Extreme Grid QF09 100 km North of Broken Hill By Rex VK7MO and Dave VK3HZ A relatively easy ISCAT-B QSO over 753 km so the opportunity was taken to do an ISCAT-A test to see if we could get a better understanding of why ISCAT-A sometimes fails to decode correctly. As it happened (or as Murphy would dictate) ISCAT-A worked perfectly on this occasion. But we are starting to have some ideas on why ISCAT-A sometimes gives errors and this issue is discussed in more detail at the end of this report. Our best guess is that it is related to short-term Doppler variations perhaps caused by turbulence buffeting the aircraft. Given the issue of decoding errors on ISCAT-A we recommend that ISCAT-B be preferred for 10 GHz aircraft scatter and run in 15 second periods. Location VK7MO was located just into QF09 at 30.99978 South and 141.18733 East. The land was essentially flat but at least no nearby trees (Fig1) so take-off was down to 0.2 degrees. VK3HZ was located at Hesket which brings him within his 50 km radius to qualify for a grid (#78). Fig 1: VK7MO s Operating Location at QF09
Flight Paths and Common Coverage Fig 2: The path of propagation is in Orange. The common coverage area is defined by the Heywhatsthat 2D limits in purple. The aircraft paths are in Blue with the Red aircraft paths showing those that were being flown. VOZ1447 was the first aircraft detected and is just going out of ADSB range from VK7MO as indicated by turning Red -- this is slightly beyond the 2D limit showing the radio refractive index is a help. VOZ1394 was used for the second ISCAT-B detection and JST773 for the ISCAT-A test. RESULTS Good data highlighted in Green and DFs for good data highlighted in Yellow. ALL.TXT Files at VK7MO ISCAT-B UTC Date: 2014 Aug 06 --------------------- 234059 Transmitting: ISCAT- VK3HZ VK7MO
First Aircraft 235815 0-20 10.4-108 0 VK7MO VK3HZ -7 15 3 9 2.2 235835 Transmitting: ISCAT- VK3HZ VK7MO -20 235845 0-20 2.6 65 0 9QZ 3 0 1 1.1 235900 Transmitting: ISCAT- R-20.. Later Aircraft 001715 4-5 5.9 22 0 * RRR 4 3 9 4.5 001730 Transmitting: ISCAT- 73 001745 0-20 6.5-194 0 73 3 5 6 2.2 001800 Transmitting: ISCAT- VK3HZ VK7MO 001801 Transmitting: ISCAT- VK3HZ VK7MO -5 001815 0-20 9.3 345 0.WMMNK4E15C 11 0 4 1.1 001830 Transmitting: ISCAT- ISCAT - A Changed to ISCAT-A 002000 Transmitting: ISCAT- VK3HZ VK7MO 002015 0-20 9.8-151 0 L55 3 1 1 4.5 002045 0-20 3.1-205 0 W/KI.NAI 8 0 2 2.2 002115 2-10 13.2 11 0 * VK7MO VK3HZ 12 2 10 4.5 002134 Transmitting: ISCAT- VK3HZ VK7MO -10 010015 0-20 15.4 86 0 P8L0.VP3 8 0 2 2.2 005947 0-20 0.0 0 0 0 0 0 0.0 010045 0-20 14.3 528 0 X4O 3 1 3 2.2 010115 0-20 9.8-97 0 FQX 3 0 1 2.2 010145 1-11 13.2 75 0 * VK7MO VK3HZ 12 6 10 4.5 010215 0-20 8.7-108 0 VK7MO VK3HZ 12 5 10 4.5 010245 0-20 0.0 0 0 0 0 0 0.0 All aircraft were crossing roughly at right angles and it is seen that the Doppler varied over a similar range on both ISCAT-A and B It is seen that both ISCAT-A and ISCAT-B decoded without errors. ALL.TXT Files at VK3HZ UTC Date: 2014 Aug 06 ISCAT-B 235245 Transmitting: ISCAT- VK7MO VK3HZ 235800 0-20 4.8 194 0 VK3HZ VK7MO 12 10 10 1.1 235816 Transmitting: ISCAT- VK7MO VK3HZ -7 235830 2-6 3.7-194 0 * VK3HZ VK7MO 12 9 10 2.2 Second aircraft 01700 0-20 7.0 129 0 R-20 5 10 10 1.1 001716 Transmitting: ISCAT- RRR 001730 1-5 3.1-43 0 * 73 3 10 10 1.1 001745 Transmitting: ISCAT- 73 001800 0-20 14.3 108 0 H5Z4O8KA7QVX6XQBL 17 0 2 2.2 Changed to ISCAT-A
001815 Transmitting: ISCAT- VK7MO VK3HZ 002130 3-7 3.1-32 0 * VK3HZ VK7MO 12 10 10 2.2 002155 Transmitting: ISCAT- VK7MO VK3HZ -7 002200 0-20 0.0 0 0 0 0 0 0.0 002215 Transmitting: ISCAT- VK7MO VK3HZ Another aircraft 010200 2-8 3.1 43 0 * VK3HZ VK7MO -10 16 10 10 2.2 010230 2-11 4.2-183 0 * VK3HZ VK7MO -10 16 10 10 4.5 010245 Transmitting: ISCAT- R-7 Note: The maximum variations in Doppler that produced good decodes were +194 Hz to -194 Hz on ISCAT-B and +129 Hz to -183 Hz on ISCAT-A. So on this small data set there is no evidence that we are losing signal due to Doppler going outside the passband on ISCAT-B. This differs from our findings in shorter distance tests and perhaps indicates that at long range the weaker side scatter of the aircraft scatter pattern in the limiting factor. ISCAT-A or ISCAT-B for 10 GHz Aircraft Scatter ISCAT-B uses a bandwidth of 1809 Hz (most of an SSB passband typically from 300 to 2300 Hz) sends at 32.3 characters/second while ISCAT-A uses half the bandwidth (907 Hz) but sends at half the rate (16.5 cps). When an aircraft crosses the path at near right angles the Doppler can be up to +/- 500 Hz and thus the broader bandwidth ISCAT-B takes the signal out of a narrow SSB passband. To overcome this Joe K1JT developed the narrower bandwidth ISCAT-A. ISCAT-A is about 1 db more sensitive but needs bursts of twice as long to decode. While we have in the past had good results with ISCAT-A we have noticed that it tends to be less reliable on signals that we expect should decode correctly. On the first of our long range attempts at QF49 which involved aircraft crossing at about 45 degrees we first used ISCAT-A and both stations received a number of bursts of relatively strong signal that failed to decode correctly as per examples below: 011300 2-9 15.4-22 0 * VK8LP-VK3HY 12 0 7 2.2 011330 1-15 10.9-108 0 * UK3HZ VK7MO 12 1 7 8.9 013800 1-12 10.9-151 0 *?VK2H VL6MO 12 0 4 4.5 014500 1-13 10.9-97 0 * VK3HY VK6MN 12 1 8 4.5 The message in this case should have been VK3HZ VK7MO and as can be seen there is often an error one character off indicating the decode is out by one tone. Sometime later we moved to ISCAT-B and it decoded correctly as below: 030530 2-7 4.2-65 0 * VK3HZ VK7MO 12 6 10 4.5 030600 0-20 7.6-194 0 R-15 5 8 10 4.5 033100 0-20 5.9 65 0 R-15 5 4 6 2.2 033130 0-20 9.8-65 0 73 3 6 6 8.9 033200 0-20 3.7-151 0 73 3 5 6 4.5
Interestingly the current tests at QF09 showed no evidence of this issue with both ISCAT-A and ISCAT-B giving accurate decodes. ISCAT establishes the frequency reference and sync by using a Costas array. Such arrays were developed in the 1960 s to improve detection in Sonar systems. They are in effect a combination of tones in both frequency and time to produce a sync and frequency reference compared to modes such as JT65 which uses time with a single frequency. An advantage of using the Costas array for sync is that it can be sent very rapidly and so correct for short-term changes in Doppler. But there is a limit. In the case of ISCAT-B the tones are spaced 44 Hz apart and the Costas array sent every 1.1 seconds so the Doppler has to be of the order of 22 Hz per second to cause a problem. However in the case of ISCAT-B the tones are spaced 22 Hz apart and Costas array sent only every 2.2 seconds thus it could cope with no more than about 6 Hz per second Doppler change ie ISCAT-A is around 4 times more susceptible to short term Doppler errors. On tests between VK7MO and VK3GHZ ISCAT-A has normally been used without problems. However we have gathered some data on the short-term Doppler shift by using 1270 Hz single tones as shown in Fig XX below. If such short term Doppler variations were occurring then they could well explain the decoding errors with ISCAT-A in the QF49 tests referred to above. At this stage we can only speculate on the reasons for these short-term Doppler changes. Perhaps the aircraft is going though turbulence and the height drops rapidly causing these short term Doppler fluctuations. If so this might explain why the ISCAT-A issue occurs on only some occasions. (Note: In the QF49 example while there were 6 files that showed the decoding errors all these were from the same aircraft.) Fig #: Data from VK7MO to VK3GHZ over a 570 km path. Shows short term Doppler variations (the kink) of up to 20 Hz over a few seconds. CONCLUSION Our preliminary conclusion is that it is safer to use ISCAT-B as it is less sensitive to short-term Doppler variations and does not appear to limit the decodes due to wide Doppler variation. Thus ISCAT-B is to be preferred for 10 GHz aircraft scatter.