#Q65 Latest Tests #Q65


Hasan Schiers N0AN
 

WSJT-X 2.5.0 RC3 Test Results (Q65-30A 50.275) 

(Data below represents a snapshot of a long test run from this morning. Conditions were normal to below normal based on our prior months of testing.)

Decode Settings: Fast, Enable Averaging, AutoClear Average after Decode

-----------------------------------------------
KB7IJ > N0AN: Path Length 667 Miles 
KB7IJ Power Output 0.5 Watts to 7 EL LFA @ 50' 1/2" Hardline: Icom 7300
N0AN Power 10.0 Watts to 5 EL LFA @ 55' 1/2" Hardline (GAsFET Preamp in Shack) > TS-590SG

KB7IJ has a high noise environment receiving through Dallas, and some local Solar Panels/
All testing done from 5:30 a.m. thru  approximately 9:00 a.m.

KB7IJ ran with me for approximately 2 hours at 0.5 watts output  on 50.275 2000 Hz 1120z to 1315 z


112000 -17 0.2 2001 : N0AN KB7IJ RRR  q32

112200 -16 0.2 2001 : N0AN KB7IJ RRR  q3

112400 -19 0.2 1998 : N0AN KB7IJ RRR  q32

112500 -17 0.2 2001 : N0AN KB7IJ RRR  q3

112700 -17 0.2 2001 : N0AN KB7IJ RRR  q3

112900 -23 0.2 2001 : N0AN KB7IJ RRR  q32

113100 -16 0.2 2001 : N0AN KB7IJ RRR  q32

113200 -23 0.2 2001 : N0AN KB7IJ RRR  q3

113300 -24 0.1 1998 : N0AN KB7IJ RRR  q3

113500 -22 0.2 1998 : N0AN KB7IJ RRR  q32

113600 -21 0.2 2001 : N0AN KB7IJ RRR  q3

113700 -24 0.2 2001 : N0AN KB7IJ RRR  q3

113800 -25 0.2 2001 : N0AN KB7IJ RRR  q3

114000 -26 0.2 1998 : N0AN KB7IJ RRR  q32

114200 -22 0.2 1995 : N0AN KB7IJ RRR  q32

114400 -23 0.2 1998 : N0AN KB7IJ RRR  q3

114600 -21 0.2 2001 : N0AN KB7IJ RRR  q32

114800 -23 0.1 2001 : N0AN KB7IJ RRR  q32

115000 -24 0.1 2001 : N0AN KB7IJ RRR  q32

115200 -18 0.2 2001 : N0AN KB7IJ RRR  q3

115400 -21 0.2 2001 : N0AN KB7IJ RRR  q3

115600 -25 0.2 1998 : N0AN KB7IJ RRR  q32

115800 -27 0.2 2001 : N0AN KB7IJ RRR  q32

120000 -18 0.2 2001 : N0AN KB7IJ RRR  q3

120400 -16 0.2 1998 : N0AN KB7IJ RRR  q3

120800 -22 0.2 2001 : N0AN KB7IJ RRR  q3

121100 -17 0.2 2001 : N0AN KB7IJ RRR  q3

121300 -11 0.2 2001 : N0AN KB7IJ RRR  q3

121400 -19 0.2 1998 : N0AN KB7IJ RRR  q3

121500 -20 0.2 1998 : N0AN KB7IJ RRR  q3

121600 -10 0.2 2001 : N0AN KB7IJ RRR  q3

121700 -25 0.2 2001 : N0AN KB7IJ RRR  q3

122000 -28 0.2 1998 : N0AN KB7IJ RRR  q33

122400 -28 0.2 1998 : N0AN KB7IJ RRR  q34

122900 -27 0.2 1998 : N0AN KB7IJ RRR  q35

123100 -26 0.2 2001 : N0AN KB7IJ RRR  q32

123400 -21 0.2 1998 : N0AN KB7IJ RRR  q3

123500 -21 0.2 1998 : N0AN KB7IJ RRR  q3

123800 -29 0.2 2001 : N0AN KB7IJ RRR  q33

124000 -23 0.2 2001 : N0AN KB7IJ RRR  q32

124100 -23 0.2 2004 : N0AN KB7IJ RRR  q3

124400 -32 0.2 2001 : N0AN KB7IJ RRR  q33

124800 -32 0.2 2001 : N0AN KB7IJ RRR  q34

124900 -23 0.2 1994 : N0AN KB7IJ RRR  q3

125000 -21 0.2 2001 : N0AN KB7IJ RRR  q3

125100 -12 0.2 1510 : N0AN K5GZR EL29  q0

125200 -23 0.2 2001 : N0AN KB7IJ +11  q32 (Note bogus decode report on my end)

125700 -32 0.2 2001 : N0AN KB7IJ RRR  q35

125800 -15 0.2 2001 : N0AN KB7IJ RRR  q3

130300 -28 0.2 2004 : N0AN KB7IJ RRR  q35

130500 -23 0.2 2001 : N0AN KB7IJ RRR  q3

130600 -25 0.2 2001 : N0AN KB7IJ RRR  q3

130800 -25 0.2 2001 : N0AN KB7IJ RRR  q32

131100 -29 0.2 1998 : N0AN KB7IJ RRR  q33

131200 -10 0.2 1998 : N0AN KB7IJ RRR  q3

131500 -28 0.2 2001 : N0AN KB7IJ RRR  q33



We have  been testing Q65 for months every morning. We have evaluated the following modes for terrestrial work: Ranked from worst to best decode performance,  at lowest possible tx power.

Q65-15A : RC3 has worked much better than earlier releases. Surprisingly effective. We had no problem finishing a qso every morning we attempted one, and ran for 30 minutes to verify band conditions.

Q65-30A : Clearly superior to 15A

Q65-30B : Better than 30A, but not night and day.

Q65-120E: Dramatically better on terrestrial paths than all others tested, including 120A, but hard on equipment (transmit heating)

Overall 30A is an excellent choice. Random CQ's on 50.275 do work, but nothing like FT8, of course. We are seeing (or were) increasing activity on 275 30A. ....until FT8 started opening to DX on 6m. Everyone disappeared (and rightly so...I worked 67 European stations one day and 8 JA's on another)  :-)
Operator density is the key to enjoying random qsos.

========================================================
Further testing with K5GZR 864 miles: Q65 30A  RC3
M2 6M5X @ 25' with 100' of Davis Bury Flex : 
N0AN 70 w  (he has high noise)

131900  -8  0.2 2008 :  N0AN K5GZR EL29    q3 100 watts out
132000 -16  0.2 2008 :  N0AN K5GZR EL29    q3
132100  -8  0.2 2008 :  N0AN K5GZR EL29    q3
132200  -6  0.2 2008 :  N0AN K5GZR R-20    q3
132300 -14  0.2 2008 :  N0AN K5GZR R-20    q3
132400 -19  0.2 2008 :  N0AN K5GZR R-20    q3 Now 25 watts out
132600 -25  0.2 2008 :  N0AN K5GZR R-20    q32
132800 -28  0.2 2001 :  N0AN K5GZR R-20    q3
133000  -9  0.2 2008 :  N0AN K5GZR R-20    q3 Now 15 watts out
133500 -32  0.1 2005 :  N0AN K5GZR R-20    q35
133800 -19  0.2 2002 :  N0AN K5GZR R-20    q3
133900 -11  0.2 2002 :  N0AN K5GZR R-20    q3
134000 -19  0.2 2002 :  N0AN K5GZR R-20    q3
134100 -12  0.2 2002 :  N0AN K5GZR R-20    q3
134300 -23  0.2 1999 :  N0AN K5GZR R-20    q32
134400 -20  0.1 2002 :  N0AN K5GZR R-20    q3
134600  -3  0.2 2002 :  N0AN K5GZR R-20    q3
134700 -17  0.1 2002 :  N0AN K5GZR R-20    q3
134800  -7  0.2 2002 :  N0AN K5GZR R-20    q3 Power Returned to 100 watts out
134900  -7  0.2 2002 :  N0AN K5GZR R-20    q3
135000 -13  0.2 2002 :  N0AN K5GZR R-20    q3

Link to *.jpg of Testing Group and Wattmeter showing Power Output at KB7IJ



Thanks to all the following for the extensive tests:
KB7IJ, Rich
K5GZR, Rick (Geezer)
ND0B, Bill
WB4HIE, James
NM3G, Rick

The results reported here have been consistent with all our prior testing  (hundreds of hours over several months), with perhaps one discernable change: Q65-15A works much better now than in the early versions. We have no idea why.

Summary:

When a contact absolutely must be made, Q65 is the mode to use. It makes use of every propagation mode that may be present, and is perhaps unique in this respect.

Q65 makes use of :
Ionoscatter 
Troposcatter
Meteor scatter
Tropospheric Ducting
Sporadic E
F2
Airplane Scatter
Ground Wave
EME
Line-of-Sight

The propagation mode doesn't seem to matter. If there is a signal present, no matter what propagation mode exists at that instant, Q65 will take advantage of it. 

(Do NOT take this out of context, it is not a replacement for MSK144 when there is a plethora of meteors available)

We have observed 2 or 3  propagation modes being used at the same time, based on the signature of the signals on an SDR graph, showing amplitude vs. time, over repeated rx sequences. Each propagation mode, looks different on the signal history graph. Classic trapezoid for some meteor scatter, random low level peaks for Ionoscatter, etc., etc.

Pros and Cons vs. Other Modes:

1. Q65 does not appear to be well suited to a busy passband with overlapping signals,  which FT8 excels at. Q65 sensitivity is superior to FT8, FT4, MSK144 etc. 

2.  FT8, while permitting a highly shared passband, does not handle strong signal excursions well at all. (meteor pings destroy an FT8 decode)

3. The way MSK144 is currently used,  (Time Division Multiplexing),  the frequency cannot be shared (as a passband). If someone is in local or Es is in, MSK144 requires a QSY or changing sequences.

4. Q65-30A, for example, is about 200 Hz wide, so if the available passband is 300 to 2800 Hz, quite a few qsos can take place that do not overlap. Some overlap appears to be tolerated, but not a lot and not with large signal differences.

No mode, not even Q65, can work when none of the above propagation modes exist. It may appear to be magical, but it's not.  :-)

Our experience indicates that with patience and persistence, Q65 in its various configurations will outperform nearly all the other WSJT-X modes on 6 meters and requires very modest station setups to successfully complete digital contacts of extraordinary distance at very low power levels.

The typical 100w output setup with a small yagi, can easily complete 1000+ miles contacts every morning, if one's local noise is not excessive.

73, N0AN
Hasan