"50eroids" should read "50 periods", and "on periods" should read "non periods", fldigit should read fldigi.
Sorry - must be the wine - just got back from a family dinner! Skip KH6TY ----- Original Message ----- From: "kh6ty" <[EMAIL PROTECTED]> To: <digitalradio@yahoogroups.com> Sent: Thursday, November 27, 2008 8:19 PM Subject: Re: [digitalradio] Re: How Can We Push HF Emcomm Messages to the Field? > > >> If the signals are in the marginal range, how do you do the coordinating >> between the stations? To date, we have been able to use a cell phone. >> >> How do you calculate the error rate (such as the 6% mentioned)? > > We send 50eroids.(..............................................). > Anything > that is not a period is easily recognized as an error. Three "on-perionds" > equates to a 6% error rate. >> >> If I understand this correctly, the test was between a 5/8 vertical to >> quad for vertical polarization vs. quad to quad for horizontal, wouldn't >> it be about right to see 6 dB difference considering that you are >> increasing the path budget with the inclusion of the quad? > > The test was a vertically polarized quad to a 5/8 wavelength whip, and a > horizontally polarized quad to a horizontally polarized quad. The quad had > 7.5 dBi of gain versus perhaps 5 dBi of gain for the vertical whip. That > makes up about 2.5 dBi of the 6 dBi, and the rest is an approximation of > the > S/N as measured by flidit in both cases. As you can imagine, it is > extremely > difficult to make exact quantitative measurements under such conditions, > but > even modeling shows the 6 dB that Cebik references. Our experience is that > the 6 dB is about correct. >> >> Several ways to do this is with quad to quad vertical and quad to quad >> horizontal polarization, or some other gain antenna that can switch >> properly between polarizations. I wonder if you would see such a >> difference? > > Based on two different modeling programs, and our own simple tests, I > think > so. The most significant finding is that we lose communicaton over about > 30 > miles using vertical-to-vertical, but easily over 70 miles using > horizontal-to-horizontal, even though the horizontal antenna on the mobile > end is 5 feet higher than the whip is. In the end, anectodal evidence from > others also suggests a 15 to 20 mile range with vertical whips, and we > already know we can exceed 70 miles in flat country using a low, > horizontally-polarized quad instead of a vertical, and that is all that is > important to our purpose. It would be nice to have more and better > controlled tests, but you can just imagine the difficulty in arranging for > such tests without doing it on an antenna range. You have to switch > polarization on both ends, and one existing antenna may be on a tower, 50 > feet in the air. Of course, any such tests are possible, but the > difficulty > of finding people to participate is difficult, at best. As far as we are > concerned, together with the common knowledge that all weak signal > communications on 2m use horizontal polarization, TV stations use > horizontal > polarization because long ago it was found to be better for propagation, > and > the confirming results from modeling, are sufficient enough reasons to > insist on using horizontal polarization for distances longer than a > repeater > can provide. Add to that the probability that many existing vertical beams > are not mounted on rotators, and the change to horizontal polarization > appears to be well worth the effort, based on available information. You > can > also include the possibility that using a horizontally polarized quad > provides a lower takeoff angle close to ground that a yagi, and you can > see > why there are many reasons to insist on using horizontal polarization. > Finally, in a serious emcomm situation, NBEMS only needs to reach > connectivity with the Internet for email delivery or POTS for phone > delivery, so any available forwarding station will suit the purpose, > whether > a part of an organized emcomm effort or not. The need is only to get the > message to the EOC or other recipient, and all existing weak signal 2m > stations are using horizontal polarization. > > Our main interest is emcom messaging, and even a single dB of advantage > may > mean getting the traffic through or not, so we have use the best methods > at > our disposal, and the preponderance of evidence says that horizontal > polarization has an advantage over vertical polarization. > > 73, Skip KH6TY > NBEMS Development Team > >> >> 73, >> >> Rick, KV9U >> >> >> kh6ty wrote: >>> Hi Rick, >>> >>>> Have you found that DominoEX is the best overall digital mode for FM? I >>>> know that PSK modes can have doppler errors from aircraft, but >>>> otherwise >>>> seem pretty good for weak signal. >>>> >>> >>> Yes, definitely! DominoEx is a frequency shift keying mode, not a phase >>> shift mode, but doppler problems are still sometimes a problem, but not >>> nearly as much as on PSK31 or PSK63, so that is one reason why we now >>> use >>> DominoEx. Once the reflected signal arrives 180 degrees out of phase >>> with >>> the direct signal, it cancels out the direct signal for a while and >>> there >>> is >>> no mode that is going to print under that condition. The wider, >>> multitone >>> modes have less problem because the data is redundant and spread over >>> the >>> width of the signal, but even they are no completely immune. However, on >>> our >>> twice-weekly net, since we switched to DominoEx, the number of multipath >>> problems is considerably down, even on SSB. Initial tests suggest that >>> MFSK16 might even be better on FM, since it is the most sensitive mode >>> we >>> currently have with almost enough speed for messaging. It is completely >>> unusable on VHF SSB, though, because many transceivers in the field are >>> not >>> frequency-stable enough to stay tuned. On FM, the carrier frequency >>> sweeps >>> over the entire passband, so only the audio frequency stability is >>> important. DominoEx is especially valuable for drifting signals on SSB, >>> because it can tolerate mistuning of 50% of the signal width. The >>> IC-746Pro >>> and the FT-857D, if without an optional TCXO, just drift too much to be >>> usable on SSB, but are OK on FM, even though the S/N of FM is worse than >>> on >>> SSB. Note that any multipath cancellations simply cause repeated blocks >>> when >>> using ARQ, so they only slow down the transfer while the reflected >>> signal >>> is >>> moving across the direct signal. >>> >>> Last night on our net, we had positive confirmation of the better >>> performance of SSB over FM. The error rate between two stations was >>> running >>> at 6% on SSB, but when we all switched to FM, there was zero copy. The >>> fact >>> that there were any errors at all on SSB indicated that the stations >>> were >>> fringe area to each other, so it was a good demonstration of the >>> advantage >>> of SSB over FM. One station was beaming toward me and the other was 45 >>> degrees away from the beampath of that station. It was the same as if a >>> station with a high gain yagi were pointed away from me and even if I >>> pointed directly at him, he was not radiating enough energy in my >>> direction >>> for me to copy him. We have to make more tests, but I think the secret >>> of >>> the OptimizedQuad is that the pattern is bulbous instead of being >>> pencil-shaped - more like an omnidirectional pattern, but with gain over >>> a >>> wide beamwidth. Stacking OptimizedQuads vertically would increase the >>> gain >>> by 2.5 dB and still retain the wide beamwidth. It sure is interesting >>> stuff! >>> >>> >>>> Your point is well taken that many of the hams who participate in >>>> public >>>> service activities, may tend to be the younger ones who are Technician >>>> class and can mostly operate on 6 meters and up with their vertical >>>> antennas and FM only rigs. The number of hams with the >>>> multimode/multiband rigs is increasing, at least in our area. It is not >>>> easy to get them to try SSB, much less SSB digital though. >>>> >>> >>> I have found that the main problem is lack of VOX with the FM >>> transceivers, >>> which cost under $200 for a single band one, so you need to spend >>> another >>> $100 for a SignaLinkUSB interface in order to use macros to do the PTT >>> switching. The FM Transceiver Interface solves that problem for only >>> $10. >>> I >>> have built 10 of them which I will be giving out to the first few people >>> who >>> want to join the net but have only FM transceivers, but they also need >>> to >>> have an OptimizedQuad, or small yagi, horizontally polarized. >>> >>> >>>> The claim about the ground gain for horizontal antennas may be true but >>>> I have not seen this definitely tested. Have you done some comparisons >>>> with low 2 meter antennas, such a mobile to low base antenna with V and >>>> H and found H consistently better? I don't hold too much stock in >>>> software modeling and only would go with empirical data for that kind >>>> of >>>> test. >>>> >>> >>> I have done only one test so far, as it is difficult to arrange, since >>> both >>> stations have to switch polarization, but that first test did show a >>> huge >>> advantage using horizontal polarization. Range on FM between a 5/8 >>> wavelength whip mounted on a Prius and my quad turned for vertical >>> polarization was only 25-30 miles, depending on whether or not the >>> mobile >>> was clear of trees, but 70 miles was a piece of cake between the >>> OptimizedQuad and my own quad turned for horizontal polarization. We >>> could >>> have gone even farther if we had time. Next opportunity, we hope to be >>> able >>> to keep going. I am now more than convinced that the difference is real. >>> There was once a reference, which I cannot find, that found that a quad >>> near >>> a ground surface retains a low takeoff angle, but the takeoff angle of a >>> yagi of the same gain increases to as much at 40 degrees off the >>> horizon, >>> which means a quad may be the best choice for portable operation anyway. >>> >>> >>>> We will probably bite the bullet eventually and put a rotor back up on >>>> the low tower and maybe go with a Gulf Alpha 11 element V and H antenna >>>> for some reasonable gain. Then we could do the test. The ham that was >>>> going to help us lost his QTH and will not be able to relocate his VHF >>>> antenna farm. Of course they are quite high so maybe there would not >>>> have been as much difference in such a case. One of the best known VHF >>>> ops in my Section says that after running many tests he has never found >>>> either polarization is any different. But he has high antennas so maybe >>>> that accounts for it. >>>> >>> >>> Yes, high antennas are probably the reason. At seven wavelengths from >>> real >>> ground, the disadvantage to using vertical polarization over horizontal >>> drops from 6 dB at two wavelengths to only one dB at seven wavelengths, >>> but >>> portable stations or mobiles generally are not going to be able to get >>> have >>> antennas much higher than 2 wavelenghts. The jury is also out whether >>> horizontal polarization is an advantage over several hundred miles. I >>> will >>> not be able to test this until the coastal tropo scatter season comes >>> back >>> in the spring. >>> >>> If your yagi has more gain that you need, you can just rotate it 45 >>> degrees >>> and cover both polarizations, but with a 3 dB gain loss on both. >>> >>> >>>> We hope at least soon do some digital mode comparisons on 2 meters, >>>> whether SSB or FM. >>>> >>>> 73, >>>> >>>> Rick, KV9U >>>> >>> >>> That would be great! We need as much information as we can get, >>> especially >>> since lower South Carolina is quite flat, with no hills until you get to >>> the >>> upper part of the state. We do know this for sure - using a sensitive >>> digital mode with either SSB or FM greatly extends the range over using >>> phone, simply because the digital mode can copy under the noise level >>> and >>> phone cannot. The average modulation of a phone signal is only 30%, or >>> maybe >>> 50% with compression, but the passband needs to be over 2 KHz. With a >>> narrow >>> digital mode, the DSP filters in the software (and at IF if available) >>> can >>> be used to narrow the noise window by at least four times, improving the >>> S/N >>> by 6 dB or more and still use 100% average modulation for another 3 dB >>> or >>> more improvement in S/N. You simply cannot do this with phone and remain >>> intelligible, and you cannot use redundancy with phone as you can with >>> digital modes. >>> >>> >>> 73, Skip KH6TY >>> NBEMS Development Team >>> >>> >> >> >> >> Internal Virus Database is out of date. >> Checked by AVG - http://www.avg.com >> Version: 8.0.138 / Virus Database: 270.9.2/1785 - Release Date: >> 11/13/2008 >> 9:12 AM >> >> > > > > Internal Virus Database is out of date. > Checked by AVG - http://www.avg.com > Version: 8.0.138 / Virus Database: 270.9.2/1785 - Release Date: 11/13/2008 > 9:12 AM > >
