Report of the ARRL Ad Hoc HF Digital Committee
Dissenting Recommendation

Prepared by Committee Member Howard Teller, KH6TY

Why is a dissenting recommendation necessary?

The majority recommendation of the ARRL hfdigital committee represents the interests of a small special interest group, Winlink, representing 0.7% of the FCC licensed radio amateurs in the United States, and NOT the interests of ALL radio amateurs, as requested by ARRL president, Jim Haynie.

The ARRL hfdigital committee majority recommendation [hfdigital:284] was composed and written jointly by the Winlink author, W5SMM, and the Winlink Network Administrator, K4CJX, and then rubber-stamped, without comment except for typo corrections, by the Winlink supporters on the committee, WA1LOU and K0PFX. All alternative recommendations were totally ignored  by the chairman and Winlink supporters. At no time during the committee discussions did WA1LOU or K0PFX submit any independent comments or suggestions except for corrections. The chairman of the committee, who is also the Winlink software author, consistently shut off discussion prematurely and forced a vote, which was naturally won by the Winlink majority, eventually resulting in the resignation of the widely respected Peter Martinez, G3PLX, in protest for having his views silenced.

The committee majority recommendation therefore represents the views of a special interest group, and NOT the interests of all radio amateurs.

This dissenting recommendation is submitted as a recommendation that at least attempts to consider the needs of ALL radio amateurs, including Winlink.


Winlink consists of a network of fully automated digital transceivers, providing free email gateways to the Internet using the ham bands in competition with commercial LEO satellite services which do the same thing, but charge for messaging on a per-minute basis.

This network of fully automated transmitters is causing historically high levels of interference to all other radio amateur activities on HF because the automated side of an email gateway is incapable of "listening first", or frequency sharing, as radio amateurs usually do, and are required to do, by FCC regulation, and the other side, which is supposed to be manned by a "listen first" live operator, finds it unnecessary to be concerned about anyone else already on the frequency, because the protocol chosen by Winlink is capable of overpowering anyone else on the frequency, and keep hammering away using ARQ until the Winlink station dominates the frequency. The result is high levels of QRM to all others. In addition, the manual stations escape detection by never transmitting any callsign identification in case they try to connect with an automatic station and fail. They just create QRM and leave the air unidentified.

The majority recommendation proposes to expand this network of robot transmitters to completely cover the ham bands without restriction, including the phone band segments, with the exception of the CW and beacon regions. This chart of the current coverage of the 20 meter band by Winlink robots indicates why interference by these robots is so high and currently disrupts so many other traditional radio amateur communications. It is easy to visualize what it would be like if the majority recommendation were accepted by the Board. There would be no space left on the HF bands for ragchewing, DX chasing, award-chasing, contesting, or experimentation, free from constant interference from Winlink robot stations, such as suffered currently by digital operators, and more recently, on 30 meters by CW operators, who find it more and more difficult to operate without a Pactor station suddenly coming on the frequency in the middle of their QSO.



The hfdigital committee was chartered to address only Minute 63 as follows:

"Minute 63. On motion of Mr. Frenaye, seconded by Mr. Bodson, it was unanimously VOTED that the President is authorized to appoint an ad hoc working group to study the new HF data modes in the Amateur Radio Service. The Terms of Reference are to develop recommendations for introduction of voice-bandwidth data modes and to advise the board on amateur-Internet linking and HF automatic control with a final report to the Board by January 2003."

Recommendation for Introduction of Voice-bandwidth data modes

It is recommended that voice-bandwidth data modes NOT be introduced, nor even allowed, on the crowded HF bands, except for digital voice applications, because they unnecessarily use a disproportionate amount of bandwidth for the benefit of a faster data rate.

Data provided by the Winlink Administrator for the first three weeks of 2003, in which 13,182 emails were transferred by Pactor 1 and Pactor 2, which are 500 Hz wide, compared to 5,745 emails transferred by Pactor 3, which is 2400 hz wide, or "voice-bandwidth", shows only a 30% savings in time using a voice-bandwidth data mode, which uses 500% more space, than using a 500 Hz-wide data mode, such as Pactor 2:

 Pactor 1 or Pactor 2

 3631.9        2.9
 7076.9        2.7
 10123.9       1.9
 14076.9       1.7
 18103.9       1.7
 ______   ___
 Total   10.9

 Avg:   2.18 minutes/message for Pactor 1/2 over 13,182 messages

 Pactor 3

 7103.7        1.8
 10141.2       2.5
 14106.7       1.3
 18108.7       1.2
 _______  ___
 Total   6.8

 Avg:   1.7 minutes/message for Pactor 3 over 5,745 messages

Therefore, voice-bandwidth data modes are an inefficient use of the spectrum for average-length email transfers, such as those handled daily by Winlink.

The only other known use for voice-bandwidth data modes is for image transfers, which is understandable, as images are known to utilize large amounts of data. The only currently known voice-bandwidth image data mode is HDSSTV, which can send an SSTV-size picture, with a very low error rate, in 30 seconds, using a bandwidth of 2400 Hz. However, the HDSSTV software author has confirmed that the same image, at the same low error rate, can be sent in less than 2 minutes, using a bandwidth under 500 Hz.

In other words, just by accepting to wait longer for an image to arrive, as is the current practice in SSTV, it is not necessary to use a voice-bandwidth transmission mode.

Since space is at a such premium on the crowded HF bands, it is irresponsible to use voice-bandwidth data modes which serve merely to reduce waiting time, either for image transfer, or email transfer, because it deprives other users of other modes of space in which to operate. The VHF or UHF bands are a more appropriate place to use voice-bandwidth data modes, as is ATV, which requires so much bandwidth it is only allowed on the UHF bands, where space is not at such a premium.

Recommendation: Voice-bandwidth data modes SHOULD NOT be introduced on the HF bands.

Amateur-Internet Linking

Recommendation: The current FCC regulations regarding amateur-internet linking are adequate, and no changes are necessary. The dissenting recommendation is in agreement with the majority recommendation on this issue.

HF Automatic Control

HF automatic control needs to be confined to a single, contiguous, space on the bands where the automatic robot stations can be avoided by others and where they do not cause interference to others.

If it is agreed that voice-bandwidth data modes should not be used because they are an inefficient use of bandwidth for email transfers, then the current FCC-allocated sub-bands for automatically controlled digital stations are adequate for the current level of automated email gateway operations, if networks, such as Winlink, make any serious attempt at all in efficiently utilizing frequencies on a first-come-first-serve basis.

Winlink professes to be the largest email gateway network in the world, and it numbers only 4,500 users out of 660,000 FCC licensed radio amateurs, or 0.7%. Any HF spectrum space reserved for activities such as WINLINK should be proportional to their population size as compared to the total amateur population desiring to use the same spectrum for other purposes. The current FCC sub-bands for automatically controlled digital stations on 20 meters is currently accepted by all IARU regions, and provides even more space than the proportion of automated network users on HF would ordinarily be entitled to.

Recommendation: Stations using Automatic Control on the HF bands should be confined to the current FCC sub-bands for automatically controlled digital stations and FCC regulation 97.221(c) should be repealed to prevent the robot stations from spreading randomly all over the bands where they can create interference to others trying to use the bands for traditional ham radio activites.


It is highly improper to exceed the charter of the committee, so this dissenting recommendation provides no bandplan recommendation.

However, the majority recommendation does include a bandplan as Appendix A, which seriously misrepresents the Region 1 proposed bandplan, presented to the committee by Dave Sumner, K1ZZ, by conveniently omitting remarks and limitations which confine store-and-forward operations, such as Winlink, to the FCC-defined sub-band for automatically controlled digital stations on the 20m band and completely bans those operations on the 160m, 40m, and 30m bands, and then declares that the majority recommendation "harmonizes where possible" with the latest proposed Region 1 bandplan.

It is clear that the majority bandplan submission is really a veiled attempt by Winlink to gain ARRL support for Winlink robot transmitters operating all over the HF bands and therefore should be disregarded.


Submission of such a bandplan is clearly outside the charter of the hfdigital committee, and should be left up to a new committee, officially chartered to address Minute 64, which must consist of members representing the entire range of operating interests - CW, data, and phone - and not dictated by a committee already unbalanced in favor of a special interest group.

 The ARRL Board should be promoting bandplans that encourage the development of more spectrum-efficient modes, such as PSK31 and MFSK16, which accomplish the essential task of communications in less bandwidth, as opposed to promoting wider data modes which add to the interference and congestion on HF bands instead of reducing it.

Howard Teller
March 23, 2003

Zerobeat - Zerobeat Discussion Forum - ARRL Band Plan Proposal - ARRL Divisions-Contact YOUR director

If the ARRL has its way, you'll hear a lot of this in the bands!