Hawaii <--> Mainland Tropo

[Glenn Elmore]

As many of you may have noticed, the last week has been an active one in regards to the Hawaii to Mainland tropo path. Although this condition seems to occur yearly, sometimes more than once, we have only recently had WSPR stations on both end ready and waiting to show the advantage of the mode compared to CW, SSB and FM that have been used previously. 

The very first contact across the path was on 2m in 1957 between W6NLZ in Palos Verdes, CA and  KH6UK on Oahu. Since then the path has been spanned on VHF through microwave amateur bands. To help alert everyone to activity, CW beacons on numerous bands have been added on the slopes of Mauna Loa, on the Big Island, at about 8200' ASL.  As the duct that occurs generally is stronger and detected earlier at higher elevations, this has been very useful in alerting North American amateurs that something interesting is going on. The call associated with this site is KH6HME.
Here are some related links:

• Gary, WA6MEM, a member of this group, has a KH6HME page.
• An older Ham-Radio page.
• And an older QSL.net page.
• There is a Tropo Ducting Report page where reception reports and QSOs are discussed as openings occur.

As we have found for WSPR activity on 2m and above, the Hepburn Tropospheric Forecasts pages are very useful.

About a week ago the predictions started looking promising for an opening and this time KH6IMB in Volcan, HI at about 3700' feet was ready with 2m WSPR as were several here in California. If you have been watching the 2m data base lately you may have noticed this first spot

2016-06-30 19:10

N3IZN

144.490549

-27

0

DM13ji

100

KH6IMB

BK29jk

4061

257

Unlike some of the wrong-band reports we see in the WSPR database, this one was real and the first one across the path this year.  It was followed by many others, including two spots between Chris, N3IZN like this one

2016-07-01 06:44

KH6IMB

144.490491

-8

1

BK29jk

2

N3IZN

DM13ji

4061

60

 and one-way spots with Gary, WA6MEM, such as this

2016-07-01 03:34

WA6MEM

18.106121

-16

-1

DM03ts

50

KH6IMB

BK29jk

3966

256

In this last spot there *is* a band error, it was actually on 2m, Wil, KH6IMB's, WSPR band was out of synch with his radio for a few hours and reported the wrong band.
Although the CW beacons from KH6HME were copiable on several bands for several days here in Northern California, and at N6GN, no WSPR spots resulted even though WA6M who is on a ridge with ocean horizon was transmitting.  This time the best duct was south of us.

I asked Wil, KH6IMB, for a .wav file of one of the stronger spots to examine for any signs of spectral spreading. I'm interested in this because I have started on a WSPR/JT65/JT9 beacon project that may replace the CW beacon currently operating at KH6HME on 70cm.  More on this as it gets closer to completion. 

Here is a screen shot

I'm also attaching a .wav file that you can decode with WSPR/WSJT-X and/or listen to on an audio device.  If you listen you'll hear what a nice strong CW QSO this could have supported.  Of course, the nice thing about WSPR is that it can decode down almost a factor of 1000 lower so can provide a spot well before there's any sign of other modes.

Glenn n6gn

[Leigh VK2KRR]

Glenn,

Excellent post and report, thank you. Just so u know I've copied it on to the VHF WSPR group on Facebook for others to have a read of.

With KH6IMB there, you never know but some openings could be discovered where the Volcano beacons are not even heard. Interesting times ahead.

Leigh VK2KRR

[Bo, OZ2M]

Hi Glenn and the list

Like you I am into beacons and have been for a good number of years now. Indeed the reception of the Hawaiian beacon contains every aspect of ham radio. However, I am triggered by your comments that the CW beacon was audible for days but no WSPR spots and that the WSPR decoder is 1000 times more sensitive. Isn't the lack of WSPR spot a proof that sensitivity is not the only metric? How about robustness? WSPR has 1000 times poorer robustness compared to CW. I don't know if the number is 1000 but I am sure you get my point. In communications it is all about link probability, i.e. how often is the link open; is it 1%, 5% 10% ..., 95% or 99% etc. There is no doubt that you can get a WSPR decode via EME but not very often and in this case the S/N reading will be "low". But if you instead used e.g. JT4B-G (WSPR and JT4A are VERY much alike RF-wize) with the exact same ERP you are more likely to get a decode and with a "high" S/N.How can this be? Well the blurring of the tones is a challenge to the decoder and the wider tone spacing in the modulation helps the decoder. If the tones are blurred to much the decoder will miss out tone(s) or symbol(s) resulting in less possible decode(s) and subsequently lower S/N calculation. You can try this out by deliberately displacing the transmitted tones even with a "loud" signal and under your control. I did this, on tone 3, by mistake four years ago when implementing JT4 in the Next Generation Beacons platform. In WSJT I only saw three tones but it still decoded much to my surprise.

But by all means please give it a try. This is the way we move forward.

The whole issue of link probability is why I developed PI4 for use in a mixed mode beacon transmitting PI4 + CW + carrier. This is the best compromise between digital capabilities, the fast ID possible from a partial CW and carrier measurements. I am sure there are situations where e.g. WSPR will have a higher link probability than PI4. On the other hand there are also situations where this will be reversed e.g. aurora and rain scatter where WSPR will fall short. Besides the situations where the propagation is so short lived that even a the necessary fractional reception of the 110 s long WSPR transmission will not be long enough.

73
Bo, OZ2M

[Glenn Elmore]

Bo,

I think you don't have a clear picture of what we experienced. 

There are two different paths in question.
1) KH6HME CW beacons on 2m, 432, 1296 MHz currently operating continuously (perhaps on timers to save power). Roughly 1 kW ERP located at ~8200' ASL (2500 meters).
2) KH6IMB home station on 2m WSPR. Wil came on well after the KH6HME beacons had been spotted for some time.  Perhaps similar ERP but located at only 3700' ASL (1130 meters).

The duct on this trans-Pacific path has been watched and used for years and has been found to be a fairly complex system. In particular, height is extremely important to communications over the path.  Almost always, higher is better.  It is not uncommon for the KH6HME signals to be very loud with absolutely nothing at QTHs near sea level.  The same is generally true for the California end. Stations with elevation are generally the first, and sometimes the only ones, to copy the beacon or QSO KH6HME when there is an operator present.  I don't think a single number can be applied to the difference just as a blanket statement as to the best frequency can not be made. I think often 2m is the strongest but that isn't always the case. Even this time, I think I saw Gary, WA6MEM mention that the 70cm CW beacon was always stronger than KH6HME's 2m signal.

Regarding the "best mode" for this beacon, I certainly don't know the answer.  There is a history of CW only and certainly many of those monitoring for HI/North America tropo openings may be limited to decoding CW.  In bringing up the prototype 70cm beacon I'm building, I've run it locally at 1 microwatt with several different modes going. This has included WSPR-2, CW, JT65 and JT9-1. I added  a little telemetry into the CW message (20 WPM) and asked WW6D to see if he could copy it. Doug is a skilled CW operator and at 1 microwatt he was spotting the WSPR at S/N=-9 dB.
That's somewhere around the limit of a skilled ear and  represents an equivalent "brain-filtered" CW bandwidth of around 300 Hz.  Perhaps some operators can do slightly better but 200-300 Hz effective bandwidth seems a reasonable number for an accomplished CW operator.  In the same situation, WSPR can clearly go down at least another 20 dB, WSPR-15 could probably go almost 30 dB lower. 

The KH6HME CW beacon presently has deliberate chirp on the signal.  It  IDs and sends a little bit of carrier with the total signal going from about 144.277800 to about 144.778300. It won't even fit in a 300 Hz filter. I think the chirp is there for legacy reasons. People liked to recognize it is as KH6HME and not a birdie.

The system I have built up is based on Ultimate3 QRSS/WSPR capabilities.  With current code that can support

QRSS mode (plain on/off keyed slow CW)
- FSK/CW mode (frequency shift keyed slow CW)
- DFCW mode (dual frequency CW)
- WSPR mode (Weak Signal Propagation Reporter)
- WSPR-15 experimental WSPR mode with 15-minute frames
- Opera modes (8 speeds)
- PI4 beacon mode
- JT9 modes (5 speeds)
- Slow-Hellschreiber (frequency shifted slow-Hell)
- Full-speed Hellshreiber
- Half-speed ("DX") Hellshreiber
- CW (plain CW)
- FSK (0-999Hz shift, fast-speed FSK CW)
- Customisable FSK patterns
- manually-keyed CW/FSK transmitter
- JT65 modes
- ISCAT-A and B

I've only tried three of these so far but once it is in place, it isn't difficult to change the selections and frame to include more or fewer of these.  I don't yet know which of these will actually decode best across the path. One of my reasons for asking Wil for the .wav file was to be able to closely examine the spectrum of a strong-ish signal. I wanted to see if there was any spectral spreading due to Doppler shift from the effective path length changing that might cause problems for some of these modes. At N6GN I've heard the CW beacon many times and worked CW/SSB over this path on 2m and 70cm  and wasn't aware of anything 'strange' but I'd also not had a chance to look closely, within fractions of a Hz, of a signal to see if the path was modulating the signal.  I still haven't studied the files Wil sent closely but from what I've seen so far and from what Leigh VK2KRR reports for long tropo paths across the Great Australian Bight, nothing like that appears to occur. We really don't know.

The point of this is to give the possibility of improving understanding of this long tropo path.  It needs to support the function of the existing 70cm beacon, thus it will certainly run CW as part of its repertoire.  It can also run any of the above modes. WSPR is of particular interest to me because of the relative ease in decoding, the already moderately large base of 2m WSPR stations on the coast of North America and the worldwide Internet database.  All of this might serve to improve not only the detection but the communications about this interesting propagation among all concerned. Presently the real-time discussions have been somewhat limited. WSPR might be an addition and improvement on this. We might indeed learn more than we now know.

But for the original detection function, a decode of *any* digital mode is not essential. Currently, it's the presence of a signal on the correct frequency and beam heading that serves to alert most userss. A dead carrier would likely serve that, though there is I suppose still the "birdie" question and legality without an ID of some kind. To an "analog" ear, A WSPR or JT9 signal is close enough to a steady carrier that most users may not even notice.  With JT65 thrown in, there is a sort of replacement for the "musical" chirp of the present 2m beacon. It should be unique enough that people listening without the ability to decode a digital mode will have no doubt that they are hearing a signal from the KH6HME site, even before or if they don't copy the CW portion of the transmission. 

My hope is that with the flexibility of the replacement beacon we might get even more interest and learn even more. 

I have a plan in mind for a different approach to add functionality and bands, should this initial 70 cm attempt prove fruitful. In any case, I expect things to change and don't yet have the answers as to which way is best.

Glenn n6gn

--
You received this message because you are subscribed to the Google Groups "2 Meter WSPR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 2-meter-wspr...@googlegroups.com.
To post to this group, send email to 2-mete...@googlegroups.com.
Visit this group at https://groups.google.com/group/2-meter-wspr.
For more options, visit https://groups.google.com/d/optout.

[Leigh VK2KRR]

Bo,

As Glenn has mentioned, it might seem to be the same path, but with this tropo path the height differences at the Hawaiian side that Glenn mentioned, play a significant role in signals being heard or not.

A quote taken from a paper located on my website http://www.vk2krr.com/AR%20Article%20Bight%20propagation%20Mch%202001.doc
"The Hawaii/California path appears to depend significantly on the height of the operator at the Hawaiian end.  The report by Overbeck[i] and reinforced by later reports of contacts over this path, pointed out the importance of being between 1500 and 2600 m above sea level in Hawaii.  Contacts from points at lower altitudes are much less frequent.  Overbeck suggested that the maximum height of the duct at the California end was about 450 m.  Such observations were repeated by Tynan[ii] and Pocock[iii]."

 

I am surprised and happy to see that KH6IMB was even heard at all due to his location, so this is quite exciting.

Comparison tests will be able to be performed when there is a co located WSPR beacon at the same site as the KH6HME beacons. Should be good.
Across this path at 2m or 70cm, as the path is so long there is no chance of other propagation modes like AU, aircraft, meteors etc. As it is a stable tropo path that is being observed, and particularly because it exceeds the distance of these other prop modes which can cause doppler issues, WSPR will work fine.

Leigh VK2KRR

[Bo, OZ2M]

Hi

OK about two different paths. I don't have much experience myself with higher altitude locations like mentioned. Here in Europe paths like the ones you have CA <-> HI, i.e. long and across hot water, are rare. Canary Islands to south Ireland and south west England is closest however lacks the altitude I believe. This being said my understanding is that altitude is not always an advantage when it comes to sea ducts at last not seen across the North Sea. Furthermore, may e.g. 23 cm exhibit louder signals than 2 m due to the size of the duct. In any case I would imagine that the characteristics of a very long path like CA <-> HI is made up of a multitude of ducts resulting in "some" distortion of the signal. This doesn't favor a clean signal thus signal/decoder robustness may be relevant even for a "loud" signal. Therefore a less sensitive modulation may have a higher link probability than a more sensitive yet less robst modulation may have. So the conclusion is probably to try out a multitude of modulations preferably simultaneously.

An article worth reading is an article by Ray, W2RS, about the capabilities of the human ear http://rudius.net/oz2m/ngnb/thehumanear_raysoifer_w2rs.pdf. However, for automated monitoring this doesn't help much :-)

If the S/N number(s) is the only thing that matters then JT9-30 and Opera4H is the solution. However, I will challenge that. Previously I have made this page http://rudius.net/oz2m/ngnb/sequence.htm for comparing the different sequence/modulation possibilities. I think it is a good idea to include CW because the partial/fast ID and it is quite robust too. Indeed different modulations vs band may prove successful and perhaps even vs time of the year. It does make it a bit more tricky for the listeners. A sequence like this https://www.qrz.com/db/IT9X may not be beneficial for anyone. Worth taking into account is also who is the audience? Is it the analog DXers, mixed mode DXers or digital DXers.

Bo

[n6gn]

CW beacons are making it to the mainland, a few QSOs and KH6IMB is showing on the WSPR map so I've turned toward HI with WSPR.
Glenn n6gn

[Glenn Elmore]

Interesting waterfall. Note the few seconds of signal at 22:55.

Only N3IZN was reported as transmitting at this time.   I think this may be some kind of ACS backscatter possibly related to  tropo over the Pacific.

Glenn n6gn

On 07/26/2016 03:42 PM, Glenn Elmore wrote:

Obviously there is a path to the south. :-)

2016-07-26 22:38

N3IZN

144.490547

-12

0

DM13ji

100

KH6IMB

BK29jk

4061

257

Maybe elevation is too low up here.

Glenn

[n3...@aol.com]

Thanks for the heads up.

2016-07-26 22:54

N3IZN

144.490547

-24

0

DM13ji

100

KH6IMB

BK29jk

4061

257

2016-07-26 22:42	KH6IMB	144.490424	-19	-1	BK29jk	50	N3IZN	DM13ji	4061	60

2016-07-26 22:38

N3IZN

144.490547

-12

0

DM13ji

100

KH6IMB

BK29jk

4061

257

[wtan...@mac.com]

Started the monitoring protocol.

432? Sorry.. Don't ask! At least it came out of the box and the boom is mostly assembled!

[Glenn Elmore]

Hey, that's a start!

When you have it on, let us know. I'm particularly interested to know
if you can see any sign of the KH6HME 70cm beacon on a waterfall
(assuming it is too weak for your ears).

Glenn n6gn

[n3...@aol.com]

I don't hear or see the beacon, but I started TXing at 23:28 will let it go for an hour or so. WSPRnet is down but if you see me or decode me let me know on this reflector.

I was decoded in VK land on 6 again, so I'll switch back to 6 at 02:00 or so unless I hear the beacon or get an email.

Chris

--
You received this message because you are subscribed to the Google Groups "2 Meter WSPR" group.

To unsubscribe from this group and stop receiving emails from it, send an email to 2-meter-wspr+uns...@googlegroups.com.
To post to this group, send email to 2-meter-ws...@googlegroups.com.

[Glenn Elmore]

I was aiming my comment to KH6IMB. I wanted to know if he could copy KH6HME at his QTH which is very much not line of sight to Volcano
No implication that the path was in to CA.
Glenn n6gn

To unsubscribe from this group and stop receiving emails from it, send an email to 2-meter-wspr...@googlegroups.com.
To post to this group, send email to 2-mete...@googlegroups.com.

[n3...@aol.com]

No worries Glenn, I was aiming at KH6IMB as well. There are reports of the beacon being copied on the tropo ducting reports page, by the time I got home all was quiet. Thought I would give it a try.

I keep looking for the beacon once I see nothing on WSPR and nada on 144 or 432, so looks like we missed that opening.

AIS reports out to 600+ km. so maybe.

Switching back to 6.

Chris

-----Original Message-----
From: Glenn Elmore <n6...@sonic.net>
To: n3izn <n3...@aol.com>
Cc: 2-meter-wspr <2-mete...@googlegroups.com>
Sent: Mon, Aug 15, 2016 6:18 pm
Subject: Re: {2 Meter WSPR} Re: Hawaii <--> Mainland Tropo

I was aiming my comment to KH6IMB. I wanted to know if he could copy KH6HME at his QTH which is very much not line of sight to Volcano
No implication that the path was in to CA.
Glenn n6gn

[n3...@aol.com]

I have a 2.4 Ghz AREDN / HamMesh node and I'm seeing the Qualcom club's node come in some 35 miles away. Strong enough for a WSPR decode, if some one was to run that.

[n3...@aol.com]

The 144 beacon was in S3 this morning. Though no noise associated with it. So I put it on 144 WSPR pointed to HI. Don't know about the website yet.

Chris 

Sent from AOL Mobile Mail

[Glenn Elmore]

Chris,

When you wrote noise I first interpreted it as "noise on the tropo forum".  But from your posting there

S3 but none of the hash I usually hear when the beacon is in

I think you are talking about something that you often see coincident with a duct. 

Can you describe this? This is the first I've heard of it.

Glenn n6gn

[n6gn]

Yesterday, Feb 16 2017, Fred Honnold, K6IJ/KH7Y,  successfully installed the QRPLabs U3s-based 70cm multimode beacon into service at the KH6HME site at 8200' ASL elevation on Mauna Loa on the Big Island of Hawaii.  This beacon replaces the previous CW-only beacon and adds WSPR, JT9 and JT65A modes to the transmissions. This is the same hardware that we have been testing since last August
 at multiple locations including N6GN, WA6M, K6PZB, WW6D and WB7ABP. All functions seem nominal and Fred measured around 25 watts into the 432 coax as expected.  Schedule and frequency is shown in the attached image.

With the added weak signal capabilities of the K1JT modes, it is hoped that this may allow a roughly 20 dB earlier indication of duct formation and may give additional insight into the physics of the duct itself. Combined with the West Coast WSPR stations and the round-the-clock spotting provided into the the WSPRnet.org database DX conditions may be discovered earlier and more often than before.

A description of the beacon, upconverter, PA and U3s programming is being maintained on the KH6HME 70cm Beacon pages. 

-de n6gn

[n6gn]

I'm happy to report that N3IZN has just posted successful reception of the first WSPR spot from new KH6HME multi-mode 70cm beacon into the WSPR database:

 2017-03-15 23:40 KH6HME 432.301532 -28 0 BK29 20 N3IZN/RX DM13ji 4043 60 

As far as I know, this marks the longest 70cm WSPR spot ever recorded, as of this date.  Certainly longer QSOs have been achieved but not WSPR.  Given the weak signal, -28 dB S/N, it would seem that only WSPR would have been capable of making the path under these conditions.  Hopefully WSPR and this beacon will serve as an early warning of duct formation that will be useful to many stations and many modes in the future.

See the  KH6HME 70cm Beacon pages for more detail of the beacon.  I believe actual transmitted frequency is within a few tenths of a Hz of 432.301530 but until Chris describes the expected accuracy of his station it is not possible to determine if the 2 Hz error is path induced or just accuracy of an LO/clock at N3IZN.

Good job Chris. Congratulations!

For reference, here is Hepburn mapping for the Eastern Pacific near the time of the spot.

A short write-up of the beacon should appear in the May 2017 QST, World Above 50 MHz column. 

Glenn n6gn

[Leigh VK2KRR]

Well done guys, excellent!! Im keeping my eye on things.

Leigh VK2KRR

[n6gn]

And another,   2017-03-30 03:30

KH6HME

432.301539

-22

0

BK29

20

N3IZN/RX

DM13ji

4043

60

 

in spite of Hepburn showing only