Dick, > It will be interesting to see how your results alter with version 0.3.2.0 and > future > versions when released.> Dick G4BBH
Unfortunately, I haven't figured out a way to test ARQ modes like Winmor without a more elaborate setup than the sound card path simulator. I can only record / playback and test Winmors connect request mode. Two PC's and an outboard HF simulator between them would be useful for this sort of thing. It would allow one to test other modes like ALE-400 ARQ. I should be able to use virtual audio cable to record my own signals straight from the sound card during a live Winmor connect / file transfer and that would allow me to capture the different modes as Winmor cycles through them. I can then use those one-sided signals to see how they perform with various HF path simulations. The problem is that Winmor might not acknowledge those signals, but the constellation score might. That should give some indication of how good or bad throughput might be. Tony -K2MO ----- Original Message ----- From: "ferrymanr" <g4...@btinternet.com> To: <digitalradio@yahoogroups.com> Sent: Tuesday, September 29, 2009 6:43 AM Subject: [digitalradio] Re: Winmor Path Simulaitons >A new version of RMS Express has just been released (29 September) and may >affect things considerably once users have updated as there are several >changes. > > > --- In digitalradio@yahoogroups.com, Tony <d...@...> wrote: >> >> All, >> >> For what it's worth, I ran Winmor "connect request" signals through Moe >> Wheatley's path simulator. These tests are usually pretty good indicators of >> how digital modes perform and compare over a real HF channel. >> >> That said, the results of these tests may not reflect Winmors real potential >> so take them with a grain of salt. Running the path simulator on the same PC >> as I did with a mode that's sensitive to sound card performance and CPU >> usage is probably not the best for optimum results. >> >> I haven't figured out a way to test Winmor other than the "connect request" >> PSK mode so these tests are at best, a glimpse of what to expect when >> connecting with other stations under a variety of conditions; sound card >> issues aside. >> >> I should point out that some of these HF path simulations are severe with >> enough signal distortion to make one slow to a snails pace, even on CW! >> >> Tony -K2MO >> >> _________________________________________ >> >> >> Minimum signal-to-noise: >> >> The minimum SNR (white noise test) was approximately -6db / 3KHz BW. About >> the same sensitivity as PSK63. >> >> +++++++++++++++++++++++++++++++++ >> >> HF Path Simulation Tests: >> >> In the following tests, the path simulators signal-to-noise ratio was set a >> fixed level of +6db. This produced a Winmor connect request score of 85% >> with 100% decode (no HF path distortion). >> >> +++++++++++++++++++++++++++++++++ >> >> Selective fade performance: >> >> In this test, the simulator produces a deep notch that randomly sweeps >> across the spectrum simulating multipath conditions. The mode connected 75 >> percent of the time. >> >> Simulation: Selective Fading >> Path delay: Path #1 (0) Path#2 - 0.5ms, Path#3 - 0.6ms >> Frequency spread: Path #1 0.5Hz, Path #2 0.55Hz, Path #3 0.6Hz >> >> CONNECT REQUESTS: 75% DECODE >> >> ++++++++++++++++++++++++++++++ >> >> This High Latitude path test simulates a moderate disturbance over the polar >> ionosphere. The 10Hz signal spread causes the majority of digital modes to >> fail. It takes a slow MFSK mode like Olivia to get through. >> >> Simulation: High Latitude Moderate >> Path delay: 3ms >> Frequency spread: 10Hz >> >> NO DECODE >> SCORE - 20's >> >> +++++++++++++++++++++++++++++ >> >> This test simulates mid-latitude HF circuits during severely disturbed >> conditions at mid-latitude. Winmor connects were close to 100%. >> >> Simulation: Mid-Latitude Disturbed >> Path delay: 2ms >> Frequency spread: 1Hz >> >> CONNECT REQUESTS: 90% DECODE >> SCORE: 80's. >> >> ++++++++++++++++++++++++++ >> >> This low Latitude path test simulates a highly disturbed Spread-F paths that >> occur near the equator. Once again, 10Hz signal spread / timing delays cause >> the majority of modes to fail. >> >> >> Simulation: Low-Latitude Disturbed >> Path Delay: 6ms >> Frequency Spread: 10Hz >> >> NO DECODE / SCORE - 20's >> >> ++++++++++++++++++++++++++ >> >> Winmor connected near 100% with the less severe low-latitude path. >> >> Simulation: Low-Latitude Moderate >> Path Delay: 2ms >> Frequency Spread: 2Hz >> CONNECT REQUESTS: 90% DECODE >> SCORE: 70's >> > > >
