ICAO NDB Specifications

[ChrisSmolinski]

I stumbled on this while trying to figure out why NDBs use 1020 Hz MCW for the ID. Which I was researching while looking into the possibility of writing software to perform automated decoding of the entire NDB band, a la Amalgamated DGPS. There's a lot of variability in NDBs, carrier frequency, MCW frequency, CW speed, and repetition rate of the ID. That is good and bad. It is good because it means there is more chance for NDBs to (eventually over time) not overlap each other, so you can pick out a weaker station even if a stronger station is on the same frequency. It is bad because computers and software are not as clever as humans at decoding CW under poor and variable conditions.

I never could find an explanation as to why this (dumb to me) choice was made. But here's the ICAO rules on NDBs. I figure this is as good of a place as any to post them?



3.4 Specification for non-directional radio beacon (NDB)

3.4.1 Definitions
Note.— In Attachment C, guidance is given on the meaning and application of rated coverage and effective coverage and on coverage of NDBs.
Average radius of rated coverage. The radius of a circle having the same area as the rated coverage.
Effective coverage. The area surrounding an NDB within which bearings can be obtained with an accuracy sufficient for the
nature of the operation concerned.
Locator. An LF/MF NDB used as an aid to final approach.
Note.— A locator usually has an average radius of rated coverage of between 18.5 and 46.3 km (10 and 25 NM).
Rated coverage. The area surrounding an NDB within which the strength of the vertical field of the ground wave exceeds the minimum value specified for the geographical area in which the radio beacon is situated.
Note.— The above definition is intended to establish a method of rating radio beacons on the normal coverage to be expected in the absence of sky wave transmission and/or anomalous propagation from the radio beacon concerned or interference from other LF/MF facilities, but taking into account the atmospheric noise in the geographical area concerned.

3.4.2 Coverage

3.4.2.1 Recommendation.— The minimum value of field strength in the rated coverage of an NDB should be
70°microvolts per metre.
Note 1.— Guidance on the field strengths required particularly in the latitudes between 30°N and 30°S is given in 6.1 of Attachment C, and the relevant ITU provisions are given in Chapter VIII, Article 35, Section IV, Part B of the Radio Regulations.
Note 2.— The selection of locations and times at which the field strength is measured is important in order to avoid abnormal results for the locality concerned; locations on air routes in the area around the beacon are operationally most significant.

3.4.2.2 All notifications or promulgations of NDBs shall be based upon the average radius of the rated coverage.
Note 1.— In classifying radio beacons in areas where substantial variations in rated coverage may occur diurnally and seasonally, such variations should be taken into account.
Note 2.— Beacons having an average radius of rated coverage of between 46.3 and 278 km (25 and 150 NM) may be designated by the nearest multiple of 46.3 km (25 NM) to the average radius of rated coverage, and beacons of rated coverage over 278 km (150 NM) to the nearest multiple of 92.7 km (50 NM).

3.4.2.3 Recommendation.— Where the rated coverage of an NDB is materially different in various operationally significant sectors, its classification should be expressed in terms of the average radius of rated coverage and the angular limits of each sector as follows:
Radius of coverage of sector/angular limits of sector expressed as magnetic bearing clockwise from the beacon.
Where it is desirable to classify an NDB in such a manner, the number of sectors should be kept to a minimum and preferably should not exceed two.
Note.— The average radius of a given sector of the rated coverage is equal to the radius of the corresponding circle- sector of the same area. Example:
150/210° – 30° 100/30° – 210°.

3.4.3 Limitations in radiated power
The power radiated from an NDB shall not exceed by more than 2 dB that necessary to achieve its agreed rated coverage, except that this power may be increased if coordinated regionally or if no harmful interference to other facilities will result.

3.4.4 Radio frequencies

3.4.4.1 The radio frequencies assigned to NDBs shall be selected from those available in that portion of the spectrum between 190 kHz and 1 750 kHz.

3.4.4.2 The frequency tolerance applicable to NDBs shall be 0.01 per cent except that, for NDBs of antenna power above 200 W using frequencies of 1 606.5 kHz and above, the tolerance shall be 0.005 per cent.

3.4.4.3 Recommendation.— Where two locators are used as supplements to an ILS, the frequency separation between the carriers of the two should be not less than 15 kHz to ensure correct operation of the radio compass, and preferably not more than 25 kHz in order to permit a quick tuning shift in cases where an aircraft has only one radio compass.

3.4.4.4 Where locators associated with ILS facilities serving opposite ends of a single runway are assigned a common frequency, provision shall be made to ensure that the facility not in operational use cannot radiate.
Note.— Additional guidance on the operation of locator beacons on common frequency channels is contained in Volume V, Chapter 3, 3.2.2.

3.4.5 Identification

3.4.5.1 Each NDB shall be individually identified by a two- or three-letter International Morse Code group transmitted at a rate corresponding to approximately 7 words per minute.

3.4.5.2 The complete identification shall be transmitted at least once every 30 seconds, except where the beacon identification is effected by on/off keying of the carrier. In this latter case, the identification shall be at approximately 1-minute intervals, except that a shorter interval may be used at particular NDB stations where this is found to be operationally desirable.

3.4.5.2.1 Recommendation.— Except for those cases where the beacon identification is effected by on/off keying of the carrier, the identification signal should be transmitted at least three times each 30 seconds, spaced equally within that time period.

3.4.5.3 For NDBs with an average radius of rated coverage of 92.7 km (50 NM) or less that are primarily approach and holding aids in the vicinity of an aerodrome, the identification shall be transmitted at least three times each 30 seconds, spaced equally within that time period.

3.4.5.4 The frequency of the modulating tone used for identification shall be 1 020 Hz plus or minus 50 Hz or 400 Hz plus or minus 25 Hz.
Note.— Determination of the figure to be used would be made regionally, in the light of the considerations contained in Attachment C, 6.5.

3.4.6 Characteristics of emissions
Note.— The following specifications are not intended to preclude employment of modulations or types of modulations that may be utilized in NDBs in addition to those specified for identification, including simultaneous identification and voice modulation, provided that these additional modulations do not materially affect the operational performance of the NDBs in conjunction with currently used airborne direction finders, and provided their use does not cause harmful interference to other NDB services.

3.4.6.1 Except as provided in 3.4.6.1.1, all NDBs shall radiate an uninterrupted carrier and be identified by on/off keying of an amplitude modulating tone (NON/A2A).

3.4.6.1.1 NDBs other than those wholly or partly serving as holding, approach and landing aids, or those having an average radius of rated coverage of less than 92.7 km (50 NM), may be identified by on/off keying of the unmodulated carrier (NON/A1A) if they are in areas of high beacon density and/or where the required rated coverage is not practicable of achievement because of:
a) radio interference from radio stations;
b) high atmospheric noise;
c) local conditions.
Note.— In selecting the types of emission, the possibility of confusion, arising from an aircraft tuning from a NON/A2A facility to a NON/A1A facility without changing the radio compass from “MCW” to “CW” operation, will need to be kept in mind.

3.4.6.2 For each NDB identified by on/off keying of an audio modulating tone, the depth of modulation shall be maintained as near to 95 per cent as practicable.

3.4.6.3 For each NDB identified by on/off keying of an audio modulating tone, the characteristics of emission during identification shall be such as to ensure satisfactory identification at the limit of its rated coverage.
Note 1.— The foregoing requirement necessitates as high a percentage modulation as practicable, together with maintenance of an adequate radiated carrier power during identification.
Note 2.— With a direction-finder pass band of plus or minus 3 kHz about the carrier, a signal to noise ratio of 6 dB at the limit of rated coverage will, in general, meet the foregoing requirement.
Note 3.— Some considerations with respect to modulation depth are contained in Attachment C, 6.4.

3.4.6.4 Recommendation.— The carrier power of an NDB with NON/A2A emissions should not fall when the identity signal is being radiated except that, in the case of an NDB having an average radius of rated coverage exceeding 92.7 km (50 NM), a fall of not more than 1.5 dB may be accepted.

3.4.6.5 Unwanted audio frequency modulations shall total less than 5 per cent of the amplitude of the carrier.
Note.— Reliable performance of airborne automatic direction-finding equipment (ADF) may be seriously prejudiced if the beacon emission contains modulation by an audio frequency equal or close to the loop switching frequency or its second harmonic. The loop switching frequencies in currently used equipment lie between 30 Hz and 120 Hz.

3.4.6.6 The bandwidth of emissions and the level of spurious emissions shall be kept at the lowest value that the state of technique and the nature of the service permit.
Note.— Article S3 of the ITU Radio Regulations contains the general provisions with respect to technical characteristics of equipment and emissions. The Radio Regulations contain specific provisions relating to necessary bandwidth, frequency tolerance, spurious emissions and classification of emissions (see Appendices APS1, APS2 and APS3).

3.4.7 Siting of locators

3.4.7.1 Recommendation.— Where locators are used as a supplement to the ILS, they should be located at the sites of the outer and middle marker beacons. Where only one locator is used as a supplement to the ILS, preference should be given to location at the site of the outer marker beacon. Where locators are employed as an aid to final approach in the absence of an ILS, equivalent locations to those applying when an ILS is installed should be selected, taking into account the relevant obstacle clearance provisions of the PANS-OPS (Doc 8168).

3.4.7.2 Recommendation.— Where locators are installed at both the middle and outer marker positions, they should be located, where practicable, on the same side of the extended centre line of the runway in order to provide a track between the locators which will be more nearly parallel to the centre line of the runway.

3.4.8 Monitoring

3.4.8.1 For each NDB, suitable means shall be provided to enable detection of any of the following conditions at an appropriate location:
a) a decrease in radiated carrier power of more than 50 per cent below that required for the rated coverage;
b) failure to transmit the identification signal;
c) malfunctioning or failure of the means of monitoring itself.

3.4.8.2 Recommendation.— When an NDB is operated from a power source having a frequency which is close to airborne ADF equipment switching frequencies, and where the design of the NDB is such that the power supply frequency is likely to appear as a modulation product on the emission, the means of monitoring should be capable of detecting such power supply modulation on the carrier in excess of 5 per cent.

3.4.8.3 During the hours of service of a locator, the means of monitoring shall provide for a continuous check on the functioning of the locator as prescribed in 3.4.8.1 a), b) and c).

3.4.8.4 Recommendation.— During the hours of service of an NDB other than a locator, the means of monitoring should provide for a continuous check on the functioning of the NDB as prescribed in 3.4.8.1 a), b) and c).
Note.— Guidance material on the testing of NDBs is contained in 6.6 of Attachment C.

[pinto vortando]

Interesting stuff, thanks for posting.

FWIW,   IIRC morse identifier dots are to be .1 sec duration and dashes 3x as long.  The time within a letter between dots and dashes is to be the length of a dot.  Time between letters of the ID group to be equal to the length of 3 dots.  NDB to transmit a 3 letter ID at intervals equal to 64 dots, LOM to transmit 2 letter ID (first 2 letters of associated ILS) at interval equal to 48 dots.
Now,  how to explain the beacons that miskey month after month???

As to the 1020 cycle tone, IIRC it has been in use for a long time, all the way back to the age of the radio range stations.  My guess is that a tone widely separated from the 3000 cycle tone used for the airway markers ID had to be employed so as to eliminate any possibility of confusion.  The pilot had to be sure of station passage.  Why 1020 cycles?  Again, my guess is since 1020 is an odd multiple of the 60 cycle power line frequency that it may have to do with cancellation of 60 cycle modulation of the signal which can be a concern.    

btw,  VOR and ILS stations also use 1020 cycle ID.  

[ChrisSmolinski]

I bet you're right, it has something to do with 60 Hz. Looking back, I see that even in the 1930s, the "LF/MF Four-Course Radio Ranges" system developed in the 1930s used 1020 Hz. This is a multiple of 60 Hz, perhaps they chose 1020 Hz as the closest harmonic they could get to 1000 Hz? (as a reasonable frequency) And once set, that frequency choice has been with us since.

[pinto vortando]

As mentioned in my post, the old radio range stations used 1020 cycle modulation and as you say it has been with us ever since.  Trying to find facts on these old stations today is tough as info is sparse.  It looks as though the 90 degree displaced  A (.-) and N (-.) signals were rotated by a goniometer which back in those days would have been motor driven.  So again, the 60 cycle power line frequency comes into play as the motor rotation is a function of the line frequency and number of poles.    The answer is buried somewhere back there in the past.  Check out Adcock antenna,  low frequency radio range.   

As for my speculation about the need to be a range station tone far enough removed from the airway markers so as to avoid confusion between them,  it is a question of which came first. Thinking about it,  it was probably the range stations first as lighted beacons were used back then to mark the airways.  So when the radio airway markers came along, 3000 cycles was chosen for them.  Just guessing but interesting to ponder this stuff.  Try to imagine trying to fly "on the beam" with static crashes and other variables distorting the signal with terrain looming all around...     

[jFarley]

Thanks, Chris!

[skeezix]

Here's what the FAA has to say about their NDBs:

https://www.faa.gov/documentLibrary/media/Order/6740.6.pdf


Spoiler- No historical information given on 400/1020 Hz modulation.

[NJQA]

This site ( http://k4che.com/FL8/FL81.htm ) claims that the choice was based on the use of an electric motor to generate the tone.  I don't know that I am convinced that is correct, but it is as reasonable as anything else.  The good thing about a 1020 Hz tone is that it doesn't fall on an adjacent channel carrier frequency, making it easier to hear the ID when you have stations on both channels.

[ChrisSmolinski]

Last night on the Ryver chat, skeezix posted this link to a paper: http://nvlpubs.nist.gov/nistpubs/jres/5/jresv5n4p897_A2b.pdf

It looks like they originally used 50 or 60 Hz, then switched to 1000 Hz produced by feeding 500 Hz AC to the transmitter. My guess is that they later switched to 1020 Hz as it was obtainable from 60 Hz power, either via a motor or electronic means.

It's nice that it isn't exactly 1000 Hz, but it still falls very close to the adjacent carriers. I would have preferred 900 or 1100 Hz, but we're stuck with what we have. Plus there's a +/- 50 Hz tolerance in the spec today, so it could be anywhere. Throw in the fact that the carrier is also not exactly on, and the two MCW signals can end up in a lot of places. Which is great for human decoding, since it increases the chance that two stations on the same or adjacent channels will end up with enough separation on the sidebands so you can pull out an ID. But tougher for automated decoding, which is how I started going down this rabbit hole. 

[pinto vortando]

In actual practice, there is enough slop between carrier frequency tolerance, modulation offsets (400 or 1020) and their tolerances along with varying ID repetition rates that under the right conditions and with a receiver with good filtering,  3 or 4 beacons all on the same assigned frequency can be heard simultaneously from widely different points on the compass. 

[skeezix]

In another paper

APPLYING THE RADIO RANGE TO THE AIRWAYS
RP155
By F. G. Kear and W. E. Jackson
Washington, August 8, 1929

II. INTERFERENCE BETWEEN RADIO RANGES

Two methods have been successfully adopted to differentiate be- tween radio ranges. The first method was to allocate different fre- quencies in the band from 285 to 350 kc. to each radio range. Par- ticular care was taken to separate adjacent beacon frequencies by at least 12 kc. When utilizing the ordinary aircraft receiver using two stages of shield grid tube, radio-frequency amplification, detector and two stages audio with a 6-foot rod antenna, it was possible to tune in either a beacon 100 miles ahead or 100 miles behind, when both were operating simultaneously on frequencies separated by 8 kc. The average power output from a beacon installation is 2 kw.

Read through a bunch of old NBS papers int the NIST Digital archives and its nothing short of absolutely fascinating. Even dipped into eLORAN while in there.


In a side note, there was a station called WWV on 290 kHz.  That was published in
Field-intensity measurements at frequencies from 285 to 5,400 kilocycles per second
RP429
By S. S. Kirby and K. A. Norton
January 19, 1932