secondary business
135.7–137.8 Fixed, Waterborne, Unlicensed Radio Navigation [Amateur]
5.64 Fixed service radio stations are only permitted to use A1A, F1B, A2C, A3C, F1C, or F3C class transmitters within the 90-160 kHz (148.5 kHz in Zone 1) band allocated to this service, and waterborne mobile service radio stations are only permitted to use A1A, F1B, A2C, A3C, F1C, or F3C class transmitters within the 110-160 kHz (148.5 kHz in Zone 1) band allocated to this service. Exceptionally, waterborne mobile service radio stations may also be permitted to use J2B or J7B class transmitters within the 110-160 kHz (148.5 kHz in Zone 1) band.
5.67 Additional Allocation: In Mongolia, Kyrgyzstan, and Turkmenistan, the 130-148.5 kHz band is also allocated for radio navigation services as a secondary business. Within these countries and between them, this service has equal operating rights.（WRC-07）
5.67A: Amateur radio stations using frequencies within the 135.7-137.8 kHz band shall not have a maximum radiated power exceeding 1 watt (e.i.r.p.), and shall not cause harmful interference to any radio navigation station operating within the countries listed in Section 5.67.（WRC-07）
5.67B is restricted to fixed and maritime mobile services in Algeria, Egypt, the Islamic Republic of Iran, Iraq, Lebanon, the Syrian Arab Republic, Sudan, South Sudan, and Tunisia, within the 135.7-137.8 kHz frequency band. In the above countries, amateur radio use of the 135.7-137.8 kHz frequency band is prohibited. Countries that are authorized to use this frequency band should take this into account.（WRC-12）
Satellite communication system
136 kHz antennas are almost all shortened antennas, and their efficiency is almost less than 1%. This means that to meet the legal requirement of E.I.R.P. of 1 W, you often need several hundred to thousands of watts of power.
The most common antenna in this frequency band is a vertically polarized antenna (also known as a Marconi antenna).
The picture shows the T-shaped antenna of WD2XNS. The website for WD2XNS
The picture shows the Titanic's "E" antenna.
The image shows a typical T-shaped antenna design.
This antenna can be summarized from top to bottom as:
A horizontal conductor that forms a load for the ground potential
Main downward-pointing energy radiation line
The inductor responsible for matching.
The horizontal top section and the ground form a capacitor, resulting in negligible radiation. To increase the capacitance value, multiple parallel conductors can be arranged side by side to improve radiation efficiency.
The vertical portion is almost the only part that can effectively radiate. Due to the long wavelength of the 136 kHz band, even with horizontal antenna heights of tens of meters, ground effects are still significant at this time.Only the vertical part can effectively radiate energy.Therefore, the design should aim to make the vertical part as high as possible. An ideal Marconi antenna should be at least 1/4 wavelength, which is over 500 meters. This height is practically impossible for amateur radio operators. Therefore, antenna designs typically compromise with practical installation conditions and allow ham radio enthusiasts to fully utilize their subjective initiative and take advantage of the surrounding environment in their design.
For such a short antenna, there should theoretically be very small parasitic resistance (typically a few ohms) and a large negative inductance. To compensate for the large inductance, an inductor is used at the bottom to resonate with the antenna. In order to maximize radiation efficiency, a high Q-value is sought, typically achieved using a large, self-made, hollow inductor. Another effect of a high Q-value is a very narrow antenna bandwidth. To ensure that the antenna can resonate at 135.7-137.8 kHz, an adjustable inductor design is also required.
The variable inductance of WD2XNS uses an internally wound inductor and a reduction motor to fine-tune the inductance value.
The inductor for W5JGV has a series of terminals designed for adjustment.
Similar ferrite inductors are also feasible, but ferrite inductors generally have a lower Q-factor than air inductors, typically only in the tens. The inductor shown here is wound using PC95 material and has a value of 6 mH.
Due to the large size of inductors, it's not common to load an inductor in the center; however, some ham enthusiasts have successfully done so.
The ON7YD vertical antenna with inductive top-load. The original L1 coil was divided into two parts: L1 = 2.3 mH and L2 = 1.9 mH. Measurements using PA0SE and DK8KW showed an increase in antenna field strength of approximately 4 dB.
At this point, if the inductor completely cancels out the impedance, the antenna impedance is not necessarily 50Ω. It still needs to be combined with a transformer (also known as a "balun" in ham radio) or other matching circuit to further transform the impedance.
The picture shows the transformer used by WD2XNS, which is made of FT240-77 magnetic core material. FT240-77 is a common manganese-zinc ferrite material used for frequencies below 1 MHz, and can be replaced with domestically produced MXO-2000.
Here are some useful reference websites:
ON7YD, Longwave 136 kHz antenna introduction(Highly recommended)
Short-wave tuning
Furthermore, there are other types of antennas.
The G3YMC loop antenna
@BH6BEZ In a 2200m practical test, an 80-meter long wire antenna with a terminal load was used. With a power of 10 watts, a signal strength of FST4W-120 -11dB was achieved at a distance of 110 km.
Common patterns
CW
The most basic and universal mode, but it's not common in the longwave band because CW requires a much higher signal-to-noise ratio than other modes.
QRSS and DFCW
QRSS It is a very slow CW mode, named after "QRS," which stands for "Please Reduce Sending Speed." In this mode, the receiver's bandwidth is significantly reduced, and the WPM rate slows down to a speed that is easily readable by humans, typically less than 2-3 wpm.
Typically, we use "dot time" to quantify QRSS speed, which is the number of seconds required to send a single "dot." For example, 3 dots means sending each "Di" for 3 seconds, which equals 0.4 wpm. 60 seconds worth of dots equates to 0.022 wpm. Therefore, transmitting my callsign "BH3PTS" using 60 seconds worth of dots would take a full 1 hour and 3 minutes.
Because the noise received is proportional to the receiver bandwidth, using a bandwidth less than 1 Hz is common in QRSS. This allows the receiver to select a very slow spectrum rate, resulting in an average of one audio pixel per second. By averaging the audio in this way, the noise level (which is random and averages to zero) will be lower than the signal level, allowing signals below the ambient noise floor to appear on the spectrum, which can then be visually identified by the human eye on the spectrum. However, because the bandwidth used is narrow, frequency stability is crucial for both the receiver and the transmitter.
DFCWSimilar to QRSS, the key difference is that DFCW uses very small frequency offsets (even as little as 0.1 Hz) to represent points and lines instead of lengths. This has the advantage of reducing transmission time, but requires a wider bandwidth, a more stable crystal oscillator, or a DDS. Because two frequencies are used, each frequency is susceptible to noise and interference.
Currently popular QRSS receiving softwareFSKview ，Spectrum Lab、Argo、LOPORA And QRSSpig
PSK31
PSK31 was developed and named by British amateur radio operator Peter Martinez, G3PLX, and was introduced to amateur radio in December 1998. PSK31 works well on phase-modulated propagation paths and is resistant to fading. （QSB） It has good performance. However, it may be affected by propagation patterns (such as cross-polar paths), where scintillation or multipath effects disrupt the continuity of signal phases.
Based on experience, the signal-to-noise ratio required for decoding PSK31 is often significantly lower than the minimum signal-to-noise ratio that a human ear can detect CW signals, typically by 6 dB. However, PSK31 also has its drawbacks.Requires linear transmission equipment and amplifiers.Therefore, it cannot be transmitted using inexpensive and efficient D-class amplifiers or power amplifiers. As various FSK (Frequency Shift Keying) modes gained popularity, PSK31 gradually faded into obscurity.
JASON
JASON Originating from 2005Based on Steve Olney's work (VK2ZTO) Developed IFK (In-Phase Frequency Shift Keying) Technology. This technology effectively avoids errors caused by receiver detuning and frequency drift, while also enhancing resistance to multipath propagation and inter-symbol interference. Specifically, information is encoded using the absolute value of the difference between the two transmitted frequencies. The advantage of this method is that it does not require precise initial tuning; even a few Hertz of tuning error can be tolerated. Conservatively estimated,JASON requires a frequency stability of ±31 ppm at a frequency of 136 kHz.Additionally, because the frequencies are transmitted one at a time using FSK mode,No linear amplifier is required.A single-class MOSFET transmitter is sufficient to meet the requirements. Although Jason transmits fewer characters per minute (approximately 2.5 characters), it still has some advantages compared to QRSS.
JASON designed a total of 16 different frequency offsets. After transmitting one tone, the next tone will adjust the appropriate frequency offset up or down as needed, depending on the setting of the USB/LSB switch. With 16 possible frequency offsets, the system requires 17 tone slots (i.e., notes), and any overflow will result in tone looping.
Each bit (where a bit refers to a change in the signal, specifically frequency changes) is encoded as 4 bits (i.e., one half-byte, also known as a nibble). However, since 4 bits cannot satisfy the needs of the entire alphabet, we use two half-bytes to represent a character.
JASON achieves character synchronization, meaning how to determine the high and low nibbles. This information is encoded by using the highest bit of each nibble. The high nibble uses '1xxx' (binary), while the low nibble uses '0xxx' (binary). xxx Represents actual transmitted data.
Therefore, 6 bits are sufficient to represent a character, allowing for the encoding of up to 64 symbols. I chose the character set from x'20' (space) to x'5f' (underscore) in the ASCII code. This allows us to transmit all uppercase letters, 10 digits, and commonly used punctuation marks.
The JASON standard speed (Normal) settings are as follows:
The default center frequencies for the receiver (Rx) and transmitter (Tx) are 800 Hz.
The interval between each of the 17 tones is 3 FFT bins, to ensure orthogonality.
For standard speed settings, each FFT bin corresponds to approximately 0.084 Hz, so the distance between frequency slots is approximately 0.252 Hz, and the total bandwidth occupied is approximately 4.038 Hz.
Each period (i.e., each transmitted tone) lasts approximately 11.89 seconds (i.e., the inverse of the FFT bin width).
Therefore, at standard speed settings, JASON achieves a throughput of approximately 2.5 characters per minute. While this is slower than QRSS, it still offers some advantages.
The parameters for each of the three JASON modes are as follows:
| Speed | Pitch duration (seconds) | Frequency interval (Hz) | Bandwidth (Hz) | Character count per minute (Turbo off) | Character count per minute (Turbo on) |
|---|
| Slow | 95.2 | 0.03 | 0.5 | 0.3 | 0.6 |
| Normal | 11.9 | 0.25 | 4 | 2.5 | 5 |
| Fast | 1.5 | 2 | 32 | 20 | 40 |
JASON can -25 dB achieve effective communication under appropriate signal-to-noise ratios.
JASON's website:Jason's official website
WSPR
WSPR(Pronounced "whisper") is Weak Signal Propagation Reporter The acronym. It is a protocol implemented in a computer program, specifically designed for weak-signal radio communication between amateur radio operators. The protocol is defined by Joe Taylor（K1JT）A program was designed and written in 2008 for use at medium frequencies.（MF）And high frequency（HF）Conduct low-power transmission experiments in the frequency band to test the performance of radio wave propagation paths. Due to its high sensitivity, WSPR quickly became 136 kHz It has been widely used in low-frequency weak signal communication.
The operation of WSPR is similar to beacon transmission mode, where each transmission sends the operator's call sign, grid locator, and transmit power (in dBm).
A key feature of WSPR is its ability to communicate over long distances using very low transmit power, making it an ideal tool for testing radio wave propagation, conducting remote signal monitoring, and collecting large amounts of transmission data. Its sensitivity can reach 2500 Hz -28 dB.
WSQ2
Introduction
WSQ is a weak signal QSO mode designed for LF (low frequency) and MF (medium frequency) bands, aiming to provide similar sensitivity to WSPR while being faster than modes like JASON. It's suitable for real-time communication in very weak signal conditions. WSQ was jointly developed by Con ZL2AFP and Murray ZL1BPU in 2013, using a new incremental frequency modulation technique.（IFK）Design, and combine with text compression technology, to enable it to operate at a signal-to-noise ratio of less than -27 dB.（SNR）To enable effective communication. ZL2AFP hoped to achieve the sensitivity of WSPR, while also having a faster transmission speed than the then-popular JASON, in order to adapt to two-way QSO communication. The original version of WSQ was written in late 2013. Compared to common digital modes at the time, WSQ offered higher signal reception sensitivity and faster communication speeds. In particular, compared to JASON (-25db) and DominoEX (-16db), WSQ had a significant advantage in symbol transmission efficiency, allowing for faster information exchange even under very weak signals.
Unlike WSPR, which requires complex error correction, WSQ achieves this through the use ofIFK Incremental Frequency KeyingThis design avoids errors caused by the receiver's detuning and frequency drift, and also resists multi-path propagation interference and inter-symbol interference. ThroughLong-term accumulationWSQ can effectively suppress impact noise and reliably receive signals at -27 dB SNR. It uses aVariable-length encoding (Varicode)Frequently used characters are transmitted using fewer symbols, which significantly increases transmission speed.
WSQ uses 33 frequencies, with a frequency spacing of 1.953125 Hz, and employs MFSK modulation with phase coherence. Each symbol lasts for 2.048 seconds. Each symbol carries more information, so even though the symbol rate is low (0.512 bps), the transmission speed can reach 5 to 7 letters per minute, which is significantly higher than other modes such as JASON. The design of WSQ allows it to operate efficiently in noisy environments, especially for weak signal communication in the LF/MF band. The WSQ is primarily designed for real-time QSO (two-way communication) and is not suitable for use as a signal detector or beacon.It is suitable for weak signal communication in the 160m, 630m, and 2200m bands, especially in environments with deep fading and slow signal degradation.
The original version of WSQ2ZL2AFP WSQ
Improved version of DL4YHFThe extended version of WSQ / WSQCall by DL4YHF
JT9
While JT65 is well-suited for longwave propagation with very weak but slow variations, its initial design for EME and tropospheric scattering results in a very wide frequency range, making it unsuitable for transmission within the narrow 136kHz band. Joe K1JT, using the same logic encoding as JT65, developed JT9, which uses a 9-FSK encoding optimized for MF and HF bands. It can also be used on 136kHz. JT9 has a sensitivity approximately 2dB higher than JT65A, with a minimum decoding capability of -27dB (2500Hz bandwidth). Furthermore, it occupies a bandwidth less than 16Hz.
FST4 and FST4W
FST4 and FST4W are two modern digital communication protocols specifically designed for longwave and mediumwave. In these frequency bands, the protocol's base sensitivity is superior to that of WSJT-X modes using similar sequence lengths, approaching its theoretical maximum data throughput rate. FST4 is primarily optimized for bidirectional communication (QSOs), while FST4W is suitable for transmitting quasi-CW beacons similar to WSPR. Unlike modes like EbNaut, which require strict independent time synchronization and phase locking, FST4 and FST4W do not have these requirements.
These two new modes use 4-GFSK modulation. FST4 provides T/R sequence lengths of 15, 30, 60, 120, 300, 900, and 1800 seconds, while FST4W omits lengths below 120 seconds. The names for each sub-mode are FST4-60, FST4W-300, etc., with the numbers indicating the sequence length (in seconds). The message payload is the same as FT4, FT8, and MSK144, containing 77 bits of data. The WSPR-style messages for FST4W contain 50 bits of data. Similar to other 77-bit and 50-bit modes, the message format displayed by WSJT-X is also the same. Forward error correction uses low-density parity check.（LDPC）The code consists of 240 bits, including information and parity bits. The transmission is composed of 160 symbols: 120 symbols carrying information, each symbol containing two bits, separated by five groups of predefined 8 synchronization symbols.
The basic parameters for each mode of FST4 and FST4W are shown in the table. The sensitivity threshold (i.e., the signal-to-noise ratio when the decoding success probability is 50% under a bandwidth of 2500 Hz) was determined through experiments with additive white Gaussian noise.（AWGN）Simulations are performed on the channel.
| Launch time(s) | Symbol Length (s) | Frequency interval (Hz) | Bandwidth usage (Hz) | FST4 Signal-to-Noise Ratio (dB) | FST4W Signal-to-Noise Ratio (dB) |
|---|
| 15 | 0.060 | 16.67 | 67.7 | 20.7 | |
| 30 | 0.140 | 7.14 | 28.6 | 24.2 | |
| 60 | 0.324 | 3.09 | 12.4 | 28.1 | |
| 120 | 0.683 | 1.46 | 5.9 | 31.3 | 32.8 |
| 300 | 1.792 | 0.56 | 2.2 | 35.3 | 36.8 |
| 900 | 5.547 | 0.180 | 0.72 | 40.2 | 41.7 |
| 1800 | 11.200 | 0.089 | 0.36 | 43.2 | 44.8 |
FST4-60 has a sensitivity approximately 1.7 dB higher than JT9, primarily because it uses multi-symbol block detection under appropriate conditions. When decoding AP signals using FST4, this difference can reach 4.7 dB. The sensitivity of FST4-120 and longer sequence lengths is relatively higher. FST4W-120 has a sensitivity approximately 1.4 dB higher than the standard WSPR, while with a 30-minute sequence, FST4W-1800 can achieve a threshold SNR of -45 dB. We strongly recommend that users who are currently using JT9 and WSPR in the LF and MF bands switch to FST4 and FST4W.
You may also consider other applications for these new modes, besides those discussed here. Please note that these modes have very narrow bandwidths; to achieve the sensitivity levels listed in the table, the drift and path-induced Doppler shifts must be smaller than the pitch interval, and this must be maintained throughout the entire sequence length. For example, the short sequence sub-mode FST4-15 performs very effectively on a 50 MHz ionospheric scattering path. In extreme applications where transmission time is long, VK7MO and VK7ZBX successfully used FST4W-1800 to perform non-line-of-sight optical scattering communication across a path of up to 153 km, using an LED array, a Fresnel lens, and a photodetector.
Other modes
闽ICP备2021006864号-7
闽公网安备35020602002794号