Spotted on wrong gridsquare? #IssueReport #WSPR


Steve marsh (M0NMA)
 

I am helping G8IOA who has beacons running on 10,6,4,2,and 70cms all running about 2W using antennas mounted about 3m AGL. The hope is that these will help identify propagation windows that open from time to time, particularly on the higher bands. See GB3SA.UK for our DX records

Anyway, G8IOA is occasionally spotted as transmitting from JO01 rather than IO92SP. This has happened many times and from different stations - some running WSJT-X and others using other software. I know that there can be decoding errors, but seems unlikely a the error is always JO01, so I presume that the programs must be picking up errors from another database somewhere? Does anyone know why this is happening? A list of the 'wrong spots' from the last 5 weeks are listed below (using WSPR.Rocks, in case you are not familiar with it - a great WSPR tool). There were 8505 spots in total, so only a small proportion were wrong. They are all reporting the wrong frequency as well, as G8IOA doesn't transmit on 10Mhz.

Local   (y-m-d) TX txGrid RX rxGrid MHz W SNR drift km mode km/W spotQ version
2021-08-17 09:48 G8IOA JO01 PA3EGH JO21qx 10.140165 5 -25 0 309 2 62 19 2.4.0
2021-08-17 09:48 G8IOA JO01 G8ORM IO93cu 10.140101 5 -21 0 323 2 65 27 2.1.2
2021-08-17 09:48 G8IOA JO01 G8ORM IO93cu 10.140169 5 -18 0 323 2 65 33 2.1.2
2021-08-17 09:28 G8IOA JO01 G8ORM IO93cu 10.140099 5 -24 0 323 2 65 22 2.1.2
2021-08-17 09:28 G8IOA JO01 G8ORM IO93cu 10.140236 5 -22 1 323 2 65 25 2.1.2
2021-08-16 17:28 G8IOA JO01 G8ORM IO93cu 10.140101 5 -21 0 323 2 65 27 2.1.2
2021-08-16 13:48 G8IOA JO01 PD0OHW JO33lc 10.140238 5 -20 0 446 2 89 40 2.0.0
2021-08-14 20:28 G8IOA JO01 DL2YED JO31op 10.140241 5 -14 0 432 2 86 53 2.5.0-rc5
2021-08-14 09:48 G8IOA JO01 DK8JP JO31gk 10.140242 5 -25 0 387 2 77 24 2.1.0
2021-08-11 13:28 G8IOA JO01 PD0OHW JO33lc 10.140238 5 -19 0 446 2 89 42 2.0.0
2021-08-11 11:28 G8IOA JO01 DF2JP JO31fp 10.140236 5 -23 0 381 2 76 27 2.5.0-rc5
2021-08-11 11:08 G8IOA JO01 ON3URE JO20cw 10.140237 5 -25 0 234 2 47 14 v1.1.0-dev
2021-08-11 08:08 G8IOA JO01 DK8JP JO31gk 10.140242 5 -29 0 387 2 77 15 2.1.0
2021-08-08 13:08 G8IOA JO01 LA1ZM JO38ig 10.140235 5 -24 0 839 2 168 56 2.3.0
2021-08-07 09:48 G8IOA JO01 ON3URE JO20cw 10.140237 5 -22 0 234 2 47 18 v1.1.0-dev
2021-08-07 09:48 G8IOA JO01 ON7KB JO21ei 10.140119 5 -22 0 237 2 47 18 2.0.0
2021-08-07 09:28 G8IOA JO01 DF2JP JO31fp 10.140101 5 -18 0 381 2 76 38 2.5.0-rc5
2021-08-07 09:28 G8IOA JO01 DF2JP JO31fp 10.14017 5 -19 0 381 2 76 36 2.5.0-rc5
2021-08-07 09:08 G8IOA JO01 ON7KB JO21ei 10.140117 5 -21 0 237 2 47 20 2.0.0
2021-08-07 08:08 G8IOA JO01 DK8JP JO31gk 10.140243 5 -24 0 387 2 77 26 2.1.0
2021-08-05 23:08 G8IOA JO01 OZ7IT JO65df 10.140237 5 -22 -1 859 2 172 67 2.1.0
2021-08-05 14:48 G8IOA JO01 HB9TMC JN46lj 10.140152 5 -20 0 812 2 162 72  
2021-08-05 14:28 G8IOA JO01 HB9TMC JN46lj 10.140152 5 -19 0 812 2 162 77  
2021-08-04 13:48 G8IOA JO01 DL2YED JO31op 10.140101 5 -20 0 432 2 86 38 2.5.0-rc4
2021-08-04 13:48 G8IOA JO01 DL2YED JO31op 10.140239 5 -21 0 432 2 86 36 2.5.0-rc4
2021-08-04 13:08 G8IOA JO01 DL2YED JO31op 10.140101 5 -17 0 432 2 86 45 2.5.0-rc4
2021-08-04 13:08 G8IOA JO01 DL2YED JO31op 10.14024 5 -16 0 432 2 86 48 2.5.0-rc4
2021-08-03 21:28 G8IOA JO01 DK8JP JO31gk 10.140244 5 -29 0 387 2 77 15 2.1.0
2021-08-03 14:28 G8IOA JO01 HB9TMC JN46lj 10.140171 5 -20 0 812 2 162 72  
2021-08-02 16:48 G8IOA JO01 DL2YED JO31op 10.1401 5 -18 0 432 2 86 43 2.5.0-rc3
2021-08-02 16:48 G8IOA JO01 ON3URE JO20cw 10.140098 5 -25 0 234 2 47 14 v1.1.0-dev
2021-08-02 16:48 G8IOA JO01 ON3URE JO20cw 10.140237 5 -16 0 234 2 47 26 v1.1.0-dev
2021-08-02 02:48 G8IOA JO01 DL2YED JO31op 10.140101 5 -27 0 432 2 86 22 2.5.0-rc3
2021-07-30 16:48 G8IOA JO01 M0UNI IO82rw 10.140094 5 -21 -1 288 2 58 24 2.4.0
2021-07-30 16:48 G8IOA JO01 M0UNI IO82rw 10.140165 5 -16 0 288 2 58 32 2.4.0
2021-07-25 20:28 G8IOA JO01 M0UNI IO82rw 10.140096 5 -23 0 288 2 58 21 2.4.0
2021-07-24 00:08 G8IOA JO01 ON7KB JO21ei 10.14019 5 -20 0 237 2 47 21 2.0.0
2021-07-24 00:08 G8IOA JO01 ON7KB JO21ei 10.140257 5 -19 1 237 2 47 22 2.0.0
2021-07-19 15:48 G8IOA JO01 ON3URE JO20cw 10.140236 5 -25 0 234 2 47 14 v1.0.0
2021-07-19 12:48 G8IOA JO01 G8ORM IO93cu 10.1401 5 -25 -1 323 2 65 20 2.1.2
2021-07-18 13:28 G8IOA JO01 ON3URE JO20cw 10.140099 5 -26 0 234 2 47 13 v1.0.0
2021-07-18 12:48 G8IOA JO01 G8ORM IO93cu 10.1401 5 -18 0 323 2 65 33 2.1.2
2021-07-18 12:48 G8IOA JO01 G8ORM IO93cu 10.140168 5 -20 0 323 2 65 29 2.1.2
2021-07-18 09:48 G8IOA JO01 DL2YED JO31op 10.140104 5 -18 0 432 2 86 43 2.5.0-rc3
2021-07-18 09:48 G8IOA JO01 DL2YED JO31op 10.140169 5 -17 0 432 2 86 45 2.5.0-rc3
2021-07-18 08:28 G8IOA JO01 G8ORM IO93cu 10.140101 5 -22 -1 323 2 65 25 2.1.2
2021-07-18 07:28 G8IOA JO01 G8ORM IO93cu 10.140102 5 -25 -1 323 2 65 20 2.1.2
2021-07-17 21:08 G8IOA JO01 DK8FT JN58oe 10.140097 5 -25 -1 820 2 164 50 2.3.0
2021-07-17 21:08 G8IOA JO01 DK8FT JN58oe 10.140094 5 -26 -2 820 2 164 46 2.2.0
2021-07-17 18:28 G8IOA JO01 G8ORM IO93cu 10.140098 5 -22 0 323 2 65 25 2.1.2
2021-07-17 18:08 G8IOA JO01 G8ORM IO93cu 10.140101 5 -21 -2 323 2 65 27 2.1.2
2021-07-17 17:28 G8IOA JO01 G8ORM IO93cu 10.140235 5 -22 0 323 2 65 25 2.1.2
2021-07-16 13:08 G8IOA JO01 HB9TMC JN46lj 10.140148 5 -20 0 812 2 162 72  
2021-07-16 12:08 G8IOA JO01 HB9TMC JN46lj 10.140152 5 -18 0 812 2 162 81  
2021-07-14 14:08 G8IOA JO01 ON3URE JO20cw 10.14011 5 -24 0 234 2 47 16 v1.0.0

73
--
Steve M0NMA


Bill Somerville
 

On 17/08/2021 10:10, Steve marsh (M0NMA) wrote:
I am helping G8IOA who has beacons running on 10,6,4,2,and 70cms all running about 2W using antennas mounted about 3m AGL. The hope is that these will help identify propagation windows that open from time to time, particularly on the higher bands. See GB3SA.UK for our DX records

Anyway, G8IOA is occasionally spotted as transmitting from JO01 rather than IO92SP. This has happened many times and from different stations - some running WSJT-X and others using other software. I know that there can be decoding errors, but seems unlikely a the error is always JO01, so I presume that the programs must be picking up errors from another database somewhere? Does anyone know why this is happening? A list of the 'wrong spots' from the last 5 weeks are listed below (using WSPR.Rocks, in case you are not familiar with it - a great WSPR tool). There were 8505 spots in total, so only a small proportion were wrong. They are all reporting the wrong frequency as well, as G8IOA doesn't transmit on 10Mhz.

Hi Steve,

do you have evidence that the stations spotting him as sending from JO01 actually received a message with that grid square? Perhaps check with one of the spotters and ask for an extract of their ALL_WSPR.TXT for the relevant time. I ask because the first thing is to check if the online spot database has substituted that grid for some reason.

73
Bill
G4WJS.


Dave Garber
 

I noticed that all wrong grids were spotted on 30m.   could that wspr setting be incorrectly sending jo01 ?  worth double checking

Dave Garber
VE3WEJ / VE3IE


On Tue, Aug 17, 2021 at 6:36 AM Bill Somerville <g4wjs@...> wrote:
On 17/08/2021 10:10, Steve marsh (M0NMA) wrote:
I am helping G8IOA who has beacons running on 10,6,4,2,and 70cms all running about 2W using antennas mounted about 3m AGL. The hope is that these will help identify propagation windows that open from time to time, particularly on the higher bands. See GB3SA.UK for our DX records

Anyway, G8IOA is occasionally spotted as transmitting from JO01 rather than IO92SP. This has happened many times and from different stations - some running WSJT-X and others using other software. I know that there can be decoding errors, but seems unlikely a the error is always JO01, so I presume that the programs must be picking up errors from another database somewhere? Does anyone know why this is happening? A list of the 'wrong spots' from the last 5 weeks are listed below (using WSPR.Rocks, in case you are not familiar with it - a great WSPR tool). There were 8505 spots in total, so only a small proportion were wrong. They are all reporting the wrong frequency as well, as G8IOA doesn't transmit on 10Mhz.

Hi Steve,

do you have evidence that the stations spotting him as sending from JO01 actually received a message with that grid square? Perhaps check with one of the spotters and ask for an extract of their ALL_WSPR.TXT for the relevant time. I ask because the first thing is to check if the online spot database has substituted that grid for some reason.

73
Bill
G4WJS.





Karl Beckman WA8NVW - AFA5VB
 

#Steve (M0NMA) and G8IOA
WSPR requires the station operator to input the grid locator detail that is to be transmitted,  Assuming the software is operating correctly, it can neither insert an arbitrary grid locator nor alter the 4 or 6 digit locator entered by the station operator. 
First try viewing the data from a different perspective: Sort the 'JO01' list by receiving call sign to see which receiving stations OFTEN spot the wrong grid square location. You may discover a pattern!
1)  It is possible that those stations are obtaining the wrong grid from some other database or station registry, such as a contest history file on the spotting receive station's computer itself.
2)  Also have G8IOA verify his location data (and correct if necessary) at the more popular British and European ham info forums where he is a registered user.  Here in North America some site operators will calculate and insert a 4 character grid square based on the station's postal address if the ham user fails to fill in the box with his own valid location data.
3)  Have either of you looked at the 'wrong' grid locator JO01 area to see if it includes a residence from long-ago or previously licensed location for G8IOA.  That also points to someone fabricating a locator based on the obsolete location info rather than accepting the data actually being trransmitted.
 
Karl  WA8NVW  OH
WA8NVW@...
in WSJTX@groups.io


Dave Garber
 

I understand you checks, but it still does not explain why all jo01 spots were on 30m, but the operator was never on that band....

just curious

Dave Garber
VE3WEJ / VE3IE


On Tue, Aug 17, 2021 at 5:23 PM Karl Beckman WA8NVW - AFA5VB <wa8nvw@...> wrote:
#Steve (M0NMA) and G8IOA
WSPR requires the station operator to input the grid locator detail that is to be transmitted,  Assuming the software is operating correctly, it can neither insert an arbitrary grid locator nor alter the 4 or 6 digit locator entered by the station operator. 
First try viewing the data from a different perspective: Sort the 'JO01' list by receiving call sign to see which receiving stations OFTEN spot the wrong grid square location. You may discover a pattern!
1)  It is possible that those stations are obtaining the wrong grid from some other database or station registry, such as a contest history file on the spotting receive station's computer itself.
2)  Also have G8IOA verify his location data (and correct if necessary) at the more popular British and European ham info forums where he is a registered user.  Here in North America some site operators will calculate and insert a 4 character grid square based on the station's postal address if the ham user fails to fill in the box with his own valid location data.
3)  Have either of you looked at the 'wrong' grid locator JO01 area to see if it includes a residence from long-ago or previously licensed location for G8IOA.  That also points to someone fabricating a locator based on the obsolete location info rather than accepting the data actually being trransmitted.
 
Karl  WA8NVW  OH
WA8NVW@...
in WSJTX@groups.io



Ryan Trullinger (KC0QNB)
 

I had the same issue the other day DN90 is correct where did the other grid square come from?


Bill Somerville
 

On 26/08/2021 21:00, Ryan Trullinger (KC0QNB) wrote:
I had the same issue the other day DN90 is correct where did the other grid square come from?
Hi Ryan,

possibly someone in JO22li square has a call that is a hash collision with yours. As they are sending a 6-character grid they will be sending message pairs with a hash code in the second of each pair. I see a spot of your station with a 6-character grid also, do you have the "Prefer Type 1 Messages" option unchecked or checked?

73
Bill
G4WJS.


Ryan Trullinger (KC0QNB)
 

Prefer Type 1 messages is checked, I don't think I changed it. I changed the GRID to 4 characters in JTDX from 6 Characters, that is why you saw that. Now both apps are set to 4 characters.


Bill Somerville
 

On 27/08/2021 00:37, Ryan Trullinger (KC0QNB) wrote:
Prefer Type 1 messages is checked, I don't think I changed it. I changed the GRID to 4 characters in JTDX from 6 Characters, that is why you saw that. Now both apps are set to 4 characters.
Hi Ryan,

OK, with "Prefer Type 1 Messages" checked yo will only send single messages with 4-character grid squares. That may help a little with hash collisions, but if a station with a hash collision with your callsign is sending 6-character grids then receiving stations *may* get false decodes due to the hash collisions. It should be rare but I guess there may be a particular circumstance when a first message of a pair is not received by a number of stations, that have previously copied you, that then copy only the second message of the pair, then will spot the wrong message erroneously containing your call.

I would not worry unless it continues to happen.

73
Bill
G4WJS.