Monday, April 29, 2019

FT8 Is Not a Low Power Mode

Early this year I gave a talk on 6 meter DXing with FT8. It was well enough received that I was twice asked to repeat my performance to two other clubs. If the interest shown is any indication expect to hear a few new DX seekers from FN25 and nearby grids on 6 meter FT8 this year.

I will not echo that presentation on the blog since many of the ideas therein were covered in 2018 blog articles. Instead I will dig deeper into one topic: power. How much is enough? Does it make a difference? Is it the right thing to do with FT8? With the summer Es (sporadic E) season on the doorstep and FT8 the go-to mode for 6 meters (personal preferences aside) this is a good time to review this sometimes controversial subject.

There is a widely held view that FT8 is a low power mode. I will state right up front that this is not true, or at least no more true than for other modes. Most of the time we all use more power than is necessary to complete a QSO, whether it is with CW, SSB, RTTY, FM or other modes.

We do this most often because it is a bother to fuss with power level and we don't really know how little is enough since we don't know the listening conditions at the other end of the QSO. Many hams contend with EMI in our dense urban areas in addition to QRM and QRN. Of course there are those who run QRO for no better reason than they can and want to do so.

Those are very general statements and I want to restrict this discussion to the unique propagation conditions on 6 meters. Operating FT8 on 6 meters is not the same as on HF, and we need to do it differently to increase DXing success. Es is particularly challenging and most of the time that is the propagation by which we work DX.

Es propagation for DX

Es openings tend to be spotlight openings in which, for example, your neighbour hears the DX but you do not. This is because the area of E layer ionization that supports 50 MHz refraction back to Earth is typically small and both stations need that ionosphere patch to be between them and "visible". Extend this to lengthy multi-hop path and the effect is accentuated.

The sample plot of 6 meters QSOs in a randomly selected opening is illustrative of this effect. Don't despair if you are not on one of these lines since the area and position of ionization patches that support 6 meter propagation constantly change. Like the aurora it is not the ionized patches that are moving but rather that the pattern of E layer ionization varies.

This variability is a part of the challenge of 6 meter DXing. When your neighbour has propagation, or someone midway along the wished for DX path, you need to keep monitoring since your turn may come next.

With a big enough station you don't always have to wait. It is possible to work DX when the MUF (maximum usable frequency) between you and the DX hovers below 50 MHz. This is due to forward scatter, a diffuse refraction of signals downward from the main signal path, which itself doesn't bend downward enough to be usable. Unfortunately the signal level of scattered signals is well below that of signals below the MUF. The lower the MUF the weaker the scatter.

Power and big antennas are needed to exploit forward scatter. This is an important advantage the big guns have over everyone else. With an amplifier you will have more success with forward scatter by increasing the duration of DX openings. Of course the other station must be sufficiently equipped or although they will hear the big gun they are not able to be heard. This is often my fate when calling the big guns.

Fitting FT8 into brief openings

The following fictional sequence is the shortest possible FT8 QSO with a DX station. Counting the first CQ it is 60 seconds long. It is reasonable to add the 15 seconds for the second CQ since it confirms that the other station believes the QSO is complete. Add another 15 seconds if you don't omit Message #1 (grid square). Real 6 meter DX QSOs rarely go so fast, often requiring 2 minutes or more.
  QQ8ABC VE3VN -10
This has implications since DX openings are typically marginal and brief. Frequently QSOs are not completed due to QSB. Another signal peak may occur but often doesn't, and even if it does the other station may be busy working someone else. With 6 meter DX Es there is no tomorrow: you might not get a second chance.

Power plays a role in this, or more precisely ERP (effective radiated power). This is a measure of the flux in the direction of the opening (azimuth and elevation), which is a combination of power delivered to the antenna and antenna pattern. For example, the ERP of a 100 watt transmitter through a transmission line with -3 db loss and an antenna gain of 11 dbi in the (ionospherically determined) DX path's azimuth and elevation angles would be a little over 600 watts.

Power helps the other station hear you but doesn't help you on receive. The antenna and transmission line gain and loss apply to both transmit and receive.

Now let's look at a fictitious but representative DX opening in diagram form. The horizontal axis is time, 15 seconds (FT8 interval) per vertical bar. The vertical axis is received signal strength with an arbitrary scale. The labelled lines are the noise floors for a variety of stations. The blue curve is the signal strength of the DX station.

Station A never hears the DX. This can be due to antenna, topography, noise or all these reasons. The better equipped B has just enough time to complete a 60 second FT8 QSO requiring a large dollop of good luck.

As the station capability improves you can see that there is more time to complete the QSO, and a reduced risk of decode failures and the necessity of repeated messages. We may aspire to be E but few stations are at that level. They are the 6 meter big guns.

Power and asymmetry

At right is a screen capture of my attempted QSO with 4X4DK last summer. I copied him quite well for several minutes yet he decoded me just once. Notice how the trend line of his signal strength mirrors the arc in the above diagram. Also notice the difference in reported signal strengths: it is more than 20 db.

This is a classic asymmetry. If it were due to power alone it would imply his transmit power was ~20 kW in comparison to my 100 watts. It's unlikely he was running more than 1 kW, leaving an unexplained 13 db difference. It may be due to noise at his end. WSJT-X uses his higher noise floor to calculate my signal strength. For me this situation is quite common since I live in a quiet rural setting whereas most hams are in noisy urban areas.

No matter the reason I need more ERP to be reliably heard. Either that or I wait for a better opening which may never come. I estimate that I missed about 20 DXCC countries last year, ones that were heard but not worked. Although 56 countries on 6 meter FT8 last year is pretty good, 75 would have been very welcome indeed.

"You can't work them if they can't hear you"

The immediately preceding example turns on its head the classic saying: "you can't work them if you can't hear them." More precisely said, the wording of the cliche depends on which side of the QSO you are on. To complete a QSO both stations must copy each other. However you can only improve your own station, and that means ERP must be considered.

There are a limited number of ways to improve your ERP. One or two of these should be attainable by anyone:
  • Power
  • Transmission line
  • Antenna gain and height
  • Move to a hilltop or the tropics
The last is not an option for the majority other than by portable operation or international travel. Moving is rarely feasible when DXing is the sole justification.

Single hop Es propagation can have a relatively high elevation angle. In contrast DX paths require multiple hops and perhaps ionospheric ducting, and sometimes joining the ends together with TE (trans-equatorial) propagation across the equator. These paths are most often at low elevation angles and that requires height or a hilltop.

Since going higher incurs more transmission line loss it is worthwhile to use the lowest loss coax you can acquire. Stacking yagis for gain is far easier on 6 meters than on HF. If you go this route I recommend vertical stacking for 2 or 3 yagis so that the gain comes from narrowing the vertical beam width rather than the azimuth beam width. It will make aiming less challenging and ensure your beam covers the maximum population possible.

Antenna selection in a stack can help you deal with elevation pattern nulls between all those minor lobes. The EZNEC elevation pattern above is for my A50-6 (optimized) up 24 meters. The lowest lobe is very good at below 5° but there are many nulls due to its being 4λ above ground. It is the ionosphere that determines the elevation angle, not your antenna, so if you have a null in the wrong place you'll attenuate the DX signals. Nulls in a stack are far less deep, or you can move the nulls by selecting just one of the yagis in the stack.

An amplifier is the easiest path to a bigger signal. By all means do so whether on its own or in combination with better antennas. Since you may cause AGC derived receiver desense to your ham neighbours within the FT8 window be judicious in your use of the amplifier. It is especially important to transmit a clean signal when you are QRO so adjust it for maximum linearity and have a friend check your signal for distortion products.
My preparation for Es season

I have no time for 6 meter antenna work in this year's busy schedule. A 6 meter amplifier is also low priority. My sole improvement has been to replace the transmission line. Following re-cabling work last fall the transmission lines to the outdoor switching system from both operating positions are LMR400 in the house, LDF5-50 underground and short LMR400 jumpers to the antenna switch.

The 40 meter long run of 35 year old RG213 I used the last two years has been removed and given to a friend for use on the low bands where its loss is low. A new run of LMR400 goes from the antenna switch at the base of the 21 meter tower up to the rotation loop. The small loss in the total run should be ~2 db better than before. Eventually I intend to use Heliax end-to-end. But not this year.

Every decibel counts on the marginal openings found on the longest DX paths. While I cannot guaranty a modest 2 db improvement will help it surely won't hurt. QRO would be even better, as would bigger and higher antennas. Perhaps in the future as time and interest permit.

A development I will be following this year is the emergence of FT4 since it promises shorter QSO times. This could be a good fit for 6 meter DXing. The price of lower sensitivity than FT8 may be well worth paying to best exploit signal peaks. Lower sensitivity can, in part, be solved with QRO.

No comments:

Post a Comment

All comments are moderated, and should appear within one day of submission.