My $0.02 worth. I have the exact same GPS Hasan has. I’m also running Meinberg. I have the advantage of running Meinberg with internet time pools alongside the GPS. While the GPS show significant jitter, its availability to Meinberg consistently yields <5ms offset from time.is. Without the GPS, offset from time.is will run to 10-15ms. I realize that is not a problem for WSJT-X, but my point is that even with jitter, the UPS GPS does improve my time accuracy, even if only a smidgen (note the techno speak ☺ )
__________ Dan – K4SHQ From: Roeland Jansen [mailto:roeland.janse...@gmail.com] Sent: Wednesday, July 3, 2019 10:07 AM To: WSJT software development <wsjt-devel@lists.sourceforge.net> Subject: Re: [wsjt-devel] clock patch be aware of that if a) is 500 ms ahead and b) is 500 behind, the time skew is already 1 second between the signals. Anything more than a few ms is not good imho. Anyways, use whatever deems suitable for you. On Wed, Jul 3, 2019 at 4:33 PM Hasan al-Basri <hbasri.schie...@gmail.com<mailto:hbasri.schie...@gmail.com>> wrote: Yep, I'm aware of the potential latency issues, again, 500 ms is quite acceptable for the X suite...and I see far better than that. I haven't seen anywhere close to 500 ms errors against GPSDO (not usb) with my USB GPS. ..and that's with days of data collected and subject to statistical analysis. See the attached graphic. Not perfect, but way better and way cheaper than most other alternatives, especially if stuck away from home. 73, N0AN Hasan On Wed, Jul 3, 2019 at 9:16 AM Roeland Jansen <roeland.janse...@gmail.com<mailto:roeland.janse...@gmail.com>> wrote: the issue is not the device but the USB protocol -- it's well known for latency issues. It's super that it works for you. We have in .nl a few synchronized receivers and transmitters for repeaters and no way USB can be used there. But hey -- if it works for you, it's ok. But be aware of the mentioned issues. And 500+ms wrong times are seen more. I would not even consider it myself. On Wed, Jul 3, 2019 at 3:57 PM Hasan al-Basri <hbasri.schie...@gmail.com<mailto:hbasri.schie...@gmail.com>> wrote: Not all USB based GPS have issues. Not all software is the same. NMEATime2 does not show the issues you describe with a USB GPS. while the BKT software does. (for me) This one, from Amazon works. I have two of them. GlobalSat BU-353-S4 USB GPS Receiver (Black) The above USB based GPS used with NMEATime2, is certainly accurate enough for all the WSJT-X suite of programs....and many of us have the DT to prove it. Please don't throw the baby out with the bath water, with a pursuit of unnecessary perfection. 73, N0AN Hasan On Wed, Jul 3, 2019 at 3:07 AM Matthew Miller <mmill...@mail.umw.edu<mailto:mmill...@mail.umw.edu>> wrote: Beware those GPS receivers – I got one (unstable latency was throwing me off by as much as 550mS on NTP) and I found that when I tried to set up ntpd to use GPS it was showing jitter of around 0.8 seconds. Further research, USB GPS is a BAD IDEA because there is often buffering in the serial-USB converter chips as well as overhead with the USB protocol. Depending on what port you use it may also be connected thru one or more hubs adding additional overhead and latency. You may not “see” a hub but many computers now have internal USB hubs for touchpad, keyboard, webcams, touchscreens, Bluetooth, and a variety of other “built in” features that now all hang off internal hubs. If you are going to go with a proper GPS time sync you really need a non-USB serial GPS and PPS input…which is not trivial. Its not cheap, but I solved my problem with one of these…its plug and play then just set your computer to point at its IP address. I now have a reliable source that indicates it has only 0.5mS jitter and 3mS latency even going thru multiple network switches and a fairly busy WiFi network. https://timemachinescorp.com/product/gps-time-server-tm1000a/ There are also other cheaper DIY Raspberry Pi projects (which have some of their own challenges I didn’t feel like dealing with right now) that can be built for under $100 to make a stratum-1 time server that works without Internet but PLEASE PLEASE do not tell people to use USB GPS for time, its going to increase the problems not fix them! -Matt / KK4NDE From: Neal Pollack [mailto:nea...@gmail.com<mailto:nea...@gmail.com>] Sent: Wednesday, June 26, 2019 12:08 PM To: WSJT software development Subject: Re: [wsjt-devel] clock patch Only some are due to lack of internet. USB GPS receivers are now approx $12 USD on Ebay. I have tested a few and they work fine with NMEAtime2. Which is none simple software. The message could suggest what to do for both internet connected and off-grid users. But really, to keep the message short, it should point to a help page or file that is specific to time setting options, rather than a huge user guide, since the public no longer reads very much. With regard to time setting software, I really can't recommend BKTtimeSync for the same reasons we are discussing this issue: It works for me but is far too complicated for the average ham, and lacks sufficient docs or help. But NMEAtime2 is really simple and just works, for using GPS. For internet users, NetTime or Dimension4 for basic users or Meinberg NTP for more advanced users. BUT!!! All of the users, both internet and GPS, need to be reminded to turn OFF windows automatic time setting before using the software's above. I see too many users whose DT is swinging wildly between plus and minus 2.5 seconds, then next at 0.2 DT. This is because they have windows trying to set time while also using the above programs to set time, and the two are fighting a tug of war. If you guys put a DT warning message in the code that points to a time help page, then I volunteer to be one of several editors for the above type of help content about time sync'ing options. Cheers, Neal N6YFM On Wed, Jun 26, 2019, 3:16 AM <wsjt-devel-requ...@lists.sourceforge.net<mailto:wsjt-devel-requ...@lists.sourceforge.net>> wrote: Send wsjt-devel mailing list submissions to wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> To subscribe or unsubscribe via the World Wide Web, visit https://lists.sourceforge.net/lists/listinfo/wsjt-devel or, via email, send a message with subject or body 'help' to wsjt-devel-requ...@lists.sourceforge.net<mailto:wsjt-devel-requ...@lists.sourceforge.net> You can reach the person managing the list at wsjt-devel-ow...@lists.sourceforge.net<mailto:wsjt-devel-ow...@lists.sourceforge.net> When replying, please edit your Subject line so it is more specific than "Re: Contents of wsjt-devel digest..." Today's Topics: 1. Re: Clock patch (DG2YCB, Uwe) 2. Re: Clock patch (Roeland Jansen) 3. Re: Clock patch (Bill Somerville) ---------------------------------------------------------------------- Message: 1 Date: Wed, 26 Jun 2019 11:36:58 +0200 From: "DG2YCB, Uwe" <dg2...@gmx.de<mailto:dg2...@gmx.de>> To: "'Black Michael'" <mdblac...@yahoo.com<mailto:mdblac...@yahoo.com>>, "'WSJT software development'" <wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net>> Subject: Re: [wsjt-devel] Clock patch Message-ID: <004801d52c02$b33f1430$19bd3c90$@gmx.de<http://gmx.de>> Content-Type: text/plain; charset="utf-8" I?ve not yet seen the results, but your this approach really looks like something which could bring a substantial progress regarding the timing issue. Reason why I am supporting a warning message is that, besides field days, I saw dozens of OMs/YLs who are not aware at all that system time of their PCs needed to by synced. Alone two examples yesterday. Contacted both. Seemed to be both are ?old? men, with no computer skills. One replied the following: ?I am no longer a very young man (84 years old) who tries to tamper with the new methods; with a total ignorance of the English language??. If we like it or not: That?s the reality! Of course we could let FT8, FT4, etc. be something only for the tiny community of well-educated, scientifically trained OMs with good to excellent computer skills. But if we want to make QSOs with more OMs/YLs, then WE need to do something which helps THEM to overcome the hurdles. 73 de Uwe, DG2YCB Von: Black Michael via wsjt-devel [mailto:wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net>] Gesendet: Mittwoch, 26. Juni 2019 06:19 An: WSJT Software Development Cc: Black Michael Betreff: [wsjt-devel] Clock patch This patch shows a message if 80% of DT values are >= 0.5 seconds. https://www.dropbox.com/s/8j7qao9wa4wmzdm/clock.patch?dl=1 I picked the 50% point of FT4's 1 second tolerance but perhaps a slightly larger value might be appropriate. I think most, if not everybody, can get the > 20% < 0.5 seconds without a problem. Even with my rather extreme latency of 300-400ms I don't trigger this. de Mike W9MDB -------------- next part -------------- An HTML attachment was scrubbed... ------------------------------ Message: 2 Date: Wed, 26 Jun 2019 11:55:47 +0200 From: Roeland Jansen <roeland.janse...@gmail.com<mailto:roeland.janse...@gmail.com>> To: WSJT software development <wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net>> Subject: Re: [wsjt-devel] Clock patch Message-ID: <caj2u8oytzmmcmcauqppc9d3hk+6jxw1nxnvht-5auv5t5wt...@mail.gmail.com<mailto:caj2u8oytzmmcmcauqppc9d3hk%2b6jxw1nxnvht-5auv5t5wt...@mail.gmail.com>> Content-Type: text/plain; charset="utf-8" I did not look at the patches Mike sent but: https://time.is/ gives you how good/bad the time is. If you start up: do check like this and refuse to go on if the timeskew is too large. Add a link how to fix in the message and you force people to fix it. (not that people read notices) just my bad idea. On Wed, Jun 26, 2019 at 11:41 AM DG2YCB, Uwe <dg2...@gmx.de<mailto:dg2...@gmx.de>> wrote: > I?ve not yet seen the results, but your this approach really looks like > something which could bring a substantial progress regarding the timing > issue. > > > > Reason why I am supporting a warning message is that, besides field days, > I saw dozens of OMs/YLs who are not aware at all that system time of their > PCs needed to by synced. Alone two examples yesterday. Contacted both. > Seemed to be both are ?old? men, with no computer skills. One replied the > following: ?I am no longer a very young man (84 years old) who tries to > tamper with the new methods; with a total ignorance of the English > language??. If we like it or not: That?s the reality! Of course we could > let FT8, FT4, etc. be something only for the tiny community of > well-educated, scientifically trained OMs with good to excellent computer > skills. But if we want to make QSOs with more OMs/YLs, then WE need to do > something which helps THEM to overcome the hurdles. > > > > 73 de Uwe, DG2YCB > > > > *Von:* Black Michael via wsjt-devel [mailto: > wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net>] > *Gesendet:* Mittwoch, 26. Juni 2019 06:19 > *An:* WSJT Software Development > *Cc:* Black Michael > *Betreff:* [wsjt-devel] Clock patch > > > > This patch shows a message if 80% of DT values are >= 0.5 seconds. > > > > https://www.dropbox.com/s/8j7qao9wa4wmzdm/clock.patch?dl=1 > > > > I picked the 50% point of FT4's 1 second tolerance but perhaps a slightly > larger value might be appropriate. > > > > I think most, if not everybody, can get the > 20% < 0.5 seconds without a > problem. > > > > Even with my rather extreme latency of 300-400ms I don't trigger this. > > > > de Mike W9MDB > > > > > _______________________________________________ > wsjt-devel mailing list > wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> > https://lists.sourceforge.net/lists/listinfo/wsjt-devel > -------------- next part -------------- An HTML attachment was scrubbed... ------------------------------ Message: 3 Date: Wed, 26 Jun 2019 11:15:52 +0100 From: Bill Somerville <g4...@classdesign.com<mailto:g4...@classdesign.com>> To: wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> Subject: Re: [wsjt-devel] Clock patch Message-ID: <4c971df3-778b-86c6-a947-49c4ed9d7...@classdesign.com<mailto:4c971df3-778b-86c6-a947-49c4ed9d7...@classdesign.com>> Content-Type: text/plain; charset=utf-8; format=flowed On 26/06/2019 10:55, Roeland Jansen wrote: > I did not look at the patches Mike sent but: > > https://time.is/ > > gives you how good/bad the time? is. If you start up:? do check like > this and refuse to go on if the timeskew is too large. > Add a link how to fix in the message and you force people to fix it. > > (not that people read notices) > > just my bad idea. Hi Roeland, doesn't that miss the point that many of the poor time sync situations are where the user has no Internet connectivity? 73 Bill G4WJS. ------------------------------ ------------------------------ Subject: Digest Footer _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/wsjt-devel ------------------------------ End of wsjt-devel Digest, Vol 64, Issue 164 ******************************************* _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/wsjt-devel _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/wsjt-devel _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/wsjt-devel _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net<mailto:wsjt-devel@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
_______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel